1. Kapalikas-and-Kalamukhas-Two-Lost-Saivite-Sects-by-D-Lorenzen MLBD

THE KĀPĀLIKAS AND KĀLĀMUKHAS

TWO LOST ŚAIVITE SECTS

David N. Lorenzen

MOTILAL BANARSIDASS PUBLISHERS PVT. LTD. DELHI

First Edition : 1972 Second Revised Edition: Delhi, 1991

C D.N. Lorenzen 1972

ISBN: 81-208-0708-1

Also available at: MOTILAL BANARSIDASS 41 U.A., Bungalow Road, Jawahar Nagar, Delhi 110 007 120 Royapettah High Road, Mylapore, Madras 600 004 16 St. Mark's Road, Bangalore 560 001 Ashok Rajpath, Patna 800 004 Chowk, Varanasi 221 001

PRINTED IN INDIA BY JAINENDRA PRAKASH JAIN AT SHRI JAINENDRA PRESS, A-45 NARAINA INDUSTRIAL AREA, PHASE I, NEW DELHI 110 028 AND PUBLISHED BY NARENDRA PRAKASH JAIN FOR MOTILAL BANARSIDASS PUBLISHERS PVT. LTD., BUNGALOW ROAD, JAWAHAR NAGAR, DELHI 110 007

CONTENTS Page LIST OF ABBREVIATIONS vii PREFACE TO THE SECOND EDITION ix PREFACE xi CHAPTER I FOUR ŚAIVITE SECTS 1 Brahma-sūtra Commentaries 1 Purāņas and Other Sources 7

CHAPTER II KĀPĀLIKA SOURCES 13 Early Sources 13 Kāpālika Epigraphy 24 Śamkarācārya and the Kāpālikas 31 Śamkara and Ugra-Bhairava 32 Śamkara and Krakaca 39 Śamkara and Unmatta-Bhairava 46 Kāpālikas in Sanskrit Drama 48 Miscellaneous Later Sources 62

CHAPTER III KĀPĀLIKA CULT AND DOCTRINE 73 The Mahāvrata 73 Somasiddhānta 82 Kāpālika Bhakti 83

CHAPTER IV KĀLĀMUKHAS OF THE ŠAKTI-PARIȘAD 97 Preliminary 97 The Mūvara-köņeya-samtati 99 Other Divisions of the Parvatāvali 130 The Bhujangāvali 138

CHAPTER V OTHER KĀLĀMUKHA PRIESTHOODS 141 The Simha-parişad 141 Other Kālāmukha Inscriptions 146 Shimoga District 146 Belgaum District 150 Bellary District 153 Bijāpur District 155 Dharwar District 157 Chikmagalur (Kadur) District 157 Chitradurga (Chitaldrug) District 158 Hassan District 158

Kolar and Tumkur Districts 160 The Kriyāśaktis of Vijayanagar 161 Andhra Pradesh and Madras 165 Kālāmukhas and Vīraśaivas 167

CHAPTER VI LAKULĪŚA AND THE PĀŚUPATAS 173 BIBLIOGRAPHY 193 INDEX 203 APPENDIXES 215

LIST OF ABBREVIATIONS

ABORI : Annals of the Bhandarkar Oriental Research Institute [Poona]. ARMAD : Annual Report of the Mysore Archaeological Department. BSOS : Bulletin of the School of Oriental Studies [London]. CII : Corpus' Inscriptionum Indicarum. EC : Epigraphia Carnatica. EI : Epigraphia Indica. ERE : Encyclopaedia of Religion and Ethics, ed. James Hastings. GSS : Gorakşa-siddhānta-samgraha, ed. G.N. Kavirāja. HAS : Hyderabad Archaeological Series. HDS : History of Dharmaśāstra by P.V. Kane. HTR : Harvard Theological Review. IA : Indian Antiquary. IHQ : Indian Historical Quarterly [Calcutta]. 11J : Indo-Iranian Journal [The Hague]. JAnSB \ Journal of the Anthropological Society of Bombay. JAOS : Journal of the American Oriental Society. JBBRAS : Journal of the Bombay Branch of the Royal Asiatic Society. JBRS : Journal of the Bihar Research Society. JGRS : Journal of the Gujarat Research Society. JIH : Journal of Indian History [Trivandrum]. JOIB : Journal of the Oriental Institute, Baroda. JORM : Journal of Oriental Research, Madras. JRAS : Journal of the Royal Asiatic Society [London]. JRASB : Journal of the Royal Asiatic Society of Bengal. QJMS : Quarterly Journal of the Mythic Society [Bangalore]. SBE : Sacred Books of the East, ed. F. Max Muller. SII : South Indian Inscriptions. TAS : Travancore Archaeological Series.

PREFACE TO THE SECOND EDITION

Although The Kāpālikas and Kālāmukhas has become recognized as a standard work on these two Saivite sects of medieval India, it has never had an adequate distribution in India, despite the fact that it was originally jointly published both in India and the U.S.A. by Thompson Press and the University of California Press respectively. Hopefully this new edition by Motilal Banarsidass will largely solve this problem. The new edition also gives me the opportunity to present two new appendices that summarize the discoveries relevant to these sects and the analyses of them that other scholars have made since the book's original publication in 1972. Except for a few minor corrections, the original text of the book remains intact. The new material described and discussed in the two new appendices has not been integrated into the bibliography and index, but extensive cross-references are included in the notes to the appendices as well as complete biographical citations. The photograph on the dust jacket of this new edition is of a standing sculpture of Lakulīśa found at Alampur, (Mehabubnagar, A.P.) 10th-11th century. The photo archives of the American Institute of Indian Studies in Ramnagar, Varanasi, kindly supplied the original photograph.

New Delhi David N. Lorenzen April, 1990

PREFACE

This study attempts to give as complete as possible a description of two extinct Saivite sects-the Kāpālikas and the Kālāmukhas. In a Christian context the concept of a 'sect' embodies three essential features : a specific doctrine (including a prescribed mode of worship), a priesthood, and a well-defined and exclusive laity. The structure of Hindu 'sects' is in general much more amorphous than that of Christian ones. In most cases more emphasis is placed on doctrine and mode of worship than on organisation. The Sanskrit words most often used for the Kāpālika, Kālāmukha and Pāsupata 'sects'-the groups discussed in this study-are darśana, samaya and mata. The basic meaning of these words is 'doctrine.' Each of the three groups also had its own priesthood. That of the Kālāmukhas appears to have been the best organised. Several Kālāmukha monasteries (mațhas), each under a single head (matha-pati), controlled temples in the regions surrounding them. It is doubtful, however, whether any of the three groups had its own exclusive laity. An ordinary farmer or merchant might have called himself a Buddhist, Jain, Vaișnava, or Saivite, but probably not a Kāpālika, Kālāmukha or Pāśupata. Records indicate that persons supported priesthoods of different and even hostile 'sects' without feeling disloyal. For this reason it might be more appropriate to speak of Kālāmukha, Pāśupata and Kāpālika 'monastic orders' rather than 'sects.' Since, however, the term 'monastic order' does not usually imply a separate doctrinal or philosophical position, we will remain content with the word 'sect.' Unfortunately no religious texts of either the Kāpālikas or the Kālāmukhas have survived. Their portraits must be drawn from accounts by their opponents and, in the case of the Kālāmukhas, from the information contained in epigraphic grants to their temples. The comments on both sects by Yāmunācārya and his famous pupil Rāmānuja make the best starting point. Many of the remarks by these two Vaisnava sages about the Kāpālikas are confirmed and enlarged by the numerous descriptions of Kāpālika ascetics in Sanskrit literature. Of particular importance are two dramas-Bhavabhūti's Mālatī-Mādhava and Krșņamiśra's Prabodhacandrodaya-and two legendary accounts of the life of Śamkarācārya-Mādhavācārya's Samkara-digvijaya and Ānanda- xi

xii PREFACE

giri's Śamkara-vijaya. Although nearly all of the sources for the Kāpālikas are fictional and written from a hostile point of view, the overall picture they give is detailed enough and consistent enough to ensure that it is reasonably authentic. The discovery of two or three inscriptions from what must have been Kāpālika temples at least guarantees their existence. Apart from the remarks of Yāmunācārya and Rāmānuja, the sources for the Kālāmukhas are nearly all epigraphic ones. The majority of the grants to Kālāmukha temples have been found in what is today Mysore State and date from the eleventh and twelfth centuries A.D. Most are written in Kannada, the language of the region. In general they provide more information about the history than the doctrine and cult of the sect. The religious information that they do contain tends to discredit rather than corroborate Yāmunācārya and Rāmānuja. Most importantly, the records indicate that the Kälämukhas were an offshoot of the Pāśupatas, a sect about which a good deal is known from surviving religious texts as well as from inscriptions. Few modern scholars have paid much attention to either the Kāpālikas or Kālāmukhas. One of the earliest reputable discussions of the two sects is in R.G. Bhandarkar's Vaisnavism, Saivism and Minor Religious Systems (1913). Bhandarkar limits himself to a brief summary of the accounts of Rāmānuja, Mādhavācārya, Änandagiri, and Bhavabhūti. Although he admits that 'there appears to be a confusion between the sects of Kāpālikas and Kālāmukhas' in Rāmānuja's account, Bhandarkar seems to accept that the Kalamukhas were 'the most extreme sect.' This view, which has been accepted by many subsequent writers, is, I feel, an incorrect one. For a truer picture of the Kālāmukhas one must look to their inscriptions. No comprehensive study of Kālāmukha epigraphy has so far been attempted, but worthwhile discussions of the Kālāmukha Śakti-parișad at Belagāve are found in J.F. Fleet's 'Inscriptions at Ablur' and in A.V. Subbiah's 'A Twelfth Century University in Mysore.' A great number of Kālāmukha inscriptions are edited and translated by B.L. Rice in Epigraphia Carnatica. Other inscriptions have been edited, and often translated, by various scholars in Epigraphia Indica, South Indian Inscriptions, Indian Antiquary, and other journals. For the Kāpālikas only one modern scholar merits special mention-K.K: Handiqui. He devotes several pages to the sect in his brilliant study of the tenth century background to Somadeva's Yaśastilaka and in a note on

PREFACE xii

Somasiddhānta in his translation of Śrīharșa's Naișadhacarita. I have included a fairly comprehensive list of the many modern studies of the Päsupatas at the beginning of chapter six. In the present study I have attempted to gather together for the first time all the available source materials on the Kāpālikas and Kālāmukhas and to extract a coherent account of their history, doctrines and religious practices. With the Kāpālikas I have first presented these source materials in as readable a form as possible and saved most of my analysis of them for a separate chapter. This arrangement has necessitated a good number of repetitions of important references for which I ask the reader's indulgence. The reconstruction of Kāpālika cult and doctrine is admittedly speculative owing to the distorted and fragmentary character of the evidence. For this reason I have had to repeatedly qualify my remarks with words and phrases such as 'probably,' 'possibly,' 'perhaps,' `seems to,' 'tends to suggest,' etc. Of greatest importance is the identification and des- cription of the peculiar vow of the Kāpālikas called the Mahāvrata. It is this vow, I believe, that provides the key to a proper under- standing of many of their unorthodox ascetic practices. The subject of tantric religion is potentially a rather controversial one, and some of my comments might raise the hackles of those concerned for the image of Indian religion. The axes I have to grind do not include the wilful denigration of things Indian, and I have tried at all times simply to draw the most reasonable conclusion the evidence afforded. The presentation in a readable form of the profuse epigraphic evidence on the Kālāmukhas was a more difficult task. My main object has been to demonstrate the great importance of this sect in tenth to thirteenth century Mysore and to rescue it from the tantric limbo to which it was relegated by Rāmānuja, R.G. Bhandarkar and others. The last chapter discusses the Pasupata ancestry of the Kālāmukhas, particularly the date and life of the Pāsupata- Kālāmukha saint Lakulīśa. Many of my remarks are in the nature of criticisms and cautionary notes about the conclusions of modern scholars. Included, with some trepidation, is a criticism of Professor Ingalls' theory of the shamanistic origin of this sect. Several technical details must be mentioned. The critical apparatus I have used is based for the most part on K.L. Turabian's A Manual for Writers of Term Papers, Theses and Dissertations. Brackets

xiv PREFACE

are mainly used for my additions to translations of other scholars. Additions to my own translations are put in parentheses. The words 'Saivite' and 'Saiva' are distinguished. The former is applied to anyone who specially worships the god Siva, the latter to a Śaivite who follows Saiva-Siddhänta. The spelling of Indian place names is always a problem. In general I have used the versions given in the Government of India's Road Map of India (2d edition). For the names of small villages and other places not on this map, I have normally used the spellings given in my sources. Sanskrit words are transliterated according to the system now used by most Indologists. Nasal-consonant combinations are transliterated as in Monier-Williams' Sanskrit-English Dictionary. Thus I have written 'Samkara' not 'Sankara.' In order to avoid confusion-especially between 'c' and 'ch'-I have also made uniform the spelling of Sanskrit words in quotes and translations of other scholars, though not in the titles of their books and articles. Translations from Sanskrit are my own unless otherwise specified. Since my knowledge of Kannada is more limited, I have normally relied on the translations of Rice, Fleet, Barnett, and others for the Kälämukha epigraphy in this language. I have also had the help of Dr. H. Ullrich of Michigan State University and Professor H.S. Biligiri of Deccan College, Poona. I am particularly grateful for Professor Biligiri's excellent translation of the important record describing the exploits of Bonteyamuni of Hombal. Several other persons have contributed suggestions, criticisms and linguistic assistance. I would specially like to thank Mr. Venugopalan of Deccan College; Professor J.W. de Jong, Dr. S.A.A. Rizvi and Dr. K.H.J. Gardiner of the Australian National University; and Dr. T.R. Trautman of the University of Michigan. For instruction and encouragement in the earlier stages of my study of ancient India I am indebted to Professor J.W. Spellman of Windsor University and to Dr. J.G. de Casparis and Mr. J.E.B. Gray of the School of Oriental and African Studies. By far my greatest debt of gratitude is owed to my mentor, Professor A.L. Basham. The present work is based on a thesis prepared under his supervision and for which he gave unstinting advice, assistance and encouragement. Financial support for my studies has been provided by my parents, my wife, the American Institute of Indian Studies-which contri- buted an invaluable year in India-and the Australian National University. Neither the American Institute nor the Australian National University is to be held responsible for the contents of this work.

CHAPTER I

FOUR ŚAIVITE SECTS

Brahma-sūtra Commentaries Several Sanskrit commentators on Brahma-sūtra ii. 2. 37 criticise the doctrines and practices of religious sects which preach devotion to Śiva and philosophical dualism. Śamkarācārya (c. 788-820) mentions only the Mahesvaras.1 It is clear from his discussion that they are the same as the Pasupatas. Vācaspati Miśra (c. 850) divides these Māheśvaras into four groups-Saivas, Pāśupatas, Kāpālikas, and Kāruņika-siddhāntins.2 Bhāskarācārya (c. 850) repeats this division but replaces the Kāruņika-siddhāntins with Kāțhaka-siddhāntins.3 Other commentators are said to call this last group Kāruka-siddhāntins.4 Yāmunācārya (c. 1050), the teacher of Rāmānuja, lists together Saivas, Pāśupatas, Kāpālas, and Kālāmukhas in his Āgama-prāmāņya.5 Rāmānuja (c. 1017- 1137) repeats his preceptor's comments, in large part verbatim, in his Śrī-bhāsya.6 Most later commentators also seem to follow Yamuna's classification. Although the Kāruka-, Kāruņika-, Kāțhaka-siddhāntins are only described very cursorily, they are apparently identical with the Kālāmukhas. The comments of Yāmunācārya and Rāmānuja contain valuable information, but have been accepted too uncritically by modern scholars. In some places Rāmānuja's Śrī-bhāșya is ambiguous and his remarks about the Kālāmukhas do not always harmonise with what is otherwise known about them.7 Also, he does not maintain consistently clear distinctions between the four sects. Sometimes he seems to describe them collectively, at other times

1 Brahma-sūtra-bhāșya, ed. B. Šāstrī, ii. 2. 37. 2 Bhāmatī, ed. B. Śāstrī, ii. 2. 37. 3 Brahma-sūtra-bhāșya, ed. V.P. Dvivedin, ii. 2. 37. 4R.G. Bhandarkar, Vaişnavism Śaivism and Minor Religious Systems, p. 121. This name is also found in M. Monier-Williams' Sanskrit-English Dictionary, but we have not found a commentary containing it. 5Quoted in K.K. Handiqui's notes to Śrīharșa's Naișadhacarita, p. 644. 6Ed. R.D. Karmarkar, ii. 2. 35-37. 7Since the full text of Yāmuna's Āgama-prāmānya was not available to us, we will follow Rāmānuja's Śrī-bhāsya. The translations of this are our own.

2 THE KĀPĀLIKAS AND KĀLĀMUKHAS

individually. He first identifies the four sects which follow the doctrine of Paśupati and then adds :

All these make an analysis of reality and a hypothesis about the attainment of bliss in this world and the next which are opposed to the Vedas. They make a distinction between the instrumental and material cause (nimittopadānayor bhedam) and designate Paśupati as the instrumental cause (but not the material cause of the Universe).8

In this respect the four sects appear to be the same. This is, no doubt, an oversimplification, but each may well have propounded a dualistic metaphysics. Rāmānuja next discusses the main features of Kāpāla (Kāpālika) worship :

As the Kāpālas declare : 'He who knows the essence of the six insignia (mudrikā-sațka), who is proficient in the highest mudra, and who meditates on the Self as seated in the vulva (bhagāsana-stha), attains nirvāņa.' They define the six insignia (mudrā) as the kaņthikā (necklace),9 the rucaka (another neck ornament), the kundala (earring), the śikhā- mani (crest-jewel), ashes, and the sacred thread. A person bearing these insignia is not born again in this world.10 Yāmunācārya makes the important addition that they have two secondary insignia (upamudrā)-the skull (kapāla) and the club (khatvānga).11 Most Saivite ascetics smear their bodies with ashes and wear sacred threads, but the skull and khatvānga are mostly peculiar to the Kāpälikas. The term kundala is used for the earrings of the Kāpalikas in a number of sources, and in Bhavabhūti's Mālatī-Mādhava a female Kāpālika bears the name Kapāla- kuņdalā. 12 Large earrings made of rhinoceros horn or other material are a distinguishing feature of a related group of tantric ascetics, the Kānphatā Yogins (kān = ear, phată = split). Their earrings are of two basic types-a flat one called darsana and a round one called kundala. Both are known as mudrās.13 Statues of Lakulīśa, the

8ii. 2. 35-37. 9Yāmunācārya reads karņikā in place of kaņțhikā. Quoted in Śrīharșa, p. 644. 10ii. 2. 35-37. 11Quoted in Śrīharșa, p. 644. 12See below. pp . 56-57. 13G.S. Ghurye, Indian Sadhus, pp. 135-136.

FOUR ŚAIVITE SECTS 3

Pāśupata-Kālāmukha saint, also commonly display large earrings. The other insignia in Rämanuja's list, the neck ornaments and crest-jewel, are nowhere else specially connected with the Kapālikas. The phrase 'proficient in the highest mudra' or 'most skilful in (the use of) mudrās' (para-mudrā-višārada) is difficult to interpret. R.G. Bhandarkar explains it as he 'who is skilful in their [the six insignia's] use,'14 but it is not easy to see how these insignia can be 'used'. In tantric literature the term mudrā is one of the five Ma-sounds which designate the principal ingredients of the central tantric ritual (pañca-makāra-sādhanā) : madya (liquor), māmsa (meat), matsya (fish), mudrā, and maithuna (coition). Here mudrā has a variety of meanings. In Hindu tantras it usually denotes parched grain, kidney beans, or any cereal believed to possess aphrodisiac qualities.15 In Buddhist tantric works, on the other hand, it usually refers to the female partner in the ritual. In Buddhist tantric yoga, the four stages in the pro- duction of bodhi-citta are also called mudrās. They are karma- mudrā, dharma-mudrā, mahā-mudrā, and samaya-mudrā.16 In non-tantric religious usage, and often in tantric works as well, mudrā denotes various ritual gestures, especially ones made with the hands. More generally it simply means 'mark' or 'insignia' as in the 'set of six insignia' (mudrā-satka) mentioned above. Bhandarkar's interpretation of Rämanuja's phrase is still the best one, but most of the other meanings of mudrā are also possible. The meditation on the 'Self as seated in the vulva'17 is reminiscent of the Buddhist tantric maxim: 'Buddha-hood resides in the woman's vulva.'18 The term bhaga (vulva) also has a variety of meanings, especially in the Buddhist Tantras. Many of these texts begin with the words : 'Once upon a time the Lord of all Tathāgatas ... was dwelling in the vulvae of the vajra-women.'19 This is an example of what Bharati calls afferent sandha-terminology-the use of object words, frequently erotic ones, to 'intend' metaphysical or mystical concepts.20 Here the commentators explain bhaga as

14p. 127. 15A. Bharati, The Tantric Tradition, p. 242. 16S.B. Dasgupta, An Introduction to Tāntric Buddhism, pp. 174-75. 17 Bhagāsana-stham ātmānam dhyātvā. 18Cited by L. de la Vallée Poussin, 'Täntrism (Buddhist),' ERE, XII, 196. 19'ekasmin samaye bhagavān sarva-tathāgata-kāya-vāk-citta-hrdaya-vajra-yoșit- bhageșu vijahāra.' Trans. Bharati, p. 170. 20Ibid., p. 173.

4 THE KĀPĀĻIKAS AND KĀLĀMUKHAS

the 'void-element' (kha-dhātu) or the 'void' (sūnyatā), and also as Prajñā, the female personification of enlightenment.21 The use of the term nirvāna instead of its Hindu equivalent, mokșa or mukti, is again suggestive of a Buddhist or Vajrayāņa context although nirvāna is also used in some Saivite tantric litera- ture such as the famous, though admittedly late, Mahānirvāna- tantra. Another Buddhist connection is found in the vernacular songs (caryās) of the Sahajiyā Buddhist saint Kāņhapāda. He elevates the Kāpālika to the rank of perfected yogin.22 Vajrayāņa literature also refers to ritual paraphernalia typical of Kāpālika worship-such as bones, blood, flesh, and skulls-more often than Hindu Tantras do. Nonetheless, all Sanskrit sources claim that the Kāpālikas worship the Hindu deity Bhairava-Siva and his consort. There is little doubt, therefore, that the Kāpālikas were a Saivite sect. The Buddhist parallels indicate that they must have also had some connection with Buddhist tantrism, but, in the absence of additional evidence, it is useless to speculate about what this may have been. Yāmunācārya and Rāmānuja continue their discussion with some comments about the last of the four sects, the Kālāmukhas. Both authors should have been acquainted with these ascetics since the sect was influential in South India, particularly in the Mysore region, between the eleventh and thirteenth centuries. Yāmunācārya is believed to have spent most of his life in Kāñci. His disciple divided most of his time between this city and Śrirangam, about 150 miles to the south. He also made pilgrimages to other parts of India. Contemporary Kālāmukha monasteries in the Madras region existed at Tiruvānakkoyil in Chingleput District, Vēdal in North Arcot District, Koyil Tēvarāyanpettai in Tanjore District, and Kodumbāļūr in Tiruchchirappali District-none very far from Kāñcī or Śrīrangam.23 Rāmānuja's portrait of the sect, however, is quite different from the one obtained ffom Kālāmukha epigraphs or from the works of their parent sect, the Pāśupatas. Śrī-bhāșya ii. 2. 35-37 states : Likewise, the Kālāmukhas designate (the following) as the means of securing all desires in this world and the next : eating from a skullbowl, besmearing the body (snāna) with

21 Dasgupta, pp. 105, 120-21. 22S.B. Dasgupta, Obscure Religious Cults, pp. 57-58, 90, 103-104. 23See below, pp: 165-67.

FOUR ŚAIVITE SECTS 5

the ashes of a corpse, eating those (ashes), bearing a staff (laguda), keeping a pot of wine (surā), and using that pot for worship of the gods (deva-pūjā), etc. Much of this description seems more appropriate to the Kāpālikas. Only two items are associated with the Kalāmukhas in other sources-the bath in ashes and the staff (laguda). The bath in ashes is one of the central rituals prescribed in the Pāsupata-sūtra.24 The supposed author of this work, Lakulīśa, is held in equally high esteem by both the Pāsupatas and Kālāmukhas. His name indicates that he also carried a staff (lakula) .25 The words lakula and laguda are synonymous and etymologically identical. The Kāpālikas, on the other hand, normally carry a khatvānga or a trident (triśūla). Elsewhere in the Agama-prāmānya, Yāmunācārya speaks of a fourfold division of the tantras: Saiva, Pāsupata, Saumya, and Liguda.20 This clearly corresponds to the four sects: Saiva, Plšupata, Kāpāla, and Kālāmukha. Eating from a skull bowl and worshipping the gods with a pot of wine are items especially associated with the Kāpālikas, not the Kālāmukhas. Sanskrit sources usually portray Kāpālikas as charlatan ascetics who wander about with a skull begging bowl and drink liquo: freely for mundane as well as ritual purposes. They also wear the ashes of the dead although no source claims that they eat them. The seeming confusion in Rāmānuja's account between the Kāpālikas and Kālāmukhas was noticed by R.G. Bhandarkar who concluded (p. 128) that 'people do not seem to have made a sharp distinction' between them. G.S. Ghurye has suggested (p. 128) that by the twelfth century, the time of the greatest number of KAlamukha epigraphs, the sect 'had purged itself of, or had at least suppressed, the more objectionable practices.' Bhandarkar's theory is the more plausible one, but neither is very satisfactory. There were in fact considerable differences between the two sects, and Yāmuna and Rāmānuja must have known how to distinguish them. Ghurye's theory fails to account for the fact that the earliest Kālāmukha record, an inscription of A.D. 810, shows no more evidence of religious extremism than any of their later records. One might suggest a more sinister explanation. At the time of

24Ed. R.A. Sastri, i. 2. 25See below, p. 108. 20Quoted in Śrīharșa, p. 643.

6 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Yāmuna and Rāmānuja the Kālāmukhas were rapidly gaining popular and even royal support in South India. The two Vaisnava priests may have purposely confused the two Saivite sects in order to discredit their more important rivals. Śrī-bhāșya ii. 2. 35-37 next gives a list of religious paraphernalia prescribed in the Śaivāgamas: 'the rosary of rudrākșa seeds in the hand, a single mass of matted hair on the head, the skull-bowl, the besmearing the body with ashes, etc.' Presumably Rāmānuja means to associate these items with the Saiva sect, but with the exception of the skull-bowl, which does not belong with the Saivas anyway, all the items are part of the costume of most types of Śaivite ascetics. This list may be compared with the only significant epigraphic description of a Kālāmukha ascetic. A grant of A.D. 1252-53 from Munavaļļi in Belgaum District praises the Kālāmukha rāyarājaguru Sarveśvaradeva : 'whose body was sprinkled with ashes; who wore a small piece of cloth around the loins, and the hairy skin of an antelope; who carried a rosary of Rudrākșas. ... '27 The loincloth and antelope skin as well as the ashes and rudrāksas are standard equipment for most Saivite ascetics. No mention is made of wine pots or skull bowls. Rāmānuja's Šrī-bhāșya ii. 2. 35-37 concludes with a disparaging description of the Kāpāla vow (vrata) : Likewise, they ( ?= the Saivāgamas, the four sects, the Kāpālas) state that even men belonging to lower castes can attain the status of Brähmaņa and the highest āśrama (= samnyāsa, mendicancy) by means of certain special rites. (For it is said) : 'One instantly becomes a Brāhmaņa merely by the process of initiation. A man becomes a great ascetic (yati) by undertaking the Kāpāla vow.' This may be compared with the following verse from the Kulārņava- tantra : 'Gone is the Sudra-hood of the Sūdra and the Brahmaņa- hood of the Brāhmana (vipra); there is no division into castes for one who is consecrated by initiation.'28 Hostility to caste conscious- ness is a normal feature of tantric worship and is consistently espoused by Kāpālikas in Sanskrit literature. From a modern point

27J.F. Fleet (ed. and trans.), 'Sanskrit and Old Canarese Inscriptions Relating to Yādava Kings of Dēvagiri,' JBBRAS, XII (1876), 40. 28Ed. T. Vidyāratna, xiv. 91.

FOUR ŚAIVITE SECTS 7

of view this hostility may be commended, but for orthodox Hindu writers such as Ramanuja an attack on caste was an attack on the whole divinely ordained social order (varnāśrama-dharma). There are also limitations to this tantric rejection of caste which Rāmānuja does not choose to note. In most tantric works the denial of caste occurs only in ritual situations. In day-to-day affairs, caste distinctions are still maintained. Thus Kulārņava- tantra viii.101 says : 'In this cakra (circle of worship) there is no division into castes. Everyone (in it) is declared to be equal with Siva.' Elsewhere, however, this text prescribes different lengths of studentship for members of different classes. The Mahānirvāņa- tantra seems to accept class divisions without qualification.29 The transcending of caste barriers in a ritual context has little or nothing to do with rational materialist arguments. It is part of a mystical reversal and revaluation of all values, eine Umwertung aller Werte, valid only in the sacred circle of worship. In the supra- mundane universe of the ritual, opposites coalesce and change places-the lowest is highest and the highest lowest.30 In relation to caste, this mystical principle culminates with the apotheosis of the dombi (washerwoman) in Kānhapāda's tantric Buddhist songs.31

Purāņas and Other Sources Several Puranas and a few other works contain lists incorporating some or all of the sects in the fourfold classification of the Brahma- sūtra) commentators. The following table compares the commen- taries with these other sources :

Bhāskarācārya32 Kāpālika Pāśupata Kāțhaka- Śaiva siddhāntin

Vācaspati Kāpālika Pāśupata Miśra33 Kāruņika- Śaiva siddhäntin

29Ed. and trans. J. Woodroffe, chap. viii. 30See Bharati, p. 234, and M. Eliade, Yoga : Immortality and Freedom, p. 261. 31See Dasgupta, Obscure ... , pp. 57, 99, 102-106. 32 Brahma-sūtra-bhāșya ii. 2. 37. 33 Bhâmatī ii. 2. 37.

8 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Yāmunācārya34 Kāpālika Pāśupata & Rāmānuja35 Kālāmukha Śaiva

Yāmunācārya36 Saumya Pāśupata Lāguda Śaiva

Kūrma Purāņa Kāpāla37 Pāśupata Also Vāma, Nākula or Bhairava & Lākula Pañcarātra

Soma38 Pāśupata Also Vāma Lākura or & Bhairava Lāñjana or Vākula

Soma39 Pāśupata Also Vāma Lāṅgala & Bhairava (Lāguda)

Nārādīya P.40 Kāpāla Pāśupata Mahāvrata- Siddhānta- dhara mārga

Śiva P. Vāya- Kāpāla Pāśupata viya-samhitā41 Mahāvrata- Siddhānta- dhara mārga

Skanda P.42 Kańkāla Pāśupta Also Kālamukha Śaiva Mahāvrata

34 Āgama-prāmāņya, quoted in Śrīharșa, p. 643. 35 Śrī-bhāșya ii. 35-37. 36Quoted in Śrīharșa, p. 643. 37Ed. N. Mukhopadhyāya, i. 16 (p. 184). 38ii. 12 (p. 740). Lākura is probably a mistake for Lākula. 39 Uparibhāga. 37. 147, cited by Handiqui, p. 463. We do not know from which edition of the Purana this reference comes. Commenting on this verse, Appaya Dīkșita (cited ibid.) reads Lāguda (holding a staff) for Lāngala (a plough). Lāguda is a better reading. 40 Uttarakhanda. 31. 103, cited by A.P. Karmarkar, The Vrātya or Dravidian Systems, p. 220. 41Ed. Mallikārjunaśāstrī, ii. 24. 177. 42 Aruņācala-Mā. 10. 65, cited by Karmarkar, p. 220.

FOUR ŚAIVITE SECTS 9

Skanda P. Kāpāla Pāśupata Also Soma Sūta-samhitā43 Lākula

Svayambhu P.44 Soma Pāśupata Vā(Lā)kula Śaiva

Vāmana P. Kāpālika45 Pāśupata Kāladamana Śaiva

Mahāvratin46 Pāśupata Also Kālāmukha Śaiva Mahāpāśupata & several others

Vaśistha & Soma Also Lokāyata Lińga Ps.47 Nākula & Bhairava

Ānandagiri48 Pāśupata Also Ugra, Śaiva Raudra, Bhatța, & Jaņgama

Rājaśekhara49 Mahāvrata- Pāśupata dhara Kālamukha Śaiva

Śaktisańgama- Pāśupata Also 6 tantra50 Kālāmukha Śaiva others

43Ed. V.S. Paņaśīkara, Yajñavaibhavakhaņda. 22. 3. 44Quoted in Isāna-Śivaguru-paddhati, Pt. III, Kriyāpada, chap. i, cited by V.S. Pathak, History of Saiva Cults in Northern India from Inscriptions, p. 3. 45Venkateśvara Press edition, vi. 87. 461xvii. 10-20. 47In a verse attributed to these two Purāņas by the Tantrādhikārinirņaya, cited by C. Chakravarti, Tantras : Studies on their Religion and Literature, p. 51. 48 Śamkara-vijaya, cited by Pathak, p. 4. 49 Şaddarśana-samuccaya, cited by Pathak, p. 21. On p. 2. Pathak incorrectly attributes this verse to Haribhadra's Saddarśana-samuccaya. 50Ed. B. Bhattacharya, i. 5. 92-93.

10 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Siddha- Kāpālika Pāśupata Also siddhānta- Kālāmukha Śaiva Mahāvrata- paddhati51 dhara & 5 others

Suprabhedāgama52 Sauma Pāśupata Lākula Śaiva

Malkāpuram Śivaśāsana Pāśupata Stone Inscri.53 Kālānana Śaiva

It is evident from this table that the sects had several alternate names. The most important variants are: Lākula, Nākula and Lāguda for Kālāmukha; Soma and Saumya for Kāpāla; and Mahāvrata-dhara for both Kāpāla and Kālāmukha. The term Śivaśāsana from the Malkāpuram inscription does not necessarily refer to the Kāpālikas since this identification is based merely on an analogy with the standard fourfold division.54 The term Kankāla (skeleton) from the Skanda Purāna almost certainly refers to the Kāpālika sect, but the fifth item of this list, Mahävrata, often denotes this sect also. Kālānana (blackfaced) from the Malkāpuram inscription is merely a synonym for Kālāmukha. A tenth century grant to a Kālāmukha priest at Kodumbāļūr (Tiruchchirappali District) similarly mentions fifty Asitavaktra (black-faced) ascetics residing at his monastery.55 Although the literal meaning of Kāladamana (time-subduing) from the Vāmana Purāņa is consi- derably different, it is evidently another variant of Kālāmukha and Kālānana. The originator of the Kāladamana doctrine was named Kālāsya (black-faced).56 Most of these sources merely enumerate the sects or say that their doctrines were revealed by Siva. A few works openly condemn

51Cited by Pathak, p. 26. 52Cited by Pathak, p. 3. 53 Journal of the Andhra Historical Research Society, IV, 147, cited by Pathak, p. 3. 54The verse reads : 'upeyuşām Šaiva-tapodhanānām Kālānanānām Śivaśāsanā- nām/ vidyārthinām Pāsupata-vratānām apy anna-vastrādi-samarpaņāya//.' Cited by Pathak, p. 3. Some of the four terms in this verse may be adjectives rather than nouns. 55K.A.N. Sastri, 'The Kodumbāļūr Inscription of Vikrama-Kēsarī,' JORM, VII (1933), 9. 56Vāmana Purāņa vi. 90.

FOUR ŚAIVITE SECTS 11

the sects. In the Kūrma Purāna Siva says : 'I have declared other śāstras which are a source of confusion in this world and are opposed to the words of the Vedas. The Vāma, Pasupata, Soma, Lāngala, and Bhairava (śāstras) are declared to be outside the Vedas and are not to be served.'57 Yāmunācārya is equally critical : 'Śaiva, Pāśupata, Saumya, and Lāguda are designated as the fourfold division of the Tantras. One should not make a mixture (of these with Vedic rites).'58 The Sktında Purāna, however, at one point declares that only five of the twenty-eight agamas lead to the path of liberation : the Kālamukha, Kankāla, Saiva, Pāśupata, and Mahāvrata.59 The remarks of the Vāmana Purāna vi. 86-92 are the most interesting. It states that Brahma created four groups which worshipped Hara (Siva) and gave them each a sāstra: 'The first is known as Saiva, then Pāsupata ... , then the third Kāladamana, and the fourth Kapalika.' The text then gives a pseudo-historical account of the origin of each : Śiva himself was Śakti, the beloved son of Vasișțha. Gopāyana then became his pupil Mahāpāśupata was the ascetic Bharadvāja. His pupil was the king Somakeśvara Lord Kālāsya was the ascetic Apastamba. His pupil was named Krātheśvara Mahavratin was Dhanada. His pupil was the powerful Arnodara, a great ascetic and a Sūdra by birth. The apparent associations of teachers and doctrines are Sakti and Śaiva, Bharadvāja and Pāśupata, Āpastamba and Kāladamana (Kālāmukha), and Dhanada and Kāpālika. The reasons behind this choice of religious founders are obscure. Sakti, Bharadvāja and Apastamba are famous sages and Dhanada is the god of wealth. None are elsewhere connected with these sects with the possible exception of Apastamba. An incomplete record from Vēdal in North Arcot District, Madras, mentions a Kāļamukha Daśapuriyan of the Hārita gotra and the Āpastamba sūtra.60 The attribution of the Kāpālika śāstra to Dhanada and his powerful

57 Uparibhāga. 37. 146-47, cited by Handiqui, p. 463 (my translation). 58Cited ibid. 59 Aruņācala-Mā. 10. 65, cited by Karmarkar, p. 220. 60See V. Rangacharya, Inscriptions of the Madras Presidency, II, 1162.

12 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Śūdra disciple, Arņodara, emphasizes the worldliness and debased status of this doctrine. The four disciples-Arņodara, Gopāyana, Somakeśvara, and Krātheśvara-cannot be identified. Śiva, Mahā- pāśupata, Kālāsya, and Mahāvratin are evidently forms of Śiva. Vāmaņa Purāņa Ixvii. 1-40 tells of a war between Śiva, aided by his ganas and pramathas, and the asuras. Siva's allies included the Saivas, Pāśupatas, Kālāmukhas, Mahāvratins, Nirāśrayas, and Mahāpāsupatas. Siva extended a special welcome to the last group because they did not recognize a distinction between him and Vişņu. Here the Pāśupatas and Mahāpāśupatas seem to be separate groups.

CHAPTER II

KĀPĀLIKA SOURCES

Early Sources The earliest occurrence of the word kapālin (one who bears a skull) is probably that in the Yājnavalkya-smrti iii. 243 (c. A.D. 100- 300). This sūtra prescribes the penance for one who has killed a Brāhmaņa, a Brahmahan : 'With a skull (śiraḥ-kapālī) and a staff (in his hands), living on alms, announcing his deed (as he begs), and eating little food, the killer of a Brāhmana may be purified after twelve years.'1 Other law books prescribe much the same penance but do not use the term kapālin. An important connection between this penance and the Kāpālika faith does exist,2 but in this passage kapālin has the sense only of 'bearing a skull' and does not imply the existence of a sect or order of Kapālins. In the Maitrāyanīya Upanișad certain Kāpālins who hypocritically wear red robes (kaşāya) and earrings (kundala) are mentioned among persons with whom it is improper to associate.3 This seems to denote a member of the Kāpālika sect, but the relevant passage is definitely an interpolation or appendix to the original text and may be of fairly late date.4 The Prakrit Gāthā-saptaśatī is traditionally ascribed to the first century A.D. Sātavāhana king Hāla but was probably compiled sometime in the third to fifth centuries. It contains a verse describing a 'new' female Kāpālikā who incessantly besmears herself with ashes from the funeral pyre of her lover.5 The word 'new' (nava), unless it means simply 'young', suggests that her Kāpālika vow was taken at his death. This may well be the earliest reference to the Kāpālika sect. A Buddhist text of the early centuries of the Christian era, the

1Ed. N.R. Acharya. 2See below, pp. 73-82. 3Ed. and trans. J.A.B. van Buitenen, vii. 8. 4Ibid. pp. 88-89. SThe Sanskrit chāyā reads : 'jāra-śmaśāna-samudbhava-bhūti-sukha-sparśa-sveda- śīlāngyāḥ/ na samāpyate nava-Kāpālikyā uddhūlanārambhaḥ//' Kāvyamālā edition, vs. 408.

14 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Lalitavistara, mentions certain 'fools' who seek purification by smearing their bodies with ashes, wearing red garments (kasāya), shaving their heads, and carrying a triple-staff (tridanda), a pot, a skull, and a khatvānga.6 These must also be Kāpālikas. By the sixth to seventh centuries references to Kāpālika ascetics become fairly commonplace. The astronomer-mathematician Varā- hamihira (c. 500-575) refers to the Kāpāla vow in his Brhatsamhitā ix. 25: 'When the chariot of Rohiņi (an asterism) is intercepted (by Venus), the earth (becomes) decorated with hair and pieces of bone and seems to keep the Käpäla vow, as if it had committed sin.'7 This might refer merely to the Brahmanhan penance, but verse lxxxvii. 22 of this work seems clearly to mention Kāpālika ascetics : 'When (a tranquil omen) is in the southwest (spoke of a 'Cycle of Quarters') the arrival of a cow, a sportsman (krīdaka) or a Kāpālika is indicated, and one will obtain a bull. (There will also be) black gram, horse-gram, etc. and food.' In his Brhajjātaka xv. 1, Varāhamihira enumerates seven classes of ascetics, each born under the influence of a different heavenly body.8 He lists them as follows : the Sākyas under Mars, Ājīvikas under Mercury, Bhiksus under Jupiter, Vrddhas under the moon, Carakas under Venus, Nirgranthas under Saturn, and Vanyāśanas under the sun. The tenth century commentator Utpala (or Bhat- totpala) says that the Vrddhas are also known as Vrddha-śrāvakas or Kāpālikas.9 Utpala also mentions a similar classification made by the fifth century Jain authority, Kālakācārya. This connects the sun with Tapasvins, the moon with Kāpalins, Mars with Raktapatas, Mercury with Ekadandins, Jupiter with Yatis, Venus with Carakas, and Saturn with Ksapanakas.10 Again commenting on Varāhamirhira's text, Utpala says : 'Here the word Vrddha- śrāvaka implies the wandering ascetics who seek refuge with Maheśvara, and the word Äjīvika those who seek refuge with

6Ed. P.L. Vaidya, chap. xvii (p. 183). 7Ed. H. Kern. A slightly different version of this verse is quoted in Pañcatantra i. 234 (ed. N.R. Acharya). This version reads 'Kāpālika vow' in place of 'Kāpāla vow' and 'ashes and pieces of bone' in place of 'hair and pieces of bone.' 8Ed. and trans. V.S. Sastri. See A.L. Basham, History and Doctrines of the Äjivikas, pp. 168-71. 9'kşapākaraś candro yadā Vrddhaḥ Vddha-śrāvakaḥ Kāpālikah vrtta-bhanga- bhayāc chrāvaka-śabdo 'tra lupto draştavyaḥ.' Commentary on Brhajjātaka xv. 1 (1863 Bombay edition). 10Ibid.

KĀPĀLIKA SOURCES 15

Nārāyaņa.'11 The attribution of Nārāyana worship to the Ājīvikas is a mistake. It is apparently based on an attempt to equate them with Kalakācārya's Ekadandins.12 If the identi- fication of Vrddhas or Vrddha-śrāvakas with the Kāpālikas is correct, the claim that they worshipped Maheśvara is also correct. The Vrddha-śrāvakas are again mentioned by Varāhamihira in his Brhatsamhitā li. 20: 'When a fortune-teller is consulted by persons in the sight of a [Vrddha-śrāvaka], they do so for the sake of friends or gambling; when in the sight of a friar of decent order [suparivrāj], their query concerns a courtesan, king or wife in childbed.'13 H. Kern, the translator of this verse, renders Vrddha-śrāvaka as 'skull-wearing Śaiva monk,' presum- ably on the basis of a commentary. We have found no other examples of Kāpālikas being called by this name. The famous Chinese pilgrim Hsüan Tsang gives brief accounts of the relative strengths of the various types of Buddhists and other sects in the places he visited during his South Asian travels (c. A.D. 630-644). In Kāpiśā, modern Nuristan in eastern Afghanis- tan,14 he found over a hundred Buddhist monasteries. In addition, he says, 'there are some ten temples of the Devas, and 1000 or so of heretics (different ways of religion); there are naked ascetics, and others who cover themselves with ashes, and some who make chaplets of bones, which they wear as crowns on their heads.'15 Beal identifies these heretics as Digambara Jains, Päśupatas, and Kapāla-dhārins, i.e. Kāpālikas.16 Elsewhere Hsüan Tsang gives a general description of various non-Buddhist ascetics he met in India proper : The dress and ornaments worn by non-believers are varied and mixed. Some wear peacocks' feathers; some wear as ornaments necklaces made of skull bones ...; some have no clothing, but go naked ...; some wear leaf or bark garments; some pull out their hair and cut off their mous- taches; others have bushy whiskers and their hair braided

llIbid. The text mistakenly reads 'Māheśvara' for 'Maheśvara.' 12See Basham, pp. 170-74. 13Trans. H. Kern, JRAS, n.s. VI (1873), 87. Kern believes that this chapter may be spurious. 14See T. Watters, On Yuan Chwang's Travels in India, I, 123-24. 15S. Beal (trans.), Chinese Accounts of India, I, 117-18. 16Ibid., p. 118.

16 THE KĀPĀLIKAS AND KĀLĀMUKHAS

on the top of their heads. The costume is not uniform, and the colour, whether red or white, not constant.17

Those ascetics who wear peacocks' feathers, go about naked, and pull out their hair are probably Jains. Those who wear skull gar- lands, as Beal suggests, may well be Kāpälikas. The others are not easily identified. Hsuan Tsang visited India during the reign of Harsa-vardhana of Sthāņvīśvara (A.D. 606-647). This king's contemporary biographer Banabhatta vividly portrays the religious life of court and kingdom in his Harsa-carita and Kādambarī. These works show that both primitive and developed types of tantric worship were already widespread in the seventh century. The most archaic level of tantric worship is represented in Kādambarī by the wild Sabara tribe of the Vindhya forest whose 'one religion is offering human flesh' to Candik18 and whose chief had shoulders that 'were rough with scars from keen weapons often used to make an offering of blood' to Candikā.19 The incorporation of 'Hinduised' (Eliade) or 'Sans- kritised' (Srinivas) forms into the rituals of tribesmen such as these probably amounted to little more than the identification of their tutelary gods and goddesses with Hindu ones such as Bhairava, Kālī and Caņdikā. Elsewhere in Kādambarī Bāņa describes various religious and philanthropic acts performed by Queen Vilāsavatī of Ujjayinī in order to acquire a son :

She slept within the temples of [Candikā], dark with the smoke of bdellium [guggulu] ceaselessly burnt, on a bed of clubs covered with green grass ...; she stood in the midst of a circle drawn by [great magicians20], in a place where four roads meet, on the fourteenth night of the dark fort- night ...; she honoured the shrines of the siddhas and sought the houses of neighbouring Mātrkās ... ; she carried about little caskets of mantras filled with birch-leaves written over in yellow letters; ... she daily threw out lumps of flesh in the evening for the jackals; she told pandits the

17Trans. ibid., II, 134. 18Trans. C.M. Ridding, p. 31. Ed. P.V. Kane, Vol. I, text p. 21. 19Trans. Ridding p. 28. Ed. Kane, Vol. I, text p. 20. 20mahānarendra. Ridding's translation, 'the king himself,' is unlikely.

KĀPĀLIKA SOURCES. 17

wonders of her dreams, and at the cross-roads she offered oblation to Śiva.21

Although these rituals display a greater degree of Sanskritisation than those of the Sabara tribesmen, many of her endeavours blend tantric worship with motifs of archaic fertility magic. Crossroads, for instance, are a focal point for fertility rituals and other religious ceremonies in many parts of the world.22 A much more sanguinary amalgam of archaic magic and tantric ritual is described in the Harsa-carita. When Harsa's father falls ill, the populace of the capital city undertake various penances in order to avert his death :

Young nobles were burning themselves with lamps to propitiate the Mothers [Mātrkās]. In one place a Dravidian was ready to solicit the Vampire [Vetäla] with the offering of a skull. In another an Andhra man was holding up his arms like a rampart to conciliate Candikā. Elsewhere young servants were pacifying Mahākāla by holding melting gum [guggulu] on their heads. In another place a group of relatives was intent on an oblation of their own flesh, which they severed with keen knives. Elsewhere, again, young courtiers were openly resorting to the sale of human flesh.23 The sale of human flesh to cremation ground demons is mentioned in Bhavabhūti's Mālati-Mādhava, in the Kathāsaritsāgara, and in other Sanskrit works. Somadeva's Yaśastilaka (A.D. 959) men- tions Mahävratin heroes who sell human flesh cut from their own bodies.24 The term Mahāvratin is normally used to denote Kāpā- likas.25 The Dravidian in the above passage who offers a skull to a Vetāla must also represent a Kāpālika or closely related type of ascetic. A tantric ascetic from South India is described in great detail in Kādambarī. This Dravida-dhārmika superintends a temple of Candikā located on the road to Ujjayinī. In one spot the temple

21Trans. Ridding, pp. 55-56. Ed. Kane, Vol. I, text pp. 42-43. 22For a discussion of worship at the crossroads, see D.D. Kosambi, Myth and Reality, chap. iii. In ancient Greece Hermes was the leading god of the crossroads and also a god of fertility. See N.O. Brown, Hermes the Thief. 23Trans. E.B. Cowell and F.W. Thomas, pp. 135-36. Ed. P.V. Kane, Part II, text p. 21. 24See K.K. Handiqui, pp. 358-59. 25See below, pp. 73-82.

18 THE KĀPĀLIKAS AND KĀLĀMUKHAS

'displays the slaying of (animal) sacrifices ... with heaps of skulls (that are) like fruits'.26 The dhārmika is crippled and maimed as a result of foolish penances and fights with travellers and wild animals. The tantric character of his worship is emphasised in some of the following epithets : He had a tumor growing on his forehead that was blackened by (constantly) falling at the feet of Ambikā (the idol of Caņdikā) ... He had brought on himself premature fever with improperly prepared mercurial medicines. Although old, he troubled Durgä with requests for the boon of soverei- gnty over the Deccan ... He had made a collection of manuscripts of jugglery, Tantras and mantras (which were written) in letters of red lac on palm leaves (tinged with) smoke. He had written down the doctrine of Mahākāla, which is the ancient teaching of the Mahapāsupatas ... He manifested the disease of talking (continually about the nine) treasures (of Kubera) and became very windy (on the subject) of alchemy ... He had increased his grasp on the mantra-sādhana for becoming invisible and knew thousands of wonderful stories about Śriparvata He had many times employed woman-subduing powders ...

on old female ascetics from foreign countries who stayed (at the temple) ... 27 This remarkable passage contains one of the earliest references to Tantra manuscripts as well as to alchemy (dhātuvāda) and mercurial medicines for prolonging life (rasāyana). Mantra- sādhana (performance of mantras) is a typical tantric term. These facts show that tantric worship was fully developed by Bāņa's time and was apparently centered mainly in South India. From our point of view the references to the teachings of the Mahāpāśupatas and to the mountain Srīparvata are of special interest since they both tend to connect this devotee with the Kāpalikas and Kalamukhas. We have noted that the Vamana Purāņa seems to mention Mahāpāśupata as the form of Śiva who incarnated himself as Bharadvāja for the propagation of the Pāśupata doctrine, but that elsewhere it mentions Śaivas, Pāśu-

26Ed. Kane, Vol. II, text p. 67. 27Ed. Kane, Vol. II, text pp. 68-69. My translatien is loosely based on the renderings in Kane's English notes.

KĀPĀLIKA SOURCES 19

patas, Kālamukhas, Mahāvratins, Nirāśrayas, and Mahāpāśupatas, as separate groups.28 Several other sources lend weight to the suggestion that the Mahāpāsupatas were at least partly distinct from the ordinary Pāśupatas. Handiqui points out (p. 241) that the Mahāpāśupatas are mentioned by Udayana (late tenth century) and that Varadarāja (eleventh century) and Śamkara Miśra (c. 1600) both identify them as those Pasupatas who practised the Mahāvrata. A South Indian drama approximately contemporary with Kādam- barī, Mahendravarman's Mattavilāsa, seems to address a Kāpālika as Mahāpāśupata.29 A verse found in two Kannada inscriptions from Belgaum District dated A.D. 1148 and 1219-20 seems to identify Kālāmukhas as both Mahāpāśupatas and Mahāvratins.30 A few other inscriptions also called Kālāmukha priests Mahāvra- tins. This is a source of some confusion since Kāpālikas are usually given this title, but it is likely that the Kālāmukha and Kāpālika Mahāvratas were quite different vows.31 Since the Kālāmukhas were closely related to the ordinary Päśupatas, we feel that it is in general best to connect the Mahāpāśupatas with the Kālāmukhas and not with the Kāpālikas or Pasupatas. There is no evidence, however, that either the Kālāmukha or Pāsupata faiths were markedly tantric in character. For this reason it is quite possible that the Mahāpasupata teachings written down by Bāņa's Dravida- dhārmika were Kāpālika and not Kālāmukha doctrines. This would also agree with the Mattavilasa reference. The conflicting claims of the Kālāmukhas, Kāpālikas and Pāśupatas to the title Mahāpaśupata cannot be completely resolved without further evidence. Somewhat the same problem is encountered in the reference to, the Dravida-dhārmika's wonderful stories about Śriparvata. This famous pilgrimage site in Kurnool District, Andhra Pradesh, is the home of the Kāpālikas in Bhavabhūti's Mālatī-Mādhava

28See above, pp. 11-12. 29Ed. T.G. Sāstri, p. 26. Trans. L.D. Barnett. BSOS, V (1930), 715. Barnett did not recognize the significance of the term and translated it as 'noble Pāsupata.' In so doing he assumed that it referred to the Pasupata who appears in the play, but the context makes this unlikely. 30R.S. Panchamukhi (ed.), Karnatak Inscriptions, I, 34 and J.F. Fleet, 'A Series of Sanskrit and Old Canarese Inscriptions Relating to the Ratta Chieftains of Saun- datti and Belgaum,' JBBRAS, X (1871-74), 247. 31See below. pp. 73-82.

20 THE KĀPĀLIKAS AND KĀLĀMUKHAS

but is mentioned in eleventh century inscriptions as a Kālāmukha shrine.32 Bāņa gives a more sympathetic portrait of a Saivite ascetic in his Harșa-carita. Bhairavācārya, the saint who befriended Harșa's ancestor Puspabhūti, was also from South India (dākșiņatya) and also performed a tantric ritual appropriate for a Kāpālika. One of his three disciples, Karņatāla, was a Drāvida and another, Tīțibha, carried a skull begging bowl (bhikșā-kāpālika) in a box made of kharjūra wood.33 Bhairavācārya's name indicates that he worshipped Siva as Bhairava, the form of the god held in especial esteem by tantric groups such as the Kāpālikas. Bāņa introduces him as the 'great Saiva saint named Bhairavācārya, almost a second overthrower of Daksa's sacrifice, who belonged to the Deckan [sic], but whose powers, made famous by his excellence in multifarious sciences, were, like his many thousands of disciples, spread abroad over the whole sphere of humanity.'34 The word here translated as 'great Saiva saint' (mahāśaiva) does not seem to denote a specific sect or ascetic order. It is simply a descriptive term showing his strong devotion to Siva. The original overthrower of Daksa's sacrifice was Siva himself. In at least one source, the Vāmana-Purāņa, Dakșa is said to have refused to invite Śiva to his sacrifice because the god had become a Kapālin after cutting off the fifth head of Brahmā.35 Siva-Kapālin or Kapāleśvara is the divine archetype of the Kāpālika ascetic.36 When King Puspabhūti, a devout worshipper of Siva (parama- māheśvara), learned of this great saint Bhairavācārya, he expressed a desire to pay him homage. A meeting was arranged and Puspabhūti went to see him in a Bilva tree plantation near an old temple of the Mothers (Matrs). The description of the saint which follows is too long to quote in full, but a few of its more interesting features should be noted.37 Puspabhūti saw Bhairavācārya 'seated on a

32See below, pp. 50-51. 33Ed. Kane, Part I, text p. 46. 34Trans. Cowell and Thomas, p. 85. Ed. Kane, Part I, text p. 45. 35ii. 17 to iv. 1. In Somadeva's Kathāsaritsāgara i. 1. 23 ff. (ed. Durgaprasād and K.P. Parab), Siva is not invited because he wears a necklace of skulls. Similar ex- planations are given in the Padma and Bhāgavata Purānas. See Vișnu Purāņa, H.H. Wilson (trans.), pp. 55-56. 36See below. pp. 77-81. 37The translations are from Cowell and Thomas, pp. 263-65. The text is edited by Kane, Part I, text pp. 46-47.

KĀPĀLIKA SOURCES 21

tiger-skin, which was stretched on ground smeared with green cow-dung, and whose outline was marked by a boundary ridge of ashes.' The flashing luster of his body was like red arsenic paste 'purchased by the sale of human flesh.' His hair was twisted together (jati-krta) in ascetic fashion and was festooned with rosary beads (rudrāksa) and shells. He had a 'slanting forehead-mark, made with ashes.' His lip hung down a bit 'as if overweighted by the whole Saivite canon [Śaiva-samhitā] resting on the tip of his tongue.' He wore a pair of crystal earrings (sphāțika-kundala) and 'upon one forearm, having an iron bracelet and bound with the line of charm-thread [mantra-sūtra] of various herbs, ... a bit of shell like one of Püsan's teeth broken by holy Siva.'38 He revolved a rosary in hin right hand like a water wheel. He had a thick beard nad wore a loincloth (kaupina) and ascetic's shawl (yoga-pattaka). Constant at his side was a bamboo staff39 with a barb of iron Inserted in the end' which was like the goad for driving away Oancha. He had observed the vow of celibacy since childhood.40 'Supreme in austerities' and 'surpassing in wisdom,' he was 'like Kailāsa, having his head purified by the dust of Paupati's fcet ; like Siva's heaven, the resort of Mahesvara throngs.' One day Bhairaväcārya asked the king to assist him in the completion of the powerful spell (mahāmantra) called Mahākāla- hrdaya. He had previously begun its performance in the great cemetery 'by a crore of muttered prayers ... in garlands, clothes, and unguents all of black as enjoined in the Kalpa.' The object wax to subdue a Vetäla. The king agreed to help and duly arrived at 'the empty house near the great cemetery ... on the approaching fourteenth night of the dark fortnight' :41 In the centre of a great circle of ashes white as lotus pollen Bhairavācārya could be seen ... Seated on the breast of a corpse which lay supine anointed with red sandal and arrayed in garlands, clothes and ornaments all of red, himself with a black turban, black unguents, black amulet [pratisara], and black garments, he had begun a fire rite [agnikārya] in the corpse's mouth, where a flame was burning.

wPuşan lost his tooth during the destruction of Daksa's sacrifice. Wvaiņavena višākhikādaņdena. The meaning of višākḥikā is unclear. 40k umāra-brahmacāriņam. Cowell and Thomas incorrectly translate this as 'chaste as a boy.' 41Trans. Cowell and Thomas, pp. 90-91. Ed. Kane, Part I, text pp. 49-50.

22 THE KĀPĀLIKAS AND KĀLĀMUKHAS

As he offered some black sesamum seeds, it seemed as though in eagerness to become a Vidyädhara he were annihilating the atoms of defilement which caused his mortal condition.42

As he muttered the syllables of his charms, his three disciples and the king stationed themselves about him in the four quarters. Suddenly a spirit (purușa) rose up from a chasm in the earth. This spirit, the Näga Srīkantha, attacked the king and Bhairavā- cārya's three disciples. The king used the sword Attahāsa given to him by Bhairavācārya to fell this Nāga. The king refrained from administering the final blow, however, because the spirit wore a sacred thread. Laksmī rewarded the king for his piety with the promise that he would become the founder of a mighty line of kings. Having completed the rite, Bhairavācārya acquired 'the hair-lock, diadem, earring, necklace, armlet, girdle, hammer, and sword' and became a Vidyādhara.43 The powerful spell called Mahākāla-hrdaya (Heart of Mahākāla- Śiva) is not mentioned elsewhere in Sanskrit literature to our knowledge. From the vividness of his description, however, it seems certain that Bana had some real ceremony in mind. Most remarkable is the fact that Bāņa portrays Bhairavācārya sympa- thetically. He is not a wicked magician but a worthy ascetic and a friend and confidant of the founder of the house of Bana's patron. From this fact we can only infer that by the seventh century tantric religion, even of the so-called 'left-hand observance' (vāmācāra) type, was accepted and supported by many persons of learning and high social status. As a corollary to this, it must also be assumed that the behaviour of most of these ascetics was considerably more circumspect than their critics would have us believe. Two epigraphs from western India show that even the Kāpalikas had at least some official support in the early mediaeval period.44 Another indication of public support for this sect is found in Bhavabhūti's Mālatī- Madhava. Although the two major villains of the play are Kāpālikas, one of the heroes, the yogini Saudāmini, is also said to observe the vow of a Kāpālika.45 A wicked counterpart to Bāņa's Bhairavācārya appears in Daņdin's Daśakumāra-carita (seventh century).46 Prince Mantra- 42Trans. Cowell and Thomas, p. 92. Ed. Kane, Part I, text p. 51. 43Trans. Cowell and Thomas, pp. 93-97. Ed. Kane, Part I, text pp. 51-54. 44See below, pp. 27-31. 45Ed. and trans. C.R. Devadhar and N.G. Suru, Act I, after vs. 15. 46Ed. and trans. V. Satakopan, V. Anantacharya, and N. Bhaktavatsalam.

KĀPĀLIKA SOURCES 23

gupta, one of the ten princes of the title, met this evil ascetic in a forest near the cremation ground outside of the capital of Kalinga. The prince overheard a servant couple complaining that their master, a black magician (dagdha-siddha), gave them no time to enjoy each other's company. They called out for someone to be an 'obstacle to the magical power of this vile wizard.'47 Prince Mantra- gupta followed them in order to discover who was this siddha and what was his siddhi. After going a short way the prince saw him. His body was decorated with ornaments made of pieces of human bones and smeared with ashes; his hair, matted in ascetic fashion (jatā), shone like lightning; and with his left hand he continually threw crackling sesame and mustard seeds into a fire. The magician ordered his servant to fetch Kanakalekhä, the daughter of the king of Kalinga. When the servant had done this, the magician attempted to decapitate the princess with his sword. Mantragupta rushed forth, seized the sword, and decapitated the magician insteud.48 This story may have been the basis for the similar Kapālika cpisode in Bhavabhūti's Mālatī-Mādhava.49 Several sources attest to the early presence of Kāpālikas and similar tantric ascetics in South India. We have already noted the references in Bana's works and the association of Kāpālikas with Śrīparvata in Mālatī-Mādhava. The most important South Indian source is the Mattavilāsa. This one act farce (prahasana) was composed by the Pallava king Mahendravarman, who ruled at Kāñcī between about A.D. 600 and 630. The leading character in the drama is a Kapālin ascetic who lives at the temple of Ekāmbira- nätha near the capital. We will discuss this work in more detail below.50 A contemporary of Mahendravarman, the Śaivite nāyanār Appar, refers to Śaivas, Pāśupatas and Kāpālikas in his vernacular songs.51 The Kapālīśvara temple at Mylapore, a suburb of Madras is the subject of a song by Sambandar (c. A.D. 644-660), another of the nayanars.52 The sixth or seventh century Tamil epic Manimekalai

47'asyāņaka-narendrasya ... siddhy-antarāyaḥ.' Ed. ibid., text pp.'213-14. 48Ibid., text pp. 213-15. 49See below, pp. 56-57. 50See below. pp. 54-55. 51See M.A.D. Rangaswamy, The Religion and Philosophy of Tēvāram, Book I, p. 392. 52 Tēvāram ii. 183, cited by T.V. Mahalingam, 'The Pasupatas in South India,' JIH, XXVII (1949), 47.

24 THE KĀPĀLIKAS AND KĀLĀMUKHAS

contains another brief reference to Kāpālika ascetics.53 Kāpālikas are mentioned disparagingly in several Purāņas. Although it is impossible to date these composite texts accurately, some of them-such as the Brahmānda, Vāyu and Matsya-date back to the third to seventh centuries A.D. The Vāyu, Brahmānda and Kürma Purānas assert that when the Kali-yuga is in full sway Kāşāyins, Nirgranthas, Kāpālikas, Veda sellers, tīrtha sellers, and other heretics opposed to varnāsrama-dharma will arise.54 Brahmānda ii. 29. 116-17 claims that Svayambhū (Śiva) created Pāśupata Yoga first and Kāpāla Yoga last. The Skanda Purāna prescribes, as part of the worship of the goddess Parameśvari, the distribution of pots of wine (surāsava) to Kāpālikas and male and female slaves.55

Kāpālika Epigraphy The epigraphical sources regarding the Kāpālikas are very few. Only two inscriptions register donations to Mahāvratin ascetics who are fairly certain to have been Kāpālikas. The term Kāpālika itself, however, appears in three inscriptions from southern Mysore State dedicated to their arch rivals, the Jains. Two are from famous · Shravan Belgola and one from Tirumakūdal-Narsipur Taluk in Mysore District. The earliest is from the former site and records the death by the Jain rite of sallekhan (fast unto death) of the Western Ganga king Mārasimha III (A.D. 960-974).56 It is written in Sanskrit and Old Kannada. The author compares Mārasimha to Śiva, Lord of the Kāpālikas : Famous was the glory of Mandalika-Triņetra (a Triņetra or Siva among the mandalikas or chieftains) as if to make the ... Käpālikas arrange in a string all the newly cut off heads of the Pallavas and firmly proclaim to hostile chief- tains-'Aho! Do not allow your newly cut off heads to be added to this string; have audience and live happily in the ranks of his servants.'57 53vi. 86, cited by K.A.N. Sastri, The Coļas, p. 94. See S.K. Aiyangar, Maņimēkhalai in its Historical Setting, p. 126. 54Vāyu Purāņa, Ānandāśrama edition, Iviii. 64-65; Brahmānda Purāņa, Venka- teśvara Press edition, ii. 31. 64-66; and Kūrma Purāņa i. 30 (p. 304). Kāpālins are included among the denizens of the Kali-yuga in Matsya Purāņa (Ānandāśrama edition, cxliv. 40). 55Venkațeśvara Press edition, Prabhāsakhaņda. 87. 51-52. 56Ed. and trans. R. Narasimhachar, EC, Vol. II (rev. ed.), no. 59. Also Ed. and trans. J.F. Fleet, Sravana Belgola Epitaph of Marasimha II, EI, V (1898-99), 151-80. 57Trans. Narasimhachar. The text is in Kannada.

KĀPĀLIKA SOURCES 25

The evident intent of this remarkable passage is to show the king's ferocity against his traditional enemies, the Pallavas. It does not necessarily imply any sympathy with Saivism. The Kāpālikas seem to be either religious mercenaries or simply battlefield scavengers. The possibility that they were militant religious mercenaries is strengthened by the description of a warlike Kāpālika band in Mādhavācārya's Samkara-digvijaya.58 The strings on which the Käpālikas of the inscription arrange the heads of the king's Pallava enemies are apparently the traditional skull garlands of these ascetics. The other two inscriptions, both written in Sanskrit, date from the twelfth century. The Shravan Belgola record commemorates the death by sallekhanā of the Jain preceptor Mallisena-Mala- dhärideva in A.D. 1129.59 The inscription from Tirumakūdal- Narsipur Taluk, dated A.D. 1183, commemorates the death in the same manner of a preceptor named Candraprabha.60 Both records give a lengthy priestly genealogy of the teachers whose deaths they honor. The genealogies show that both preceptors belonged to the same priestly line. Mallisena's name does not appear in the later record, however, unless it has been defaced. Both records quote a verse about an earlier teacher named Vimalacandra who hung up a letter-presumably a polemic document of some kind- addressed to the Kāpālikas and other opponents : To the gate of the spacious palace of Satrubhayamkara which is constantly thronged with passing troops of horses and numbers of mighty elephants of various kings, the high-minded Āśāmbara (i.e. Digambara) Vimalacandra eagerly affixed a letter (addressed) to the Saivas, the Pāsu- patas, the sons of Tathāgata (i.e. Buddha), the Kāpālikas, (and) the Kāpilas.61 Śatrubhayamkara, if this is a proper name, cannot be identified. Consequently, the date of Vimalacandra is uncertain. To add to the difficulties, the list of teachers 'is not a connected and complete account, and cannot even be proved to be in strict chronological

58See below, pp. 39-46. 59Ed. and trans. R. Narasimhachar, EC, Vol. II (rev. ed.), no. 67. Also ed. and trans. E. Hultzsch, 'Sravana Belgola Epitaph of Mallishena.' EI, 1II (1894-95), 184-207. 60Ed. and trans. B.L. Rice, EC. III, no. TN. 105. 61Trans. Hultzsch, EI, III, vs. 26.

26 THE KĀPĀLIKAS AND KĀLĀMUKHAS

order.'62 For the most part, however, the list does seem to be in order since a few of the teachers can be dated. Vimalacandra is the nineteenth of forty teachers named in the Mallișena epitaph. The Pārśvanātha-carita, composed by Vādirāja in A.D. 1025, says that Vimalacandra was the disciple of Matisagara, who was the disciple of Śrīpāla of Simhapura.63 In the epitaph list, however, the eighteenth teacher is one Puspasena, who was a contemporary of number seventeen, Akalanka. Three verses which Akalanka addressed to a king named Sähasatunga are quoted. In the last verse he claims that 'in the court of the shrewd king Himaśītala, I overcome all the crowds of Bauddhas.'64 Akalanka's exploits are described in other Jain works such as the Rājāvali-kathe, the Akalanka-carita, the Akalanka-stotra, and Jinasena's Ādipurāņa.65 These works identify Himasītala's capital as Kāñcī. According to the Akalanka-carita, Akalanka defeated the Buddhists in year 700 of an era referred to as Vikramārka-Śakābdīya.66 This obscure term might denote either the Vikrama or Saka eras. If the latter- the era most often used in the region-the date would be equivalent to A.D. 777-78. If the former, it would be A.D. 642-43. Another work dates his victory sometime after the year. Sālivāhana-Śaka 710 during the reign of 'Himasitala-mahārāja.'67 This is equivalent to A.D. 787-88. B.L. Rice claims that 'the Jains have for the date [of Akalańka's victory] the memorial sentence sapta-śailādra which gives 777 Saka =855 A.D.'68 This chronogram is probably merely a mystical number. If the suggested date for Jinasena's Ādipurāņa (between A.D. 782 and 838)69 is correct, A.D. 855 is too late for Akalanka. The Pallavas ruled at Kāñcī during the eighth century, but none of their kings was called Himasītala. Nonetheless it is best to place Akalanka in this period. The king named Sāhasatunga to whom Akalanka made his claim might then be the Rāstrakūta king Dantidurga (c. A.D. 733-758).70 If Vimalacandra came not long after Akalanka, he must have lived in about the first half

62Ibid., p. 185 63Narasimhachar, EC, II (rev, ed,), intro. p. 84. 64Trans. ibid., no. 67. 65See Hultzsch, EI, III, 187. 66See Narasimhachar, EC, II (rev. ed.), intro. p. 84. 67W. Taylor, Catalogue, III, 436f., cited by Hultzsch, EI, III, 187. 68 EC, II, intro, p. 45. 69Hultzsch, EI, III, 187 70See B.A. Saletore, Mediaeval Jainism, pp. 34-37.

KĀPĀLIKA SOURCES 27

of the ninth century. Teacher number 21 in the Mallisena list, Paravādimalla, is said to have spoken in the presence of a king named Krşņarāja. This might be the Rāştrakūța king Krsņa II, who ruled between A.D. 877 and 913.71 The three inscriptions thus indicate that Kāpālikas were present in southern Mysore during the ninth and tenth centuries. This helps to give credence to the legend of Samkarācārya's encounter with some of these ascetics in the Karnata region.72 We know from Mahendravarman's Mattavilāsa that Kāpālikas already existed elsewhere in South India at the beginning of the seventh century.73 The two grants which register donations to Mahävratin ascetics who must have been Kāpālikas are from western India. Both connect these ascetics with the god Śiva in his Kapālin or Kāpāleś- vara form. The earlier grant is a copper plate issued by the early Cālukya Nāgavardhana, son of Jayasimha and nephew of Pulakeśin II, sometime about the middle of the seventh century.74 It was found in the possession of a resident of Nirpan near Igatpuri in Nasik District. Nägavardhana informed all present and future kings :

Be it known to you that ... the village of Balegräma, which lies in the district of Goparāstra, has been given by us, at the request of Balāmma-Thakkura, ... for the purpose of the (rite called) Guggula-pūjā of the temple of (the god) Kāpāleśvara, and ... to the great ascetics [Mahāvratins] who reside at that (temple).75 Balegrāma has been identified as modern Belgaum-Tarāļhā about twelve miles north-east of Igatpuri.76 The fact that the god is called Kāpāleśvara (Lord of the Kāpālas) and not Kapāleśvara (Lord of the Skull) helps to confirm that the Mahävratins of the temple were Kāpālikas. The term Guggula-pūjā probably denotes the penance of placing hot or burning guggula (bdellium, a fragrant gum) on one's head. We have noted how the subjects of Harsa's

71See Rice, EC, II, intro. p. 47. 72See below, p. 43. 73See above, p. 23. 74Ed. and trans. J.F. Fleet, Sanskrit and Old Canarese Inscriptions,' 1A, IX (1880), 123-25. Also ed, and trans. R.G. Bhandarkar, 'A Revised Transcript and Translation of a Chalukya Copper-plate Grant,' JBBRAS, XIV (1878-80), 16-28. 75Trans. Fleet, IA, IX, 125. 76Ibid., 123.

28 THE KĀPĀLIKAS AND KĀLĀMUKHAS

father performed this penance to avert his death.77 Bhairavācārya, the priest of Harsa's ancestor Puspabhūti, also practised it.78 In Somadeva's Yaśastilaka a temple of Caņdamārī is said to contain devotees who were burning guggula on their heads and also Mahāv- ratikas who were selling flesh cut from their own bodies.79 The second Kāpālika grant, another copper plate, was found in the bed of the Narmadā at Tilakwādā in Baroda District.80 It registers the gift of a village named Viluhaja for a temple of Śrīghaņțeśvara and was issued from the temple of Maneśvara at the confluence of the Mana and Narmada rivers. The donation was made in A.D. 1047 by a feudatory or officer of the Paramāra king Bhoja. The donee was 'the muni named Dinakara, a Mahāvrata- dhara who was like the Kapālin, Samkara, in bodily form.'81 We know that Kāpālikas continued to exist in Gujarat until at least the twelfth century from the Moharājaparājaya of Yaśaḥpāla.82 Several inscriptions from various other parts of India mention Kapāleśvara temples, but none of these temples are said to contain Mahāvratin or Kāpālika ascetics. A copper plate grant of the mahāsāmanta and mahārāja Samudrasena, found in Nirmaņd village in Kängara District of Himachal Pradesh, records the donation of a village to a group of Atharva Veda Brāhmaņas in Nirmaņda agrahāra.83 The gift was to support worship of Siva in the form of Mihireśvara at a temple dedicated to Kapāleśvara. A king named Sarvavarman is said to have given land 'at the former installation of the god Kapāleśvara.' Fleet could not identify any of these kings but guessed that the grant belongs roughly to about the seventh century A.D. E.A. Pires has suggested that Sarvavarman might be identical with the Maukhari king of this name who ruled c. A.D. 576-580.84 If the Atharva Veda Brāh- maņas were in fact Kapalikas, it would appear that the sect claimed to follow this Veda. Many Saivite Tantras trace their authority to it.85

77See above, p. 17. 78Ed. Kane, Part I, text p. 46. 79See K.K. Handiqui, pp. 22, 358. 80Ed. and trans. J.S. Kudalkar, 'A Note on Tilakwada Copper-plate of the Time of King Bhoja Paramara of Malwa,' in All India Oriental Conference, Proceedings and Transactions of 1st Session, Poona, 1919, II, 319-326. 81Ibid., p. 324 (my translation). 82See below, p. 52. 83Ed. and trans. J.F. Fleet, CII, III, 286-91. 84 The Maukharis, p. 91. 85See Chakravarti, pp. 10-14.

KĀPĀLIKA SOURCES 29

A Kannada inscription from Lakkunda village in Hassan District, Mysore, records the esta blishment of an image of Vāsantikā- devĩ by a certain Mallideva who was an 'ornament to the Brahman family, brahmādhirāya, (and) mūliga of Kapāleśvara-devaru of Nekunda [=Lakkunda] in Nedunād.'86 The inscription is dated Śaka 777, but the paleography is typical of the tenth century. An inscription from a modern temple of Kavāljī (Kapālin) in former Kotah principality of Rajasthan contains an introductory verse to Gaņeśa and Kapālīśvara.87 The record is dated A.D. 1288 during the reign of Hammīra, a Cāhamāna king of Ranthambhor. The half verse in praise of Kapālīśvara, written in corrupt Sanskrit, shows that the god had special powers over disfiguring diseases : May the god Kapālīśvara through compassion manifest that which is desired of (i.e. by) men and destroy the pain of bodies spoilt by leprosy, elephantiasis, and cutaneous eruptions.88

Three thirteenth century inscriptions from the Kavileśvara temple at Ambale in Mysore District contain dedications to Kapāleśvara.89 The Huzur Treasury Plates from a Vișnu temple at Tiruvalla, or Tiruvallavāl, a taluk centre in Kerala, mention a village called Kapālimangalam and a temple of Kāpālīśvara at another village named Veļür.90 A record of about A.D. 1100 from the former Bastar State refers to a village called Kapālika.91 The most famous Kapāleśvara temple is located at Mylapore, a suburb of Madras. The seventh century Saivite saint Tiruña- nasambandar is said to have revived at this place a dead girl whose bones had been kept in a pot by her father.92 The present temple is of comparatively recent construction. A holy place especially associated with the purānic myth of Śiva-Kapālin, or Kapāleśvara, is the Kapālamocana tīrtha on the

86Ed. in ARMAD 1940, pp. 145-46. 87Ed. R.R. Haldar, 'Inscription of the Time of Hammir of Ranthambhor, dated (V.S.) 1345,' EI, XIX, 45-52. 88'dadru-ślīpada-kuşha-dușta-vapu[s]ām ā[dh]i[m] vini[gh] na[nn]rņāņ kāruņyena samīhitaņ vitanutām [deļvaḥ Kapālīśvaraḥ.' Ed. ibid., p. 49. 89Ed. and trans. B.L. Rice, EC, IV, Part II, nos. Y1. 6, 7 and 8. 90Ed. T.A.G. Rao, 'The Huzur Treasury Plates Belonging to the Vishnu Temple at Tiruvalla,' TAS, II, 156 and 142. 91 Ed. R.B.H. Lal, 'Kuruspāl Stone Inscription of Someśvaradeva,' EI, X, 25-31. 92See V. Venkayya, 'Triplicane Inscription of Dantivarman,' EI, VIII, 290.

30 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Ganges at Varanasi. It was here that Siva was released from his curse and allowed to abandon his skull begging bowl.93 A copper plate grant of the Gāhadavāla king Govinda-candra (A.D. 1114-54) states that the king bathed at this place and then donated a village to a Brāhmaņa named Vyāsa.94 K.C. Panigrahi argues that the well-known Vaitāl temple in Bhubaneshwar, Orissa, was originally a Kāpālika shrine.95 This temple, built in about the eighth century, has Cāmuņdā as its presiding deity. In all likelihood it was originally named after this goddess. The Svarnnädri-mahodaya states that 'the venerable goddess Cāmuņdā garlanded with skulls exists at a spot on the west not far from the tank ... ,' and that 'she is of terrific form and is known as Kāpālinī.'96 This must refer to the Vaitāl temple. Although this solitary reference to Cāmuņdā as Kāpālinī cannot be taken as conclusive evidence of Kāpālika worship, other features of the temple-such as the fierce deities sculptured around the inner shrine and a panel of erotic couples between the walls and roof-at least indicate tantric influence. Another temple which may have been associated at one time with the Kāpālikas is the famous Pasupati temple near Kathmandu in Nepal. An inscription from this temple, written in Sanskrit and belonging to the reign of King Jișnugupta (c. A.D. 630), records a gift to 'Varāhasvamin, Dharma ... and to the Somakhaddukas in the congregation of the Muņdaśrnkhalika-Pāśupatācārya ... '97 The term Mundasrnkhalika (Wearing a Chain of Heads) is more appropriate for a Kāpālika than a Pāsupata. The term Somakhad- duka (Wearing Moon Bracelets) is unique but reminds one of Soma-siddhānta, the name of the Kāpālika doctrine.98 Sculptures of the god Kapāleśvara or Kapāla-Bhairava and the goddess Kāpālikā or Kapāla-Bhairavi are found in many early medieval temples, particularly in South India.99 Neither the

93See below, pp. 77-80. 94Ed. F. Kielhorn, 'Twenty-one Copper-plates of the Kings of Kanauj; (Vikrama-) Samvat 1171-1233,' EI, IV, 97-129. (plate no. H). 95 Archaeological Remains at Bhubaneswar, pp. 61, 233-34. 96Quoted ibid., p. 233. 97Ed. and trans. B. Indraji and G. Bühler, 'Inscriptions from Nepal,' IA, IX (1880), 174. 98See below, pp. 82-83. 99See ARMAD 1930, pp: 20, 46; ARMAD 1932, p. 11; ARMAD 1933, pp. 46-47, 95; ARMAD 1934, p. 41; ARMAD 1935, pp. 44-45, 48; ARMAD 1936, p. 32;

KĀPĀLIKA SOURCES 31

presence of such sculpture nor even the dedication of an entire temple to Kapāleśvara is proof of Kāpālika influence, but the copper plate grants from Nasik and Baroda districts show that at least some Kapāleśvara temples were at one time staffed by these ascetics.

Śamkarācārya and the Kāpālikas Some of the most valuable material about the Kāpālikas appears in the legendary biographies of the great Samkarācārya (c. A.D. 788-820). The most important, and probably the earliest, of the extant biographies are the Śamkara-vijaya, attributed to his disciple Anandagiri,100 and the Samkara-digvijaya, attributed to the famous Vijayanagar sage Mādhavācārya alias Vidyāraņya.101 Dhanapati- sūri's Dindima commentary on the latter work adds some extra detail but is mainly extracted from Anandagiri's account. A signi- ficantly different version of one of the legends is contained in a Kānphațā work, the Gorakșa-siddhānta-samgraha.102 None of these sources can lay much claim to historical accuracy. They are collections of stories handed down, embellished and invented during several centuries between the great theologian's death and their final redaction. Most modern authorities agree that the author of the Śamkara-vijaya was not Śamkara's disciple Ānandagiri but an obscure author of about the fifteenth century. Many scholars also believe that the author of the Samkara-digvijaya was not Mādhavācārya, the Vijayanagar rājaguru, but a later author who wrote under his name.103 Samkara's disciple could not have written the Śamkara-vijaya, but we can see no significant objection to Mādhava-Vidyāranya being the author of the Samkara-digvijaya, particularly since the quality of its Sanskrit verse is excellent.

ARMAD 1937, pp. 3, 45, 58; ARMAD 1938, p. 5; ARMAD 1939, pp. 55, 57, 72; ARMAD 1940, p. 34; ARMAD 1945, p. 33. 100Ed. J. Tarkapanchānana. 101Ed. with Dhanapatisuri's Dindima commentary, Ānandāśrama edition. There are several other such biographies of Samkara but they are mostly inferior and of later date. Few have been published so far. One which has, the Samkara-vijaya of Vyāsācala (ed. T. Chandrasekharan), borrows most of its verses from Mādhava, often rearranging them in illogical order. 102Ed. G.N. Kavirāja. The work is ascribed to Gorakhnäth but is a collection of essays and stories by various later authors. 103Sarkar, A History of Dasnami Naga Sanyasis [sic], p. 20. G.S. Ghurye (Indian Sadhus, pp. 82-83) accepts Mādhava-Vidyāranya's authorship and places the Śamkara-vijaya Ānandagiri in the eleventh to twelfth centuries.

32 THE KĀPĀLIKAS AND KĀLĀMUKHAS

The Dindima commentary must be later than both these works. The Gorakşa-siddhānta-samgraha (henceforth GSS) dates from sometime in the later mediaeval period. There are three separate legends. The first of these, the story of Śamkara's encounter with a treacherous Kāpālika named Ugra- Bhairava, appears in Madhava's work and in the GSS; the second, amkara's battle with the militant Krakaca of Karņāțaka, appears in the works of Madhava and Anandagiri; and the third, Samkara's debate with the casteless hedonist Unmatta-Bhairava, appears in Anandagiri and is repeated in similar wording by Dhanapatisūri.

Śamkara and Ugra-Bhairava Śamkara's meeting with Ugra-Bhairava seems to have occurred somewhere along the Krishna River, perhaps at a spot near Śrīśaila (=Śrīparvata).104 Mādhavācārya begins his tale (xi. 1-2) :

Once a certain Kāpālika there, who hid his own wickedness by adopting the disguise of a sādhu like Paulastya (=Rāvaņa, in the abduction of Sita) and had not yet completed what he had set out to accomplish, saw the muni (Samkara) whose magical power (māyā) was limitless.

Thinking that his own ambition was as good as achieved, Ugra- Bhairava approached Samkara and greeted him with fulsome praise. The Kapalika then explained what he had 'set out to accomplish' (xi. 9-12) :

I will endeavour to please Kapālin (Siva) and thereby achieve my own object. I gratified Ugra (Śiva) with arduous and severe penances for a full one-hundred years in order to go to Kailäsa with this body to sport with Iśa (Śiva). Pleased, Giriśa (Śiva) said to me: 'You will attain the (ultimate) goal which men desire if, for the sake of pleasing me, you sacrifice in the sacrificial fire either the head of an omniscient sage or the head of a king.'

104The location is not explicitly stated. In the previous sarga, however, Samkara is said to have been travelling along this river from Śrīśaila. The Kāpālika's name is not mentioned in Madhava's text but appears in the commentary and in the GSS.

KĀPĀLIKA SOURCES 33

Having said this, Mahesa hid himself. From that time on I have wandered about, my hope fixed on obtaining that, but I have not yet found a (willing) king nor a (willing) omniscient sage.

There is little doubt who he had in mind. In order to persuade Samkara to accede to his implicit demand, Ugra-Bhairava then extolled the great benefits of self-sacrifice (xi. 13-16) : By good fortune I have now seen you, an omniscient sage, travelling about for the welfare of the world. Soon the rest (of my object) will be accomplished, for the bondage of men has its termination in correct vision ( samdar śanānta). The skull of an anointed king or a lord of munis is the prerequisite for my success (siddhi). The former, however, I cannot even conceive of (obtaining). Therefore, it is up to you. In offering your head you will acquire wondrous fame in the world, and I will acquire success (siddhi). After medi- tating on the transience of the body, O Best of Men, you should do what is propitious. I cannot dare to ask (lit., my mind cannot ask) for that. Who will (willingly) abandon his own body, the fulfiller of desires? (But) you are indifferent (to worldly desires) and care nothing for the body. (You have) assumed your own body (only) for the benefit of others.

Here he even attempts to turn Śamkara's own Vedāntic doctrines against him. Ugra-Bhairava then compared himself, with specious modesty, to those men who are ignorant of the pain of others and think only of their own ends. Such men, he said, are like Indra, who stole a bone from the sage Dadhyañc to use as an axe to slay the ninety-nine Vrtras. Men like Dadhyañc, who abandon their transient bodies for the sake of others, acquire an immutable body of fame (yaśah-śarira). Their priceless virtues delight all mankind. After several more verses in the same vein, Ugra-Bhairava finally made his request (xi. 24): 'You should bestow (your) head (on me). O Lord, homage to you!' Śamkara was apparently moved by

34 THE KĀPĀLIKAS AND KĀLĀMUKHAS

the Kāpalika's plea and agreed to grant him his desire. 'What true sage,' said Samkara (xi. 25), 'who knows the human body here in this world (to be subject to) decay, would not fulfill the request of a supplicant?' Samkara had to abide by the principles of his theology. Since the soul (atman) is the only ultimate reality, it matters little what becomes of the body. It is merely the creation of mäyā. Realizing that his pupils would never allow such idealistic foolhardiness, however, Samkara advised Ugra-Bhairava to visit him in secret. The two sealed their pact and Śamkara retired to an isolated spot hidden from his pupils. In full Kāpālika regalia, Ugra-Bhairava again approached to collect his reward (xi. 30, 32) :

(Holding) a trident, with three horizontal lines (drawn across his forehead), looking about (cautiously, wearing) ornaments made of garlands of bones, with his eyes inflamed and rolling about through intoxication, the yogin (Ugra- Bhairava) went to the dwelling place of the teacher ...

Beholding that (Käpālika) in the form of Bhairava, the teacher resolved himself to abandon his body

Śamkara then 'yoked himself with the ātman (ātmānam ātmany udayunkta).' Sitting in the proper yogic position (siddhāsana), he (xi. 35) 'forgot the whole world of creation in samadhi.' When Ugra-Bhairava saw him seated in this position, his fears were dispelled and he prepared to strike with his trident. No sooner did he come near to Samkara, however, than that sage's disciple Padmapada magically knew it (xi. 38, 42) :

Then, remembering the supreme power of the Man-lion (Vișnu's Nrsimha incarnation) held by Prahlāda, which removes the affliction of those who call it to mind, that (Padmapada), well-versed in mantras (mantra-siddha), be- came the Man-lion (incarnate) and saw the ill-intentioned endeavour of (Ugra-Bhairava) .. Running up with great speed he ... tore open with his claws ... the breast of (the Kapalika) who was striking . with his trident.

This ends Mädhava's version of the encounter.

KĀPĀLIKA SOURCES 35

The Gorakşa-siddhānta-samgraha belongs to the Kānphațā or Gorakhnäth (Goraksa-nātha) tradition, sometimes called the religion of the Näth Siddhas. Many of the tantric practices of its adherents resemble those attributed to the Kāpālikas. According to the GSS (p. 16), its philosophy is 'above dualism and monism (dvaitādvaita-vivarjita).' In these circumstances it is not surprising that the GSS version of the legend of Samkara and Ugra-Bhairava reflects less favourably on the advaita sage. Here the god Śrī- Bhairava himself assumed the form of Ugra-Bhairava in order to challenge Śamkara's religious beliefs and test their sincerity. The disguised god approached him and said (p. 16) : 'Sir, you are a samnyāsin (and hence) impartial to friend and foe alike and indifferent to the (opposite) senses of word pairs such as bliss and sorrow, etc.' He immediately requested the sage's head as an offering to Śri-Bhairava. By this means he would fulfill his vow (pratijñā). Śamkara carefully considered the alternatives (p. 16) : If it is not done (as the Kāpālika demands), then there will be the ruin of monism (advaita-häni) since there will not be impartiality towards friend and foe. If it is so done, defeat is (equally certain). Even in this twofold thought there is defeat (of pure non-duality). These unhappy alternatives completely baffled the great sage, and he could say nothing. Mädhava posed more or less the same problem but avoided carrying matters to their final philosophical absurdity by the commonsense intervention of Padmapada. The GSS retains this episode but refuses to let it go at that. After he was struck by Padmapäda-Nrsimha,105 Ugra-Bhairava mani- feșted his true identity as the god Śri-Bhairava. He then addressed Śamkara in a voice as deep as thunder : Sir, (this is) a defeat for advaita. What has become of that which you said about friend and foe? As a wrestler causes his opponent to fall by falling himself, (I have) accomplished the ruin of (my) opponent's (i.e. your) doctrine through the loss of my own body. Moreover, now you yourself will also meet your doom. Stand up, stand up! You should fight!106

105The third vowel in Padmapada's name is lengthened in the GSS. 106Pp. 16-17. The Sanskrit of this and the following passage is particularly barbaric. and some of the rendering is quite free.

36 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Śamkara was completely dumbfounded (p. 17) : Thinking, 'Then (if I fight), there will be an interruption of the work (I have) commenced since, in the doctrine of the nyāsins (= samnyāsins, i.e. my own doctrine), kriyā- karaņa (action or performance of rites) is not a (proper) doctrine,' he became powerless to fight as commanded by the Kāpālika. (Thinking), 'In my own doctrine akriyā ([inaction] is proper),' he stood (motionless) in accordance with his devotion to advaita.

The Kāpālika then created a magical power of Yoga (yoga-māyā) and employed it to cut off the heads of Samkara and his four disciples. Afterwards, however, they were revived. 'Then,' says the GSS (p. 17), 'true detachment arose.' Although it is clear that the author of the GSS wishes to condemn Śamkara's insistence on akriyā (inaction or non-performance of rites), he never really proposes any practical alternative. Whether Śamkara stood up and fought or not, his doom was equally certain. In a sense this paradoxical dilemma is well-suited to a doctrine which says it is 'beyond dvaita and advaita.' Similar statements are found in other tantric texts. In the Kulārnava-tantra, for example, Śiva declares: 'Some accept dvaita and others accept advaita, (but) they do not know my essence which is beyond dvaita and advaita.'107 Neither the Tantras nor related Kanphatā literature contain much systematic philosophical speculation. For the most part they are content to loosely synthesise the arguments and hypotheses of the orthodox systems. To a certain extent phrases such as 'beyond dvaita and advaita' merely proclaim the religious superiority of tantric doctrine. They do not necessarily imply any rational philosophical position. In a sense they are rejections of all rational metaphysics. It is not knowledge, but ritual, devotion and psycho-physical discipline (Yoga) which these schools empha- sise. In this context, there is no need for the GSS to propose an alternative course of action. The author needs only to point out the inadequacy of Samkara's position. The command to stand up and fight may be interpreted simply as a demand to symbolically acknowledge the self-defeating nature of the doctrine of akriyā. Samkara's beheading is therefore the occasion for the appearance of true detachment (virāga).108 107Ed. T. Vidyāratna, i. 110. 108It seems from this that the GSS not only wishes to criticize Śamkara but also

KĀPĀLIKA SOURCES 37

The attitude of the GSS towards the Kapālika Ugra-Bhairava is ambiguous. Sri-Bhairava assumes this form to challenge Samkara, but Ugra-Bhairava himself is neither praised nor condemned. Since both the Kāpālika and Kānphatā schools belong to the tantric tradition, however, one might expect that the Kānphatā attitude would be broadly favorable. That this is the case is made clear by some subsequent passages in the GSS. In the first of these the author asserts the superiority of the doctrine of the Näth Siddhas but allows the Kāpālika faith a qualified validity (p. 18) :

Indeed, some people believe that these (Siddhas) hold the Kāpalika doctrine on account of the mention of the devotion of the Kapalika,109 but that is not actually (the case). Our doctrine is beyond all worldly ties (avadhūta).

Nonetheless, the Kāpālika doctrine was also revealed by Nātha (Siva). Nātha was the revealer of this path.

The author then quotes from the Sabara-tantra a list of twelve sages to whom the Kāpalika doctrine was revealed : Ādinātha, Anādi, Kāla, Atikālaka, 110 Karāla, Vikarāla, Mahākāla, Kāla- bhairavanātha, Bațuka, Bhūtanātha, Vīranātha, and Śrīkaņțha. These twelve had twelve disciples: Nāgārjuna, Jadabharata, Hariścandra, Satyanātha, Bhīmanātha, Gorakșa, Carpața, Avadya, Vairāgya, Kanthādhārin, Jalandhara, and Malayārjuna. These pupils were the original 'promoters of the (Kāpālika) path (mārga- pravartakas).'111 Several of these names recur in traditional Känphatā lists of the eighty-four Siddhas and nine Nathas, most notably the name of Gorakhnäth (Goraksa) himself.112 On the basis of this statement and the common features in Kāpālika and Kanphațā worship. some modern authorities believe that the latter

to claim him for the Kanphata side. The section which immediately follows this passage describes Samkara's spiritual advancement through Visnu, Śiva and Śakti worship to his final enlightenment by the Mahäsiddhas and his adoption of the Path of the Nāthas ( Nātha-mārga). 109This apparently refers to the preceding Ugra-Bhairava episode. II0S.B. Dasgupta (Obscure ... , p. 207) replaces Atikālaka with Vaikālika. We do not know from where he got this reading. mGSS, pp. 18-19. 112For a summary of the various Kānphatā lists, see S.B. Dasgupta, Obscure .... pp. 202-10.

38 THE KĀPĀLIKAS AND KĀLĀMUKHAS

school was a later 'transformation' of the older Kāpālika order.113 As a historical document, however, the late GSS is virtually useless, and the similarities between the two schools-such as meat eating, drinking wine, attainment of magical powers through Yoga, dwelling in cremation grounds, and the like-are common to the whole of the tantric tradition. We feel, therefore, that such historical speculations are of little value. The author of the GSS next poses the question (p. 20) : 'For what reason was the Kāpālika path revealed?' The answer is found in a myth. Once the twenty-four avatāras of Vișņu became intoxi- cated with pride (mada, also =wine). As mortal creatures amuse themselves, so Varāha (Boar), Nrsimha (Man-lion), and the other avatāras began splitting the earth, frightening wild animals, oppressing towns and villages, and doing other mischief. Krsna was filled with adulterous emotions, and Paraśurāma destroyed a great number of ksatriyas to punish the sin of only one of them. Nātha became exceedingly angered by these wicked actions and assumed the form of twenty-four Kapalikas.114 In the ensuing battle the Kāpalikas cut off the heads of the avatāras and carried the skulls about in their hands. This was how the school of Kāpālikas (Skull-men) arose. The loss of their heads caused the avatāras to lose their pride as well. As a result they were granted a boon. Nätha replaced the skulls and returned them to life.115 The Ugra-Bhairava legend, whatever its historical value, and this myth both reflect the very real hostility between the tantric sects and Brähmanic orthodoxy. The fact that the author of the GSS chooses the Kāpālikas instead of the Siddhas to represent the Kānphata side of the dispute suggests not only that the two sects were on friendly terms, but also that the stories were already in popular circulation. Mādhavācārya's Vedāntic version of the Ugra-Bhairava legend is certainly older than the GSS's Kānphatā account. The battle between the Vinu's avatāras and the twenty-four Kāpālikas may reflect an extension of the conflict between the

113Ghurye, p. 128. See also G.W. Briggs, Gorakhnāth and the Kānphatā Yogis, p. 218. Eliade gives a somewhat more realistic appraisal of the situation in his work on Yoga (p. 218). 114These are presumably the same twelve teachers and twelve pupils mentioned earlier. 115GSS, p. 20. Note that Krsna is singled out for his adulterous emotions ( vyabhi- cāri-bhāva), a charge more frequently aimed at the Tāntrikas themselves.

KĀPĀLIKA SOURCES 39

Jains and Kāpālikas which is described in earlier sources. In many parts of India the Vaisnavas replaced the Jains in popularity and influence and in the process absorbed many Jain beliefs and prac- tices, including hostility to the excesses of tantric Śaivism.116

Śamkara and Krakaca or Bodholbaņa-nityānanda This legend also has two versions-one by Mädhava and the other by Anandagiri. Although the broad outlines of the two accounts are identical, several important differences indicate that they may have originated from separate traditions. Anandagiri sets his story in Ujjain while Mādhava sets his somewhere in Karņātaka. Mādhava calls Samkara's Kāpālika antagonist Krakaca and Ānandagiri calls him Bodholbaņa-nityānanda. The latter Kāpālika also has a disciple named Bațuka-nātha. Ānandagiri begins his account with a lengthy debate between Samkara and Bodholbaņa- nityananda which is omitted by Mādhava, 117 and Mādhava includes some semi-historical and martial detail omitted by Ānandagiri. Since Mädhava's version is generally more coherent and complete, we will base most of our discussion on it. According to this version, Samkara had begun a march to Setu (Rāmeśvaram) in extreme South India accompanied by his best pupils and a king named Sudhanvan. This was the start of a conquest of the four quarters (digvijaya). At Rameśvaram they met a number of non-Brāhmaņical and non-Āryan Śāktas whom Śamkara defeated in a great debate. The sage honored Lord Rāmanātha and converted the Colas, Pandyas and Dravidas. Next he proceeded north to Kāñci, constructed a beautiful temple there, and suppressed the Täntrikas by spreading Goddess worship in a form authorized by the scriptures (śruti-sammata). Proceeding towards the North- east he passed through Andhra, paid homage to the Lord of the Venkața hills (Venkațācaleśa), and eventually arrived at the capital of the Vidarbha kingdom (eastern Maharashtra) (xv. 1-7) :

116The most obvious example of Jain influence on later Vaisnava attitudes is found in the philosophy of M.K. Gandhi. Although some of Gandhiji's ideas were inspired in part by European precedents, others-such as his faith in ahimsā, asceticism, cleanliness and vegetarianism-owe more to his Gujarati Vaişnava background. From quite early times Gujarat has been a centre of Jain influence, and all these beliefs derive their original impetus more from Jainism than Hinduism. 117This debate contains much religious information and will be discussed below, pp. 83-85.

40 THE KĀPĀLIKAS AND KĀLĀMUKHAS

There the king of the Kratha-Kaiikas (Vidarbhas) approa- ched him with reverence and offered his worship. (Śamkara then) caused his pupils to suppress the heretical views of the followers of the Bhairava-tantra.

These 'followers of the Bhairava-tantra' are not identified, but they might be Kāpālikas since many authors depict Kāpālika ascetics as worshippers of Siva in his terrific Bhairava form. Krakaca himself is subsequently said to 'prattle the essence of the Bhaira- vāgamas.' At the least, these Bhairava-tantra followers must have belonged to some similar group of Tāntrikas. Mādhava continues (xv. 8-9) :

Then the king of Vidarbha bowed (to Śamkara, who) desired to proceed to the Karņāta region, and said: 'That region is unsuitable for your visit since (it is filled) with many crowds of Kapālins, 'I say (this) since they cannot endure your fame and have a secret hatred towards the scriptures (śrutis). They revel in the misfortunes of the world and bear hostility against honoured men.'

Samkara's royal disciple Sudhanvan guaranteed the sage protection, however, and they advanced 'to conquer the multitude of Kāpālikas' (xv. 10-14) :

When Krakaca, the foremost of the Kapalin teachers, learned of (Samkara's) arrival, he came to meet him.

Smeared with ashes from a cremation ground (pitr-kānana- bhasman), carrying a skull-bowl in his hand, weilding a trident, and accompanied by many whose appearance matched his own, that conceited and proud (Kāpālika) spoke thus : 'Although properly ashes are worn (by you),118 for what reason do you hold that impure (clay) bowl and renounce this pure and fitting skull? Why is not Kapālin worshipped (by you)?

118 Samkara was a nominal devotee of Siva and therefore wore the traditional Śaivite ashes.

KĀPĀLIKA SOURCES 41

'If He (Kapālin-Siva) does not receive Bhairava worship with liquor (madhu) and blood-smeared lotuses which are human heads, how can he attain joy when his body is embraced by the lotus-eyed Uma, who is his equal?'

After Krakaca 'had prattled thus the essence of the Bhairavā- gamas,' King Sudhanvan ordered his officials to send him away. The enraged Kāpālika soon returned with his followers to seek retribution for this insult. As they approached he shouted : 'I am not Krakaca (= a saw) if I do not cut off your heads' (xv. 15-17) :

He sent out the countless crowds (kulas) of angered Kapālins whose cries were as terrifying as the clouds of the deluge. They attacked with weapons held aloft.

The Brähmana followers of Samkara were terrified, but the faithful Sudhanvan countered the Kāpālika advance and drove them back. Krakaca then shifted the battle to another part of the field and again threw the Brahmanas into confusion. In desperation they sought Śamkara's protection (xv. 21) :

The king of ascetics (then) reduced those (Kāpālikas) ... to ashes in an instant through the fire which arose from his humkāra (the sound hum, a mantra).

Sudhanvan rejoined Samkara and slaughtered a thousand more of their enemies. Seeing his army routed, Krakaca again approached Śamkara and said (xv. 24-25) :

'O Devotee of Evil Doctrines, behold my power! Now you will reap the fruit of this action (karman).' Closing his eyes, (Krakaca) placed a skull in the palm of his hand and briefly meditated.

After that master of the Bhairavāgamas had thus meditated, the skull was immediately filled with liquor (surā). After drinking half of it, he held the (remaining) half and thought of Bhairava.

This god instantly appeared in the form of Mahākapālin. He wore a garland of human skulls and his hair was a flaming mass of

42 THE KĀPĀLIKAS AND KĀLĀMUKHAS

matted locks (jat). He held a trident and uttered loud and dreadful laughter. Krakaca commanded him (xv. 27) :

'O God, you should destroy the enemy of your devotee with your (fierce) gaze.' Instead the enraged (Mahākapālin declared), 'How dare you offend against my own self (i.e. Śamkara),' and cut off the head of Krakaca.

This ends Madhava's account. Dhanapatisūri's Dindima com- mentary, following Anandagiri almost verbatim, continues the story to the final conversion of the Kāpālika's disciples. In this version the god Samhara-Bhairava did not immediately kill Bodholbana- nityānanda. When the god appeared Samkara paid him homage and set forth his own philosophy in order to justify his action against the Kāpālika and his disciples. Bhairava was pleased by the sage's statements, however, and commanded him : 'You should make those Kāpālikas embrace the faith of the Brāhmanas.'120 The god explained that he had become manifest because he was bound by the mantra (mantra-baddha) used by Bodholbana, not because of any merit of that ascetic (na dharmatas). Samhāra- Bhairava then vanished and the followers of the Kāpālika doctrine ( Kāpālika-matānugas)-who were of twelve sorts, Batukas, etc .- bowed down to Samkara. The sage was filled with compassion and instructed Padmapāda and his other disciples to convert the repentant heretics.121 Unfortunately neither the commentator nor Ānandagiri identifies the 'twelve sorts (dvādaśadhā) of Kāpālikas beginning with the Bațukas.' Bațuka, however, appears as one of the twelve original Kāpālika sages in the Sābara-tantra list quoted in the Goraksa-

119xv. 26. Anandagiri's account of this episode is worth comparing. Seeing that he and his pupils had been beaten by Samkara, Bodholbaņa-nityānanda approached the sage and said (chap. xxiii): '"Samhara-Bhairava is to be honored by me. I will cause him to appear by means of mantras. He will quickly devour you and your retinue." Saying this and again uttering (the sound) hum; holding a human skull with his left hand; filling that with liquor by means of a mantra; drinking half (of it) himself and giving the remainder to his pupils; and looking into the middle of the sky with round and reddened eyes-he said: "(You), who are Samhāra-kāla (Destruction-time), Bhairava, Prabhu, and Iśvara, should come and speedily devour the samnyāsin and his retinue.'" 120 Dindima commentary, vs. 8, on Mādhavācārya, xv. 28. 121Ibid., vss. 1-14.

KĀPĀLIKA SOURCES 43

siddhānta-samgraha.122 Evidently these twelve sages were considered to be the founders of twelve divisions of the Kāpālika sect. The presence of this tradition in such unrelated sources suggests that there may have been some factual justification for it. The personalities of the two legendary Kāpālikas, Krakaca and Ugra-Bhairava, are quite distinct-where the latter used guile the former chose brute force-but in appearance Krakaca, like Ugra- Bhairava, is a typical Kāpālika. He smears his body with the ashes of the dead; he carries a trident and a skull bowl; he worships Bhairava and Mahākapālin; his text is the Bhairavāgama; he honours this god with liquor and offerings of human heads; and he imagines salvation as the indescribable bliss of an endless embrace in the arms of Umā. Both the location and the large size of Krakaca's Kāpālika battalions merit additional comment. In Mādhavācārya's version the Vidarbha king warned Śamkara against going to the Karņāta region because it was populated by 'many crowds of Kapālins.'123 Dhanapatisūri glosses țhis location as the town Ujjayanī (sic), but this cannot be correct. He apparently relies for this identification on Anandagiri, who begins his version (chap. xxiii) : 'Travelling along the northern road, Śrī-Śamkarācārya ... saw the city named Ujjayinī which was filled with (persons) devoted to the Kāpālika observance.' The Karņāta region approximately corres- ponds with modern Mysore State and never included the famous Mālava city, Ujjain.124 Evidently there were two separate traditions. Reasons exist for both these places to be associated with the Kāpālikas. Neither Mādhavācārya's Śamkara-digvijaya nor Ānandagiri's Śamkara-vijaya can claim much historical accuracy. Both are products of about the fourteenth or fifteenth century and both tend to shed as much light on the religious life of India during the century or two preceding their composition as on the religious life of

122See above, pp. 37-38. 123 Mādhavācārya xv. 11. 124It is tempting to identify the Ujjayini of Anandagiri and Dhanapatisūri with the town by that name in Bellary District, Mysore, whete one of the five chief mathas of the Viraśaivas is located. The matha at this place was supposedly founded by Marulasiddha, one of the five great âcāryas of Vīraśaiva tradition. See M.R. Sakhare, History and Philosophy of Lingayat Religion, pp. 361-62. Unfortunately Anandagiri's statement that Samkara reached Ujjayini 'travelling along the northern road makes this identification less likely.

44 THE KĀPĀLIKAS AND KĀLĀMUKHAS

the time of Samkarācārya. The Karņātaka region, which seems to have been the home of Ugra-Bhairava as well as Krakaca, was domi- nated by the Kälämukhas during the eleventh to thirteenth centuries. Since no lesser authorities than Yāmunācārya and Rāmānuja associate, and perhaps confuse, the two sects, there is at least a prima facie case that Mädhava did the same. Krakaca's dress, behaviour and religious beliefs are definitely those of a Kāpālika, not a Kālāmukha, but in one important respect he and his followers have more affinity with the latter sect. In Mädhava's story Krakaca is said to command vast legions of Kāpālikas ( Kapāli-jālāḥ, Kāpālika-jālakaḥ, Kapālinām kulāni). Nearly every other story featuring Kāpālikas describes them as solitary peripatetic ascetics, occasionally joined by a single female disciple. This absence of organization may help to explain the relative lack of Kāplika epigraphy. The Kālāmukhas, on the other hand, usually established themselves in large monastic communities. It seems quite likely that Mādhava was modelling his Kāpālika legions on the brotherhoods of the Kālāmukha mathas. As in the accounts of Yamunacarya and Rāmānuja, the confusion between the two sects may have been intentional. This would help explain the absence of any mention of Kālāmukhas in Mādhava's work. Although in each chapter of Ānandagiri's Śamkara-vijaya Śamkara debates a different rival sect, the Kālāmukhas do not appear in this work either. Since the Mālava Ujjain was never a center of the Kālāmukhas, however, it is less likely that Ānandagiri was confusing the two sects. There is a tenuous connection between this town and the Käpalikas in the fact that Bhavabhūti wrote his Malati-Mādhava for the festival of Lord Kālapriya, who is usually identified with the god Mahākāla of Ujjain. The play is set, however, in Padmävatī, a town which scholars locate some 220 miles north of Ujjain near modern Narvār.125 Today Ujjain is an important center of the Kānphata yogins.126 If Gorakhnāth's commonly accepted date, c. A.D. 1200, is correct, this town may well have been a Kānphațā center by the time of Anandagiri (about the fifteenth century).127 Since Kānphatā Yogins also organise them-

125See the introduction to Devadhar and Suru's edition of the play. p. 4. 126Ghurye, p. 137. 127Although Anandagiri's date is not certain he seems to have lived sometime after Mādhavācārya. Since the Kālāmukhas were already rapidly declining in Mädhava's time, this is another reason why it is unlikely that Anandagiri was con- fusing them with the Kāpālikas.

KĀPĀLIKA SOURCES 45

selves into monastic communities, Änandagiri might have confused them and the Kapālikas. Whether Krakaca's ascetic legions are modelled on the orga- nization of the Kālāmukhas, Kānphațās, or Kāpālikas themselves, their militancy is quite striking. Military orders of religion were not unknown, however, either in mediaeval India, mediaeval Europe, or sixteenth century Japan. In Europe the Crusades produced several military orders, the most famous being the Templars and the Hospitallers. These orders not only fought against the Muslim princes of the Holy Land but also on occasion joined forces with these very princes against each other.128 In the sixteenth century Japan witnessed the Ikko Ikki or Fanatic Risings by monks of the Pure Land (Jodo) sect, who fought with their sectarian enemies and in many places even challenged the authority of the feudal lords.129 Although the mutual tolerance shown by religious groups native to India has always been remarkable, rivalry among them, especially for royal patronage, sometimes led to violence. In ancient times competition usually took the form of great public debates which often became miracle contests. One such contest between Buddha and the Ājīvika teacher Pūraņa Kassapa took place at Śrāvasti in North India. It seems to have ended with some sort of riot in which the Ajivikas were expelled.130 In some cases defeat in debate led to royal persecution. According to a South Indian legend, the famous nāyanār Nānasambandar once vanquished the Jains in debate and converted the Pāndya king to Saivism. The king then executed 8,000 Jains by impalement.131 The Kālāmukhas are frequently extolled for their debating skill. but most of their debates -- like that between Bonteyamuni and some rival logicians132 -seem to have been peaceful ones. In about A.D. 1160, a debate cum miracle contest between the Vīraśaiva leader Ēkāntada Ramayya and the Jains at Ablur in Dharwar District, Mysore, ended with the defeat of the latter. When the losers refused to abide by a previous agreement to set up a Siva idol in place of their Jina, Ēkāntada Rāmayya marched on their temple, defeated

128A.S. Atiya, Crusade, Commerce and Culture, pp. 67-68. 129G.B. Sansom, Japan : A Short Cultural History (rev. ed.), pp. 374-76. 130See Basham, pp. 84-87. 131See K.A.N. Sastri, A History of South India, p. 413, 132See below, p. 132.

46 THE KĀPĀLIKAS AND KĀLĀMUKHAS

its defenders and demolished all the buildings.133 The six major ākhādās of the Daśanāmi Nāgās are the earliest recorded examples of true religious military orders in India. These ākhādās (regiments) are still in existence. Nominally at least, their members belong to one or other of the ten orders of Saivite ascetics reputedly founded by Śamkarācārya (the Daśanāmis) .The Junā Ākhādā (Old Regi- ment) was formerly known as the Bhairava Ākhādā. Its present tutelary deity is Dattātreya but originally must have been Bhairava. The traditional date for its establishment is A.D. 1146, but Ghurye (p. 104) believes that it is descended from an older sect of Saivites, namely the Kāpālikas. This is a tempting suggestion, particularly since one of the ākhādā's centers is at Ujjain. It is difficult to see, however, how it could have survived a transition from the Kāpālika faith to Vedänta. The traditional dates for the foundation of some of the other ākhādās go back to as early as A.D. 647, but the earliest reasonably verifiable date for an actual battle involving Nāgā Samnyāsins is A.D. 1266. Most of their recorded activity belongs to the sixteenth to eighteenth centuries and culminates in a great victory over the Vaișnava Bairāgis at Hardwar in 1760.134 The preceding discussion suggests an interesting but admittedly hypothetical chain of events. An original historical debate between Śamkara and some Kāpālika ascetics either at Ujjain or somewhere in Mysore ended in a riot during which the Kāpālikas were put to flight and some possibly converted. In succeeding centuries this story was gradually elaborated until the original antagonists became a vast army of warlike monks modelled in part either on the Kālāmukha monastic orders (Mādhava) or on the newly emerging military orders of medieval India (Ānandagiri).

Śamkara and Unmatta-Bhairava This legend appears in Anandagiri's Samkara-vijaya, chapter xxiv, and in Dhanapatisūri's Dindima commentary on Mādhava's Śamkara-digvijaya xv. 28. The two accounts are almost the same and show the commentator's debt to Anandagiri. Since the published text of Änandagiri's work is corrupt, we will follow the commentary wherever possible. The introduction to the story is found only in Ānandagiri's account :135

133See J.F. Fleet, 'Inscriptions at Ablur, 'EI, V, 213-65 (inscription no. E). 134Ghurye, pp. 103-112. 135Dhanapatisūri omits this passage and grafts his story directly onto the Krakaca legend.

KĀPĀLIKA SOURCES 47

After they had been thus repudiated, the various groups (varņas) -Kāpālikas, Cārvākas, Saugatas, Kșapaņakas, Jainas, and Bauddhas-reappeared in another town. A certain Kāpālika of the Sūdra caste (jāti) named Unmatta- Bhairava (lived) there. His body was covered with ashes from a funeral pyre; his neck was ringed with a garland of human skulls; (three horizontal) streaks of lamp-black were drawn across his forehead; all his hair was fashioned into a top-knot (jatā-jūta); his waistband and loincloth were made from a tiger skin; a skull-bowl adorned his left hand; his right hand held a loudly ringing bell (ghantā); and he was chattering repeatedly 'O Sambhu-Bhairava! Aho, Kālīśa!

This classic description of a Kāpālika ascetic is followed by a derisive exposition of his hedonistic doctrine, a doctrine more suitable for a Cārvāka or Lokāyatika than a Kāpālika. Unmatta- Bhairava came to meet Samkara and proclaimed to him the superiority of the Kapalika faith over all others. He lamented the fall of Bațukanātha, Bodholbaņa's disciple, and the other Kāpālikas from the true faith and said : 'Their undoing (was their reliance on) being of the Brähmana caste. I will have nothing to do with caste. '136 He then gave a materialistic justification for his rejection of caste and suggested that there are only two real castes, the male-caste and the female-caste. He also asserted that promiscuity is the proper rule of conduct between them since the joy (ānanda) of sexual union is the true form of Bhairava, and the attainment of that joy at death is salvation (moksa).137 Śamkara listened politely to the Kāpālika's blasphemy and said (vs. 23): 'O Kāpālika, this was well said. (But) the truth should be told. Whose daughter is your mother?' Unmatta-Bhairava retorted that his mother was the daughter of a dīkșita (initiated priest) and explained his contention thus (vss. 23-28) :

O sage, He (my mother's father) extracts the toddy (surā) of the best palm trees. Though he knows well its taste, he does not wish to drink it himself but with due devotion

136Dhanapatisūri, vss. 15-16. The name Bațukanātha is from Ānandagiri. The commentary reads 'Batukas and others.' 137Dhanapatisūri, vss. 17-22. See below, pp. 90-92.

48 THE KĀPĀLIKAS AND KĀLĀMUKHAS

( śīlavān ) sells it. Therefore people always call him dīkșita.138 His daughter became my mother. By making an offering of her own body, O sage, she always caused men, who had come (to her) for the sake of pleasure, to be immersed in an ocean of bliss. Know this (person) named Unmatta-Bhairava (to be) her son. My father was also a liquor maker.139 Even the gods (suras) approach him here on earth, and they by no means run away (because they are) averse to the smell of liquor.

Having tricked Unmatta-Bhairava into this self-condemnation, Samkara good-naturedly ordered him to leave and wander about wherever he wished. To his own disciples Samkara explained (vss. 28-30):`I have come only to punish Brähmanas who have embraced a bad faith and not others. This man is not to be spoken to. Take him away quickly.' This legend provides a good example of the accusations of hedonistic licentiousness which orthodox writers are fond of leveling against tantric ascetics. These accusations have some factual basis. Sex and alcohol, for instance, do play an important part in tantric ritual. In the dakșiņācāra tradition symbolic equi- valents are substituted or the rituals sublimated into mere mental exercises, but in the vāmācāra tradition of the Kāpālikas real women and wine were employed. Nonetheless, the assertion that the Käpalikas were hedonists and that they justified this hedonism with a thoroughly materialistic philosophy cannot be accepted. Materialistic hedonism falls within the province of Lokāyata and Carvaka philosophy, not tantricism. This is not to say, of course, that many tantric ascetics were not licentious. Unmatta-Bhairava's orgasmic conception of moksa and his rejection of caste, however, are at least partly confirmed by other sources and will be discussed in more detail elsewhere.140

Kāpālikas in Sanskrit Drama Villainous Kāpālika ascetics appear in a number of Sanskrit dramas and stories. The dramas include : (1) Mattavilāsa by the

138A dīkșita makes sacrificial offerings to the gods which he does not himself consume. Unmatta's maternal grandfather sells liquor but does not drink it. 139There is an implied pun on the words surā-kara (liquor maker) and sura-ākara (treasure of the gods). 140See below, pp. 90-92. and above, pp. 6-7.

KĀPĀLIKA SOURCES 49

Pallava king Mahendravarman (c. A.D. 600-630); (2) Mālatī- Mādhava by Bhavabhūti (c. 725); (3) Caņdakauśika by Kșemīśvara (c. 900-950); (4) Prabodhacandrodaya by Krşņamiśra (c. 1050- 1100); (5) Lațamelaka by kavirāja Śankhadhara (c. 1110-50); (6) Kaumudīmitrānanda by Rāmacandra (c. 1143-75); (7) Nalavilāsa by the same author; and (8) Moharājaparājaya by Yasaḥpāla (c. 1175).141 Two very late works in which these ascetics appear are the Amrtodaya of Gokulanātha (A.D. 1693)142 and the Vidyā- pariņayana attributed to Ānandarāyamakhin (c. 1700).143 They are also mentioned in Kavikarņapūra's Caitanyacandrodaya (c. 1550).144 The Prakrit drama Karpūramanjarī by Rājaśekhara (c. 900) features a tantric 'master magician' named Bhairavananda, who `follows the kula path.'145 The Rucikaratīkā on Krsnamiśra's Prabodha- candrodaya claims that Rājaśekhara's Bhairavānanda followed Somasiddhānta, the doctrine of the Kāpālikas. 146 This is not strictly correct. 'The kula path' refers to the doctrine of the Kaulas, not the Kapalikas. Both these sects belonged to the Vāmācāra tradition, however, and had many similarities. Kaulas also appear in Yaśaḥpāla's Moharājaparājaya and Sankhadhara's Lațamelaka. None of these dramatists had much sympathy for the Kāpālikas. According to tradition, Mahendra, the royal author of Mattavilāsa, was converted from Jainism to Saivism by the famous Tamil nāyanār, Appar. 147 The king's own Tiruchchirappalli (Trichinopoly)

141A.B. Keith (The Sanskrit Drama, p. 254) places Yasahpāla during the time of 'Abhayadeva or Abhayapala, who reigned after Kumarapāla from A.D. 1229-32.' C.D. Dalal, in his introduction to Chaturavijayaji's edition of the Moharājaparājaya, places Yaśahpāla during the time of Ajayadeva or Ajayapāla, who `reigned from A.D. 1229 to 1232.' Ajayapāla, the Caulukya king of Gujarat, ruled from A.D. 1172 to c. 1176. The dates given by Dalal and apparently copied by Keith belong to the Vikrama era, not the Christian era. We do not know why Keith read the king's name as Abhayapāla rather than Ajayapāla. 142Keith (p. 343) dates this author in the sixteenth century, but this is incorrect. 143 Handiqui (Srīharsa. p. 641) follows a brief note in Sivadatta and Parab's edition of the play (p. 1) and dates Änandaräyamakhin in the first half of the seventeenth century. Keith (p. 253) gives the correct date. See M.D. Aiyangar's introduction to his edition of Ānandarāya's Jīvānanda. 144Ed. Kedāranātha and W.L.S. Paņśīkar, pp. 24-25. 145Ed. S. Konow and trans. C.R. Lanman. See especially Act I, vss. 22-25 and Act IV, vs. 19. 146Cited in G. Tucci, 'Animadversiones Indicae,' JRASB, n.s. XXVI (1930), 131. 147R. Sathianathaier, K.R.S. Iyengar, and T.M.P. Mahadevan (all in R.C. Majumdar [ed.], The Classical Age) seem to accept this tradition without question. K.A.N. Sastri, in his Development of Religion in South India (p. 42), points out that the identification is based on slender evidence.

50 THE KĀPĀLIKAS AND KĀLĀMUKHAS

record indicates that he turned to Saivism from some other faith.148 Even if he wrote his play after his conversion, he could not be expected to favor a heretical sect such as the Kāpālikas. Bhavabhūti was an orthodox Brahmana from a family which followed the Taittirīya branch of the Yajurveda. He may have had some Śaivite leanings but was not a strict sectarian.149 It is likely that Ksemīśvara, the author of Candakauśika, was also an orthodox Brāhmaņa. Krsnamiśra used his allegorical Prabodhacandrodaya to extol the merits of advaita Vaişnavism. Sankhadhara seems to have favored some moderate form of Saivism since his Latamelaka opens with introductory verses to Gaurī and Siva. Rāmacandra and Yaśaḥpāla were both devout Jains. All these writers express their contempt for the orgiastic and sadistic features of the Kāpālika cult. This is particularly true of the Jains, whose extreme asceticism made them natural enemies of tantricism. The distain of the orthodox Saivite writers, on the other hand, probably reflects a desire to disavow any association with their heretical brethren. It is significant, perhaps, that Krsna- miśra, a strong Vaişņava, attacks a Cārvāka, a Digambara, a Buddhist, and a Kāpālika but neglects to mention any of the more respectable Śaivite sects. The dramas provide some important additional information about the geographical distribution of the Kāpālikas. The Matta- vilāsa is set in Kāñcī, the capital of Mahendra. The Kāpālika ascetic of the play is said to live at Ekambam (eamvvavāsī).150 This must refer to the Ekämbira-nātha temple, which is still one of the major temples of Kanchipuram. In its present form, however, the building dates from a later time than Mahendra's. Bhavabhūti's Mālatī- Mādhava, as we have noted, is set in Padmāvatī, a town which was probably located about 100 miles south of Agra. The Kāpālika Aghoraghanta operates from a temple of Karālā situated in the cremation ground of that town, but his home is said to be the mountain Sriparvata.151 Another character, the virtuous yogini Saudāminī, is said to be 'observing the vow of a Kāpālika on Sri- parvata.'152 The heroine Mālati is eventually abducted to this place by Aghoraghaņta's female disciple, Kapālakuņdalā.

148See ibid. 149Bhavabhuti's name means either 'wealth of Siva' or 'ashes of Siva.' 150Ed. T.G. Śāstrī, p. 13. Trans. L.D. Barnett, BSOS, V, 707. 151Act I, after vs. 15. 152Ibid.

KĀPĀLIKA SOURCES 51

The famous holy center Śrīparvata (also called Śrīśailam) is located in Kurnool District of Andhra Pradesh. It is mentioned in the Mahabharata as a place sacred to Siva and Devi.153 In Matsya Purāna clxxxi. 28-29, it is listed as one of eight secret places sacred to Śiva. The Mallikārjuna linga at the site is one of the twelve jyotirlingas of Siva.154 Today the shrine is held in especial esteem by the Viraśaivas. The earliest references to Mallikārjuna worship on Śrīparvata are found in Subandhu's Vāsavadattā (c. A.D. 600)155 and the Padma Purāņa.156 Mādhavācārya claims that Samkarācārya himself visited the shrine of this god on Śrīsaila.157 In A.D. 1090 a Kālāmukha priest of the Parvatāvali named Rāmeśvara-paņdita was the head of the Mallikārjuna-silā-mațha on this mountain.158 A few years earlier, in 1057, the Western Cālukya king Someśvara I came to Śriaila and washed the feet of the Kālāmukha teacher Sureśvara-pandita in the presence of the god Mallikārjuna.159 In Bāņabhatța's Kādambarī, as we have seen, 160 the South Indian tantric priest who lived in the Caņdikā temple near Ujjain is said to know 'thousands of wonderful stories about Śriparvata.' In his Harșa-carita; Bāņa calls Harșa a 'Śrīparvata of magical powers (siddhis).'161 These two references by Bāņa indicate that Śrīparvata was already famous as a center of tantric worship by the first half of the seventh century, but Kalhana is the only author besides Bhavabhūti to specifically connect it with the Kāpālikas. 162 Some- time before the eleventh century the temple of Mallikārjuna came into the hands of the Kālāmukhas. We do not know what sort of relations, if any, they maintained with the Kāpālikas. It is not likely that the two sects were ever on very friendly terms. By about the fourteenth century the Kālämukhas had been replaced by the Vīrasaivas. The Kāpālikas seem to have become virtually extinct by this time. Most of the other dramatists lived in northern and western India.

153 Āranyakaparvan, ed. V.S. Sukthankar, iii, 83. 16-17. 154See P.V. Kane, HDS, IV, 678. 155Trans. L.H. Gray, p. 68. 156 Uttarakhanda, chap. xix, cited by Handiqui, p. 359. 157 Śamkara-digvijaya, x. 7-12. 158See inscription ed. and trans. by P. Sreenivasachar, HAS, No. 13, Part II, pp. 25-31. 159See inscription ed. by R.S. Sastry and N.L. Rao, SII, IX, Part I, no. 119. 160See above, p. 18. 161 Ed. Kane, Part I. text p. 2. 162See below, pp. 66-67.

52 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Kşemīśvara, the author of Caņdakauśika, dedicated his play to Mahīpala, who may best be identified with Mahipāla I (c. 912-942), the Pratihära king of Kanauj. This king was also the patron of Rājaśekhara. Krsnamiśra composed the Prabodhacandrodaya for Kīrttivarman (c. 1070-90), a Candella king of Jejakābhukti (modern Bundelkhand region). Sankhadhara's Latamelaka was written during the time of Govindacandra (c. 1114-54), a Gāhadavāla king of Kanauj. Rāmacandra and Yasapāla both lived in northern Gujarat during the reign of the Caulukya king Ajayapāla (c. 1172- 76). Yaśaḥpāla's Moharājaparājaya describes the conversion to Jainism of Ajayapāla's predecessor, Kumārapāla (c. 1143-72). A Kāpālika is included among the enemies of this king. In Rāma- candra's Nalavilāsa, a reworking of the Nala-Damayanti legend, some Kāpālikas appear as spies of Citrasena, a Kalacuri-Cedi king. This king is probably modeled on one of the Kalacuri kings of Tripuri (near Jabalpur, Madhya Pradesh) since one or more of these kings is believed to have fought with Kumārapāla.163 The play is set in Vidarbha. If all the information about the distribution and dates of the Kāpālikas is collated, we find that they existed throughout most of the Deccan plateau as early as the eighth century. They are connected specifically with Kāñci, parts of Mysore, western and central Maharashtra, Ujjain, the Gwalior region of Madhya Pradesh, and Kurnool District in Andhra Pradesh. They may also have been found in Orissa (Bhubaneswar) by this date. Sources later than the eighth century indicate their presence in Gujarat, Bundelkhand, the Vindhya hills, and other parts of India.164 According to Bhavabhūti, Śriparvata in Kurnool District was a particularly important Kāpālika center in his time. In later centuries, however, they were replaced at this site by the Kālāmukhas. Areas from which we have no pre-ninth century records of Kāpālikas include Bihar, Uttar Pradesh (except Varanasi), Rajasthan, Punjab, Kashmir, and West Pakistan-the whole of ancient Āryāvarta. Bengal, traditionally a stronghold of tantric worship, is unre- presented in early records, but Kāņhapāda, in his famous Old Bengali songs, calls himself a Kāpālika.165 Nowadays Kāpālikas are still rumored to inhabit the jungles of northern Bengal and

163R.C. Majumdar (ed.). The Struggle for Empire, p. 64. 164See below, pp. 63-71. 165See below, pp. 69-71.

KĀPĀLIKA SOURCES 53

parts of Assam. These rumors are undoubtedly little more than old wives' tales, but some Vämācāra ascetics do survive in this region. Although the evidence is very scanty, it appears likely that the Kāpālikas originated in South India or the Deccan. This is not surprising since the region south of the Vindhyas was dominated by Saivism from very early times. The earliest epigraphical reference to tantric worship occurs in Viśvavarman's A.D. 423-424 record from Gangdhar in south-eastern Rajasthan near Ujjain.166 Further south, in Tamil country, the early inhabitants worshipped the god Murugan with rites which might be called tantric in character.167 Murugan was later identified with Siva's warlike son Skanda. The precise date of the foundation of the Kāpālika order is impossible to establish. It is unlikely however, that these ascetics existed more than a century or two before the time of Mahendra, the author of the Mattavilasa. This period, the fifth to the sixth century A.D., also marks the time of the first development of tantric literature in the subcontinent. The Kāpālikas appear to have virtually died out by about the fourteenth century. The sect was perhaps absorbed by other Saivite tantric orders such as the Kānphațās and the Aghorīs. 168 Kāpālika characters have important roles in only four of the dramas-Mattavilāsa, Mālatī-Mādhava, Caņdakauśika, and Pra- bodhacandrodaya. The Kāpālika Satyasoma in Mahendra's Mat- tavilāsa is a wholly comic creation, reminiscent in many respects of the Kāpālika hedonist Unmatta-Bhairava. The Kāpālikas in Mālarī-Mādhava and Candakauśika, on the other hand, are nefarious rogues similar to Krakaca and Ugra-Bhairava. In the Prabodhacandrodaya the Kāpālika Somasiddhānta displays both comic and horrific traits. To some extent dramatic requirements, both practical and theoretical, have influenced the differing treatments of these ascetics. The classical theory of eight primary sentiments (rasas), which correspond to the eight basic emotions (bhāvas), encouraged

166Ed. and trans. J.F. Fleet, CI/, III, 72-79. 167See A.L. Basham, The Wonder That Was 'India, p. 314. See also J.M.N. Pillai's translation of the famous early Tamil poem 'Tirumuruganarrupadai' in J.M.S. Pillai's Two Thousand Years of Tamil Literature, pp. 55-81. 168For the Aghoris, see H.W. Barrow, 'On Aghoris and Aghorapanthis.' Journal of the Anthropological Society of Bombay, III (1893), 197-251 and also W. Crooke, 'Aghori,' ERE, I, 210-13.

54 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Sanskrit dramatists to imbue each act and character with a specific sentiment.169 In skilful hands this technique could achieve striking ritualistic effects, but it also militated against any form of realistic expression. As a result, the plots of many dramas are recapitulations and elaborations of popular legends and myths, and the characters are representatives of ideal types and sentiments, not people. The influence of the rasa theory is particularly noticeable in Mālati- Madhava. The act in which the Kapalikas appear is meant to evoke the sentiments of terror (bhayānaka, based on the emotion of terror, bhaya) and horror (bībhatsa, based on disgust, juguptsā), and these sentiments are embodied in them. The Mattavilāsa is a one act farce (prahasana) in which the comic sentiment (hāsya) naturally predominates. Satyasoma, a drunken Kapālin or Kāpälika and his equally tipsy wench, Deva- somā, engage in some classic slapstick and clever banter with a Buddhist friar, a Pāśupata monk, and a raving madman. As they enter the stage, Devasoma finds that she is too drunk to stand upright and calls for Satyasoma's assistance. Equally drunk, he falls as he lifts her up. In his befuzzlement Satyasoma calls Deva- somā Somadevā and is accused by her of having another mistress. He offers to forswear liquor to atone for his mistake, but she protests : 'O, master! Don't for my sake ruin your holy life [tapas] by breaking your vow.' He joyfully embraces her and exlaims : Dhrrna dhrrna! Reverence to Śiva! My dear- Ho, don a right jolly and quaint attire, Drink brandy [surā] and gaze in your wenches' eyes : Long life to our Lord of the Trident, who found That the road to salvation this way lies!170

Satyasoma next makes a witty attack on the asceticism of the Jains, and the two proceed to a Kañci liquor shop. He elaborately com- pares the shop with a sacrifice-yard where 'the brandy is the Soma, the tipplers are the priests' and 'the keeper of the brandy shop is the master of the sacrifice' (shades of Unmatta-Bhairava's grandfather!). The two Kapālikas are offered liquid alms, but Satyasoma discovers that his skull bowl is missing. After consi-

169See Keith, pp. 314-26. 170Trans. Barnett, BSOS, V, 703, Ed. T.G. Sāstrī, vs. 7. Dhrrna is some sort of religious exclamation, probably one invented by Mahendra. The translations which follow are all by Barnett.

KĀPĀLIKA SOURCES 55

dering the problem for a moment, he decides to follow the 'law of necessity' (āpad-dharma) and takes the gift in a cow's horn. Without the skull bowl, however, he fears he will lose the title of Kapālin (Skull-bearer). Even worse, the skull had some nice roast meat in it. As they set off in search of the skull, a Buddhist friar passes by on his way to the King's Monastery with a full alms bowl hidden under his robe. The friar's favorite pastime is looking for an 'uncorrupted original text' wherein the Buddha sanctions 'possession of women and use of strong drink.' Satyasoma and Devasoma see that he is hiding something and accuse him of taking their skull. The Buddha, Satyasoma taunts, is superior even to Kharapata, the author of the Thieves' Hand-book, for :

Your Buddha, while the Brahmans' eyes were closed, Filled up his granary by filching notions From Mahābhāratam and from Vedāntas.171

Devasomā offers her master a drink and he passes the cow's horn to the friar. Even though this fellow has wronged us, Satyasoma declares, nonetheless 'our doctrine lays chief weight on sharing our goods.' Visibly licking his chops, the friar is forced to refuse because he is afraid someone might see. The argument becomes more and more heated and Satyasoma finally threatens to make the friar's own head into an alms bowl. They begin to fight but are stopped by Babhrukalpa, the Pasupata monk, who agrees to act as mediator. In reality, however, he is more interested in win- ning Devasoma than in settling the dispute. The friar is at last forced to show the bowl which he had hidden in his robe. Satyasoma and Devasoma refuse to admit that it is not their skull and claim that the friar merely changed its color and shape. Babhrukalpa suggests they take the matter to court, but before they can do so the madman enters carrying the skull bowl which he had taken from 'a most respectable dog belonging to a Candāla.' He offers the skull to Babhrukalpa, who rejects it, but refuses to give it to its rightful owner, Satyasoma. They all attempt to trick the madman into giving it up, but he is adamant. Finally, when the friar calls him a madman, he tells Satyasoma to 'take this skull and show me the madman.' Satyasoma obligingly misdirects him and everyone parts the best of friends.

171Trans. Barnett, BSOS, V, 708 (vs. 12).

56 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Bhavabhūti's Mālatī-Mādhava is a love story, and the erotic sentiment (śrngāra-rasa) consequently predominates. For the purpose of dramatic contrast, and also to display his own virtuosity. however, the author imbues several acts with differing sentiments.172 Much of the plot seems to have been borrowed from the story of Madirāvatī in the Kathāsaritsāgara.173 The hero and heroine of the drama. Mādhava and Mālatī. are children of the ministers of the kings of Vidarbha and Padmāvati respectively. Although both parents want to see Malati and Mädhava wed to each other, a close companion of the king of Padmāvatī also desires the lovely heroine. The Buddhist nun Kāmandakī, an old friend of the two ministers, arranges for Mālatī and Mädhava to meet and fall in love and plans to get them married in secret. Meanwhile, the king summons Mälati to the palace for her marriage with his companion. In desperation Mädhava resolves to offer human flesh to the ghouls of the cre- mation ground in exchange for a boon. Act five opens with the entrance, by an aerial path, of Kapālakuņdalā, the female disciple of the Kāpālika Aghoraghanta. She offers homage to Śiva, Lord of Sakti, and describes her flight to the cremation ground :

The speed of my flight through the sky endows me with a great and charming tumultuousness. Shrill small bells jangle as they strike against the garland of skulls swinging to and fro about my neck. My pile of matted locks, though fastened by firm knots, streams out in every direction. The bell on my khatvänga staff seems to ring out with a continuous piercing scream as it whirls round and round. The wind whistling through the hollows of the row of bare skulls constantly jingles the small bells and causes my banners to flap about.174

Looking at the place around her, she says :

This is the temple of Karala. I can tell that the nearby enclosure of the great cremation area is in front of me by the smoke from the funeral pyres which smells like the

172See the introduction to Devadhar and Suru's edition of the play, pp. 35-36. 173See ibid., pp. 14-20. 174Act V, vss. 3-4 (my translation).

KĀPĀLIKA SOURCES 57

frying of garlic smeared with old nimba oil. My teacher Aghoraghanta has completed the performance of incan- tations (mantra-sādhana) and has commanded me today to bring together here all the necessities of worship (pūjā).175

Aghoraghanta has told her, she says, that today he must offer to Karālā the previously promised 'woman-gem' who dwells in this very city. Kapālakuņdalā then notices the forlorn Mādhava wandering about the cremation ground. In his left hand he holds a 'glistening chunk of human flesh dripping with clotted blood.' As the Kāpālika woman exits Mādhava enters. He laments his separation from Mälatī and offers the flesh for sale to the fiendish loeal residents. His lengthy description of their loathsome activities serves as Bhavabhūti'a vehicle for expressing the sentiment of horror (bībhatsa). As Mādhava passes near the temple of Karālā, he hears a voice crying out in distress and goes to investigate. Kapālakuņdalā and Aghoraghanta then enter with Mālatī, who wears the marks of a sacrificial victim. The Kāpālikas offer obeisance to the goddess Cāmunda and describe her fearful dance as they themselves dance about the stage. Malati's last wish- that Mädhava should remember her even in death-wins the sympathy of the cruel Kapālakuņdalā, but Aghoraghaņța remains pitiless. Raising his weapon, he calls upon Cāmunda to accept his offering. In the nick of time Mädhava rushes forward to save Mälatī. As the two men prepare to fight, soldiers are heard approa- ching the temple looking for her. This ends the fifth act called 'Description of the Cremation Ground.' In act six Kapālakuņdalā tells how Mādhava has killed her guru and swears revenge. Much later, when Malatī and Mädhava are again briefly separated as they are about to be secretly married, Kapālakuņdalā has her chance. She captures the heroine and flies off to Sriparvata with her. Act nine opens with the entrance of the yoginī Saudāminī, a former pupil of the go-between Kāmandakī. The yoginī, who has just flown up from Sriparvata, finds the de- separate Mädhava and tells him that she has intercepted Kapāla- kuņdalā and rescued his precious Mālati. The lovers are eventually reunited and all ends happily. The Candakauśika (Angry Kauśika) by Kşemīśvara is an adaptation of the puranic myth about King Hariscandra and the

175Act V, after vs. 4 (my translation).

58 THE KĀPĀLIKAS AND KĀLĀMUKHAS

irascible sage Viśvāmitra Kausika. 176 One day the king accidentally interrupted the sage's meditation. As reparation he offered his whole kingdom, but the sage was still not satisfied and demanded a final fee ( dakşiņā) as well. The king set out for Varanasi to earn it. In this city he resolved to sell himself into slavery. His virtuous wife, who had followed him to the market, rushed forward ahead of him and sold herself as a domestic slave to a Brahman teacher. Kauśika was not placated by her self-sacrifice, however, and the king in desperation declared that he would sell himself even to a Candāla. The god Dharma immediately appeared in the disguise of a Candäla and purchased the king as a keeper. Act four, like the fifth act of Mālatī-Mādhava, takes place in the cremation ground. Hariścandra's duty was to take the blankets from the dead for his new master. Dharma entered, this time dis- guised as a Kāpālika, and said :

Here am I, sir- Subsisting on alms given without asking for them and calmed by control over the five senses, I have crossed the great cremation ground of transmigratory existence (samsāra) and now roam this disgusting cremation ground. (Reflecting). It is quite suitable that divine Rudra per- formed the Mahävrata. Supreme indeed is this excellence of (those who) roam at will. But- Being exclusively devoted to alms alone, penance alone, and rites alone-all this is easy to obtain. (Being intent upon) the Self alone, however, (is a state) difficult to obtain.177

The king greeted the ascetic: 'Welcome to the performer of the Mahävrata who has undertaken a vow of lifelong chastity (nais- țhika).' The Kāpālika held several magical powers : control over a Vetala and a thunderbolt (vajra); possession of magical pills, ointments and foot salve; command over Daitya women; and knowledge of the elixer of life (rasāyana) and alchemy (dhātu- väda).178 He requested the king to guard these from interference

176Ed. and trans. S.D. Gupta. 177Act IV, vss. 26-27 (my translation). 178Act IV, vs. 31. The purpose of these items is not altogether clear. According to one commentator, the Vetala (a corpse animated by a demon spirit) serves as a

KĀPĀLIKA SOURCES 59

(vighnas). The king agreed to do so as long as there was no conflict with the aims of his master, the Candala. While the king warded off the vighnas, the Käpālika left in search of a great treasure of magical quicksilver (siddharasa) located somewhere nearby. The female Sciences (Vidyäs) then appeared and offered their services to the king, but he told them to wait upon the revered Viśvāmitra Kauśika instead. Meanwhile the Käpālika returned with a Vetāla who carried the treasure of magical quicksilver on his shoulder. This treasure, the Kāpālika claimed, could bestow immortality :

Driving away death through its use and at once attaining the path to the immortal world, the Perfected Ones (Siddhas) . enjoy themselves on the peaks of Meru, where the wishing tree (kalpa-druma) bears clusters of blossoms.179

He offered it to the king who refused to accept it for himself since this would be inconsistent with his condition of slavery. He requested the Kāpālika to give the treasure to the Candāla. Hariścandra was given one more great test of character. His wife entered the cremation ground bearing the body of their son, and the king was forced to demand the funeral blanket as his master's due. After he had snatched it away, flowers fell from the sky and the gods sang his praises. The child revived and was crowned king of his father's empire. Hariścandra and his wife ascended to heaven amidst great rejoicing. The Prabodhacandrodaya of Krsnamiśra180 is an allegorical nāțaka dedicated to the defense of advaita Vaișņavism. Most of the characters are personifications of abstract qualities such as Discrimination (viveka), Confusion (moha), Falsity (dambha), and Faith (śraddhā). The third act introduces four heretical sectarians who are the friends and auxiliaries of Passion (mahā- moha) : a Materialist (Cārvāka), a Jain (Digambara or Kșapaņaka) a Buddhist monk (Bhiksu), and a Kāpālika called Somasiddhānta. Two virtuous maidens, Tranquility (śānti) and Compassion (karuņā), enter in search of Tranquility's mother, Faith. They first meet the Jain who calls upon Faith in the form of a female

slave, the thunderbolt can be directed at will, the foot salve bestows power to walk on water or fly. We are not certain what the pills and ointment (guțikāñjana) are supposed to accomplish. 179Act IV, vs. 34 (my translation). 180Ed. V.L. Paņsikar.

60 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Digambara. Tranquility cannot accept this as her mother. Next the Buddhist introduces his own version of Faith, but this is also unacceptable to Tranquility. Somasiddhänta then enters extolling his own virtues (act III, vs. 12) :

My charming ornaments are made from garlands of human bones. I dwell in the cremation ground and eat my food from a human skull. I view the world as alternately (or mutually) separate and not separate from God (Iśvara) through eyes that are made clear by the ointment of Yoga.

The Jain Ksapaņaka, curious to hear about the Kāpālika vow (vrata), asks Somasiddhānta to explain his conception of dharma and mokșa. Somasiddhānta eagerly complies (act III, vs. 13) :

O Kșapaņaka, you should certainly consider our dharma. We offer oblations of human flesh mixed with brains, entrails and marrow. We conclude our fast by drinking liquor (surā) from the skull of a Brahman (or Brahmã). At that time the god Mahäbhairava should be worshipped with offerings of awe-inspiring human sacrifices from whose severed throats blood flows in torrents.

When the Ksapanaka fiercely repudiates this grim dharma, Soma- siddhānta castigates him in return (act III, vs. 14) :

Ah, Evil one, outcast among heretics, you whose bald head has a single tuft of hair on the top, you whose hair is pulled out (at the roots)! So, the divine Lord of Bhavani, He who creates, preserves and destroys the fourteen worlds, He the greatness of whose doctrine is revealed in the Vedānta (or Upanișads), is a deceiver! I control the gods headed by Hari, Hara and the Eldest of gods (Brahma). I can even halt the progress of the stars travelling in the sky. Know that I can submerge this earth with its mountains and towns under water and then drink up all that water again in an instant.

The Kşapaņaka again condemns the Kāpālika dharma, and Somasiddhänta threatens to gladden the Wife of Bharga (Durgā) and her troop of demons with the blood from his severed neck.

KĀPĀLIKA SOURCES 61

With the help of the Buddhist Bhiksu, the Ksapanaka succeeds in calming his adversary and asks him about the Kāpālika conception of mokşa. Somasiddhānta replies that mokșa is a condition of sensual bliss and is achieved by the union of the worshipper and his wife, who are the earthly counterparts of Siva and Śakti.181 When the Jain and the Buddhist again declare his doctrine to be false, Somasiddhänta summons his own Faith in the form of a beautiful Kāpālinī. At the command of her master she embraces first the Buddhist and then the Jain. The resistance of both is soon shipwrecked on the shore of lust. They plead with Somasiddhānta to initiate them into the most excellent doctrine of Mahäbhairava. He orders them to sit and takes up a vessel of liquor. He drinks from it and offers the remainder to his new disciples (act III, vs. 20) :

Drink this purifying nectar. It is the remedy prescribed by Bhairava against (transmigratory) existence. It is the means of cutting away the bonds of creaturehood (paśu-pāśa).

They at first refuse this polluted and improper drink, but when Faith takes a sip they can contain their eagerness no longer and together imbibe the wine, which is made 'fragrant with the liquor from the mouth of the Kapālini.' The two are soon drunk. Pleased with his work, the Kapalika says to his Faith : 'Love, we have obtained a pair of slaves purchased without capital. Let us now dance.' As they all dance about, he extols his doctrine in which the eight great powers (mahāsiddhis) are won without having to abandon the objects of the senses.182 The Ksapanaka then praises his new 'king of teachers' and kulācārya.183 The Bhiksu sees that the Jain is drunk and asks Somasiddhānta to sober him up. Soma- siddhänta does this by giving the Jain some half-chewed betel nut. The three heretics then draw up a plan to capture Faith for their king, Passion, but they soon discover that she has been joined by Vișnu-devotion and Dharma and has entered the ranks of their enemy, the good king Discrimination. Tranquility and Compassion are overjoyed at this news and set off again in search of Faith. Kāpālikas have relatively minor roles in the other plays we have noted. Śankhadhara's Lațumelaka184 is a one act farce (prahasana),

181Act II, vs. 16. See also below, pp. 90-92. 182Act III, vs. 22. 183The title kulācarya again shows the close relation between the Kaula and Kāpālika faiths. 184Ed. Durgāprasād.

62 THE KĀPĀLIKAS AND KĀLĀMUKHAS

like the Mattavilāsa. The Kāpālika in it is called Ajñānarāśi (Ignorance-heap) and has an intelligence to match. He spends his time arguing with a Digambara monk named Jatāsura. In Rāma- candra's Kaumudīmitrānanda185 a Kāpālika offers oblations of human intestines in a fire and revives a corpse. One of the heroes of the play causes the revived corpse to strike down the Kāpālika. The same author's Nalavilāsa186 features two Kāpālikas-Lam- bodara (Hanging-belly) alias Koşthaka (Stomach) alias Bhasmaka (Ashes) and his teacher Ghoraghona (Horrible-snout) alias Meşamukha (Sheep-face). Both are spies for Cițrasena, a Kalacuri- Cedi king. The spies are devious but rather amusing. In Yasahpāla's Moharājaparājaya187 five heretics-a Kaula, a Rahamāņa (Muslim), a Ghatacațaka, a Nihilist (Nāstika), and a Kāpālika-each give a one verse summary of their faith. The Kāpālika says : 'It has been declared by Narakāpālin (Human-skull-bearer, = Śiva) that who- soever always eats human meat from the skull of an excellent man obtains the place of Siva (Śivasthāna).'188 Between the rather ghoulish Kāpālikas of some of these works and the bibulous Satyasoma of the Mattavilsa there is a wide gulf, but this need not imply that either description is completely false. Tantric religion contains an amalgam of hedonistic and sadistic elements. The playwrights have simply emphasised one or other of these two elements in accordance with their artistic purposes and religious prejudices. Since hedonism lends itself easily to comic treatment, farces such as Mattavilāsa and Lațamelaka feature Kāpālika sybarites. Those authors whose aim is to horrify lay stress on the more sinister aspects of the cult. One work, the Pra- bodhacandrodaya, includes both elements. Although the account in this play is still highly tendentious and distorted, it is in many respects the most informative. We must postpone a fuller discussion of Kāpālika religion until after we have reviewed the descriptions of these ascetics in religious and narrative literature.

Miscelláneous Later Sources Stories about Kāpālikas occur frequently in katha collections such as Somadeva's Kathāsaritsāgara (c. 1063-81). In this work

185We could noi locate a copy of this work, It is summarised by Handiqui, p. 358 and by Keith, p. 259. 186Ed. G.K. Shrigondekar and L.B. Gandhi. 187Ed. Chaturavijayaji. 188Act IV, vs. 23. The Ghatacataka cannot be satisfactorily identified:

KĀPĀLIKA SOURCES 63

the story of Madanamañjarī pits a Kāpālika against the illustrious king Vikramāditya.189 Madanamañjarī, the daughter of the king of the Yaksas and wife of Kubera's brother, was amusing herself one day in a garden in Ujjayini. She was seen there by a 'hypocritical Kāpālika.'190 He immediately fell in love with her and retired to a cremation ground to attempt to make her his wife by means of a spell (mantra) and burnt offering. Madanamañjari learned of his plan through her magical power but was helpless against his Yaksa-subduing spells. Drawn by these spells she reached 'the terrible cremation ground which was filled with bones and skulls and frequented by demons.' There she saw the wicked Kāpālika. He had made a fire for oblations and a ritual circle (mandala) in which he worshipped a supine corpse.191 Madanamañjari invoked the protection of king Vikramāditya who immediately appeared accompanied by a Vetāla named Agniśikha. The king ordered the Vetala to 'kill and eat this evil Kāpālika who has carried off another's wife.' The Vetäla entered the corpse which rose up and quickly dispatched the Kāpālika. Another lecherous Kāpalika appears in the story of a young Brahman named Candrasvämin.192 One day this Brahman went to town on an errand. Meanwhile a Kāpālika came to Candras- vämin's house and happened to see his beautiful wife. The lady immediately contracted a fever and died that evening. By the time Candrasvāmin returned, his wife's relations had already placed her body on a funeral pyre. As he approached the blazing pyre he saw the Kapalika. On his shoulder the ascetic carried a 'dancing' khatvānga staff, and in his hands he held a thundering damaruka drum. When he threw ashes on the fire, the lady stood up uninjured. Drawn by his magical power (siddhi), she ran away with him to a cave on the bank of the Ganga. In the cave were two captive maidens. After putting down his khatvānga, the Kāpālika exclaimed to them : 'My vow has attained success (siddhi). I bave now obtained her without whom I could not enjoy you two even though I had obtained you.' The lady's husband Candrasvamin ad followed them there, however, and, seeing his chance, he threw the khatvanga into the

189Ed. Durgāprasād and K.P. Parab, xviii. 2. 1-33 and 209-214. 190khanda-Kāpālika. This is how Tawney renders this strange term. Böhtlinck and Roth (Sanskrit Wörterbuch) translate 'ein Quasi-kāpālika.' 191xviii. 2. 15. To a certain extent this ceremony resembles the Mahākāla-hrdaya performed by Bhairavācarya in the Harsa-carita. 192Ibid., xviii. 5. 1-22.

64 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Ganga. Without the magic of his staff, the Kāpālika was powerless. He tried to flee but was killed by one of Candrasvamin's poison arrows. 'Thus,' says Somadeva (xviii. 5. 16), 'heretics, who make a mockery of the Sivgamas for the pleasure of evil accomplishments, fall (into ruin), just as they had already fallen (into sin).' Candras- vämin released the two bewitched maidens and returned home with his wife. Another story from the Kathāsaritsāgara, that of the Brahman gambler Devadatta (v. 3. 196ff.), has as one of its central characters a Mahavratin named Jālapāda. One day Devadatta gambled away all his possessions, even his clothes, and was unable to return home to his father's house. He entered an empty temple where he saw the solitary Mahavratin, whose magic had accomplished many things, muttering mantras. Devadatta greeted him respect- fully and recited his tale of woe. The Mahävratin offered to restore Devadatta's fortunes if the gambler would assist him in becoming a Vidyädhara. The following day the Mahāvratin came and sat under a banyan tree in a corner of the cremation ground. That night he did pūjā, offered rice boiled in milk, and scattered food offerings in the four quarters. He told.his new assistant to perform the same worship every day in the same spot while saying: 'O Vidyutprabhā, you should accept this pūjā.'193 Eventually their efforts were rewarded with success. In Kathāsaritsāgara v. 2. 81 Somadeva mentions a 'Mahāvratin Kapālin' who wears matted hair, smears himself with white ashes, and has a half moon like Siva's drawn on his forehead. In yet another story (iii. 5. 74-77) a group of spies in Varanasi disguise themselves as ascetics who 'observed the Kāpālika vow.' One of them assumes the role of teacher while the others become his disciples. These disciples then go about saying : 'This teacher knows the present, past and future.' And they make sure that any predic- tions their teacher makes, come true. By this infallible method the spy-ascetic quickly wins the notice and confidence of the king. The Jain legend of Prince Brahmadatta is found in Devendra Gani's eleventh century commentary on the Uttarādhyayana194 and in Hemacandra's (1088-1172) Trişaștiśalākapurușacaritra.195 At one point in this story, the prince's friend Varadhanu is forced

193v. 3. 207. Vidyutprabhä was the daughter of a Yakşa king. 194Trans. J.J. Meyer, Hindu Tales, pp. 23ff. 195Trans. H.M. Johnson, V. 335ff.

KĀPĀLIKA SOURCES 65

to disguise himself as a Kāpālika in order to rescue his mother from the Candala quarter of a town. Several other early mediaeval works by Jain authors contain stories about Kāpālika ascetics or at least briefly mention them. In Haribhadra's (c. 750-800) Prakrit Samarāiccakahā, the gambler Maheśvaradatta becomes a Kāpālika and an expert in snakebite charms (gāruda-mantras).196 In the Pārśvanātha-caritra the goddess Kālī priases a Kāpālika who collects skulls for her. When she obtains her 108th skull she is to 'fulfill her purpose.'197 Vinaya- candra's (c. 1300) Mallinātha-caritra tells how Prince Ratnacandra finds a Kāpālika 'eagerly dancing with a sharp sword' in front of a young woman who is tied to a post. The prince rescues her and kills the ascetic.198 The story of King Devapāla in the Kathākosa, a collection of uncertain date, mentions a Kāpālika who carries a bundle of wood on his head. When the queen sees him, she recog- nises him as her husband from a former life. She had worshipped the Jina and become a queen. He had refused to do so and attained the 'miserable condition' of a Kāpālika.199 In Jambhaladatta's version of the Vetālapañcavinśati,200 written sometime before 1500, the ascetic whom King Vikramakesarin agrees to assist is called a Kāpālika. The king is requested to carry an unmutilated dead man from a tree on the bank of a river to the cremation ground where the Kāpälika is to perform a magic rite. 'When you have come,' the Kāpālika tells him, 'then here in a circle [mandala] furnished with the various instruments of worship, when I have washed the corpse and worshipped the gods and muttered a great incantation [mahāmantra], I shall attain magic power [siddhi].'201 The ascetic's actual aim is'to sacrifice the king, but the dead man, really a Vetāla, warns the king and the Kāpālika is killed instead. Many references to these ascetics take the form of poetic meta- phors or similes. Thus Trivikrama-bhatta's (c. 915) Nalacampū, a Jain version of the famous legend, compares the trees of the

196Cited by Handiqui, p. 358. 197ii. 288, cited by M. Bloomfield, 'On False Ascetics and Nuns in Hindu Fiction, JAOS, XLIV (1924), 203. There are several works by this name. We have not been able to locate Bloomfield's source. 198Ed. Hargovinddas and Bechardas, i. 40-62. This is again reminiscent of the episodes in Daņdin's Daśakumāracarita and Bhavabhūti's Mālatī-Mādhava. 199Ed. J.L. Shastri, p. 4. 200Ed. and trans. M.B. Emeneau. The earlier versions of this cycle do not speci- fically call the ascetic a Kāpālika. 201Trans. ibid., p. 11.

66 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Vindhya forest to the khatvānga staffs of Kāpālikas.202 The four- teenth century Muslim poet Abdul Rahmān uses the Kāpālika as a symbol of an absent and wandering husband in his Apabhramśa Samdeśa-rāsaka.203 Kalhaņa's Rājatarangiņī, written between 1150 and 1160, contains several such poetic allusions. During a severe famine in Kashmir the ground is said to have become covered with fragments of skulls and 'to observe, as it were, the custom of skull-carrying ascetics (kāpālika).204 After the burning of the temple of Cakra- dhara (Vişņu) in about A.D. 1125, says Kalhaņa, 'Mankha, a Dāmara from Naunagara, searched the dead bodies like a Kāpālika, and gratified himself with the objects found upon them.'205 Bhandreśvara, a rapacious tax official in the service of Samgrāma- rāja (1003-1028), is unfavorably compared to 'a fear-inspiring Kāpālika, who lives on corpses, [but] gives maintenance to his own people.'206 More interesting is an episode in the Rājataranginī which seems to identify the Pasupatas and the Kāpālikas and to connect them both with Sriparvata. This is the legend of the kings Vikramāditya- Harșa, Pravarasena II and Mātrgupta.207 Pravarasena, son of Toramäna and heir to the throne of Kashmir, went on a lengthy pilgrimage during which the throne fell vacant. King Vikramāditya- Harsa, who was apparently overlord of the region, sent his court poet Mätrgupta to fill the post. Pravarasena, still on pilgrimage, learned of this usurpation and sought to gain the kingship himself. When he arrived at Śrīparvata, 'a saint [Siddha] called Aśvapāda, who appeared in the guise of a Pasupata ascetic, offered him food prepared from roots.'208 Aśvapāda said that the prince had been his attendant in a former life and that on a certain occasion the ascetic had offered his servant a boon. Pravarasena had asked for a kingdom. Siva had then appeared and promised to fulfill this wish in another life. After imparting this information, Aśvapāda disappeared. The prince stayed at Sriparvata and performed penances in order to win the favor of Siva. Eventually Siva granted

202Ed. Durgāprasād and Sivadatta, p. 165. 203Ed. and trans. J.V. Muni, ii. 86 and iii. 185. 204Trans. M.A. Stein, Vol. II, viii. 1211. Ed. M.A. Stein (same verse numbers). 205Trans. Stein, Vol. II, viii. 995. 206Trans. Stein, Vol. II, vii. 44. 207Trans. Stein, Vol. I, iii. 125-378. 208Trans. Stein, Vol. I, iii. 267.

KĀPĀLIKA SOURCES 67

the promised boon and Mätrgupta peacefully abdicated in favor of Pravarasena. At the end of Pravarasena's long reign, Aśvapāda ordered his new disciple Jayanta, a Kashmiri Brahman, to take a letter to the king. When Jayanta complained that he was too ex- hausted from travelling to start on the great journey back to Kashmir, Aśvapāda said : 'Then bathe to-day, since I who am of the Kāpālin sect, have touched you who are a Brahman.'209 Aśvapāda then threw him into a pond. When Jayanta opened his eyes he was standing near Pravarasena's palace. The letter was quickly delivered. It instructed the king to 'go and betake yourself to Siva's abode.'210 With a great burst of light the king rose into the heavens. There has been much speculation about the identity of these three kings. It can be safely said only that they lived sometime between the fifth and eighth centuries. The legend about them is apocryphal in any case. From our point of view the important fact is that a.Pasupata ascetic who lives at Śriparvata calls himself a member of the Kāpālin sect. Kalhaņa's apparent identification of the two sects is undoubtedly a mistake, but it is an understandable one since Sriparvata is associated both with the Pāsupatas, through their offshoot the Kalamukhas, and with the Kapalikas. In the time of the three kings, the site was probably controlled by the Kāpālikas. In Kalhaņa's time it was a Kālāmukha center. This might be the source of his confusion. The idea of contact with Kāpālikas causing pollution recurs in several sources. In view of their strange habits, this is not surprising. Kşemendra (c. 1050-75), the Kashmiri polymath, includes a restriction against drinking with Kāpālikas in an attack against the tantric gurus of the Kali-yuga : 'The gurus claim that mukti (is obtained) by drinking (wine) in one vessel with artisans-such as washermen, weavers, leather makers, and Kāpalikas-during cakra-pūjā,211 by having a feast of unhesitating sexual pleasure. and by (generally leading) a festive life.'212 It is not clear why

209Trans. Stein, Vol. I, iii. 369. 210Trans. Stein, Vol. I. iii. 373. 211 During cakra-pūjā tantric adepts are required to partake of the five Ma-sounds - wine, meat. fish, mudrā, and sexual intercourse. 212 Daśāvatāra-carita, ed. Durgāprasād and K.P. Parab. x. 26. In Kșemendra's Deśopadeśa (ed. M.K. Shāstrī, iv. 3), a procuress is said to have the 'form of the great skeleton of the Kāpalika of Death.'A Mahävratin appears with some heretics and rogues in Kşemendra's Narmamālā (ed. M.K. Shāstrī, iii. 15).

68 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Kāpālikas are included in a list of artisans (silpins). Somadeva's Yaśastilaka (A.D. 959) prescribes the following penance for a Jain monk who comes into contact with a Kāpālika : 'When there is contact with a Kāpālika, a menstruating woman, a Candāla, a Śabara, or other (such persons), as a penance one should duly bathe, fast, and mutter a mantra.'213 Devanna Bhatta (c. 1200) quotes a similar passage from the Sattrimsanmata: 'When one touches Bauddhas, Pāśupatas, Jains, Lokāyatikas, Kāpilas, or Brahmans who perform prohibited acts, one should enter the water still dressed. In case of contact with Kāpālikas, restraint of the breath (prāņāyāma) is also prescribed.214 The Uśana- samhitā includes Kāpālikas in a list of heretics with whom food should not be eaten.215 Two fairly late works, the Bārhaspati-sūtra and Guņaratna's fourteenth century commentary on Haribhadra's Saddarsana- samuccaya, stress the hedonistic element of the Kāpālika faith. Guņaratna claims that the Kāpālikas are identical with the Nāstikas or Lokyatikas who enjoy wine, meat and illicit intercourse.216 The Bārhaspati-sūtra distinguishes Kāpālikas from Lokāyatikas but seconds Gunaratna's charge of dissoluteness.217 A fourteenth century Tamil work, the Śivaprakāśam of Umāpati, contains a brief disquisition on seven sects which hold that mukti is the removal of mala (impurity). These include the Pāsupata, the Mahavratin and the Kāpālika.218 Here Mahavratin probably denotes the Kālāmukhas. Another Tamil work, the Tiruvorriyūr Purānam. seems to refer to Mahāvratins in this sense.219 The twelfth century Tamil author Sēkkiļār describes a Mahāvratin ascetic who might be either a Kāpālika or a Kālāmukha. This ascetic is Siva himself in disguise. Three lines of ashes are drawn across his forehead; his head is shaved except for a tuft tied up

213vi. 3, cited by Handiqui, p. 356 (my translation). Elsewhere in this work Soma- deva describes a certain bad minister as one whose 'fame has been spread in the world by religious mendicants, snake-charmers, Kāpālikas, jugglers and consummate thugs' (iii. 183, trans. Handiqui, p. 66). 214 Smrticandrikā, ed. L. Srinivasacharya, II, 310. 215Chap. iv, vss. 23-26, cited by T.V. Mahalingam, 'The Pasupatas in South India,' JIH, XXVII (1949), 46. 216Ed. L. Suali, p. 300. 217Ed. and trans. F.W. Thomas, ii. 6, 9, 13, 18-21. 218Trans. H.R. Hoisington, JAOS, IV (1854), 125-244. 219See V. Raghavan, 'Tiruvorriyūr Inscription of Chaturānana Paņdita: 20th Year of Krishna III,' EI, XXVII, 300.

KĀPĀLIKA SOURCES 69

with a garland of bone beads; he wears kundala earrings; he has a necklace or garland of shining bone beads and a shoulder strap for yogic postures; his sacred thread is a rope of black hair; he is smeared with ashes and carries a sack of them with him; on one wrist a single bead is tied with a string (sūtra); his genitals are covered only by a loincloth; and the five marks (mudrās) of greatness shine on his feet.220 Important references to Kāpālikas occur in three Old Bengali songs (caryāpadas) by the Sahajiyā Buddhist saint Kāņhapāda (Sanskrit, Krşņapāda).221 Kāņha in fact calls himself a Kāpālin although the context makes it probable that he intends this in a symbolic sense. Two of the Kāpalin songs are addressed to the Dombi (Washerwoman) who, in symbolic terms, is the goddess Nairātmyā (Essencelessness) and Buddhist counterpart to the Hindu Kula-kuņdalinī Śakti.222 In the form of a Kāpālin yogin, Kanha becomes the lover or husband of the Dombi :

Outside the city, O Dombi, is thy cottage; thou goest just touching the Brahmins and the shaven-headed (and never reveal [sic] thyself to them). O Dombi, I shall keep company with thee and it is for this purpose that I have become a naked Kapālī without aversions .... Thou art the Dombi and I am the Kāpāli, for thee have I put on a garland of bones. The Dombi destroys the lake and eats up the lotus- stalk. I shall kill thee, Dombi, and take thy life.223 As the earthly Dombi should not be touched by the orthodox, so the divine Dombi is inaccessible to them. She lives outside the 'city' (the world of the senses). Unless she is killed (i.e. controlled) she spoils the lake (the body) and eats the lotus stalk (the bodhicitta or mind of enlightenment).224 The second song expresses similar sentiments :

Of what nature is, O Dombi, thy cleverness ?- the aristocrats are outside thee and the Kapalis are within ... Thou art

220Paraphrase of translation by Rangaswamy in his The Religion and Philosophy of Tēvāram, I, 385. 221Ed. and trans. M. Shahidullah, Les Chants Mystiques, songs no. 10, 11 and 18 (Sāstrī's numbers). Some of Känha's songs are translated into English by S. Dasgupta in his Obscure Religious Cults. 222See Dasgupta, ibid., pp. 96-106. 223Song no. 10, trans. ibid., pp. 103-104. 224See ibid., p. 104.

70 THE KĀPĀLIKAS AND KĀLĀMUKHAS

the Kāma-candāli,-there is no woman more cunning and unfaithful than the Dombi.225

The 'aristocrats' (kulina jana) are the orthodox priests. It is only the Kāpālins who can realise Nairātmya.226 In the third song Känha symbolically explains the essence of the true Kāpālin : `the yogin Kanha has become a Kāpāli, and has entered into the practices of yoga, and he is sporting in the city of his body in a non-dual form.'227 His anklets and bell (ghantā) are the āli and kāli-'the principles of all kinds of duality.'228 His earrings ( kundala ) are the sun and the moon (Upāya and Prajñā, equivalent to Siva and Sakti). The ashes he smears on his body are the ashes of passion (rāga), aversion (deśa, Sanskrit dveșa), and error (moha). His pearl necklace is supreme salvation (parama mokha). The song ends with a paradoxical verse typical of tantric 'intentional language' (sandhā-bhāșā) : 'Ayant tué la belle-mère (=le souffle) le beau-frère et la belle-soeur [=the senses] dans la maison et ayant tué la mère (=l'illusion) Kāņha est devenu porteur de crânes [kabāli, = kapālin].'229 In these songs the Kāpälin symbolizes the perfected yogin precisely because on a mundane level he is the most debased of ascetics. The verse about his murder of his mother and various relatives suggests that Känha may also have been aware of the connection between the Kapalin and the Brahmahatya vow of the law books.230 But just as one must not suppose that Kānha actually killed his mother and relatives, it is unlikely that he actually became a Kāpālika. The Kapālin, like the Dombī, is a symbolic represen- tative of the mystical doctrine of the identity of opposites. It is just possible, however, that Kanha gave this doctrine concrete embodiment and assumed the dress and habits of a Kāpālika.231 The connection between the Kapalika vow and the penance of

225Song no. 18, trans. ibid., pp. 104-105. 226Ibid. 227Song no. 11, trans. ibid., p. 90. Dasgupta notes that here the Sanskrit commen- tator derives the word Kāpālika as follows : kam mahā-sukham pālayati'ti kāpālikaḥ, 'He who nurses Ka which means Mahā-sukha is a Kāpālika.' 228Ibid., p. 58. Dasgupta paraphrases the song on pp. 57-58. 229Trans. Shahidullah, p. 118. 230See below, pp. 73-82. 231 Kaņhapāda is often identified with the Nāth Siddha Kānupa. If correct, this identification would help to bridge the gap between Känha's tantric Buddhism and the tantric Saivism of the Kāpālikas. Kānha and Kānu are both venacular variants

KĀPĀLIKA SOURCES 71

the Brahmahan is itself an example of the operation of this doc- trine.232 In cases like this the boundary between symbol and reality often becomes difficult to define. From an historical point of view, Känha's mention of the Kāpālikas is important since it is the earliest reference to these ascetics in Bengal. Känha's date is uncertain but it seems probable that he and the other authors of the Sahajiyā dohās and caryāpadas flourished during the eighth to twelfth centuries under the Palas.233 If Kānha is the same as the Siddha Kānu-pā, as seems quite possible, he must. have lived sometime after the tenth century.234

of the Sanskrit Krsna. Kanu-pa's guru was named Jalandharī-pā. In song no. 36 Kāņha mentions a Jālandharī-pa as his, or at least a former, teacher. Dasgupta is disinclined to accept these identifications (Obscure ... , pp. 392-94), but to us this seems rather stubborn-minded. 232See below, pp. 76-77. 233 Dasgupta, Obscure ... , p. 9. 234Ibid., pp. 386-93.

CHAPTER III

KĀPĀLIKA CULT AND DOCTRINE

The Mahāvrata One of the most puzzling problems about the Kāpālikas, and to some extent the Kālāmukhas as well, is their association with penance or vow called the Mahāvrata (Great Vow). Since there is reason to believe that the Kāpālika and Kālāmukha Mahāvratas wore different vows, we will discuss them separately beginning with the former. A large number of sources connect the Kāpālikas with the Mahāvrata. Jagaddhara, a commentator on Mālatī-Mādhava, explains Kāpālika-vrata or Kapāla-vrata as Mahāvrata.1 Similarly, Candapāla, a commentator on Trivikrama-bhatța's Nalacampū (p. 164), equates Kāpālikas and Mahāvratikas. A Saivite ascetic in Somadeva's Kathāsaritsāgara v. 2. 81 is called a Kapālin Mahā- vratin. Kşīrasvāmin (eleventh century), in his commentary on the Amarakośa, lists together Kāpālin, Mahāvratin, Somasiddhāntin, and Tāntrika.2 A Mahāvrata-Kāpālika named Mahāvrata, who follows the 'heretical Mahāvrata-siddhānta,' appears in Gokulanā- tha's Amrtodaya (c. 1700).3 As we have seen,4 some of the Purānas and a few other sources contain lists of sects in which the Kāpālikas (or Kālāmukhas) are replaced by Mahāvrata-dharas or Mahā- vratins. In one or two of these sources, however, Mahāvratins are listed as distinct from both Kāpālikas and Kālāmukhas. In two plays, Mattavilāsa and Caņdakauśika, a Kāpālika character refers to Siva's performance of this vow.5 In the latter work the Kāpālika himself is called a Mahāvrata-carin as well.6 We have already discussed the seventh century Cālukya grant from Nasik District. which registers a donation to the Mahāvratin priests of a Kāpāleś-

1Ed. R.G. Bhandarkar, text p. 33. 2Quoted in Śrīharșa, p. 640. 3Ed. Sivadatta and K.P. Parab, pp. 41-42. 4See above, pp. 7-11 5Mahendravarman, Mattavilāsa, ed. T.G. Śāstrī, vs. 17 and Kșemīśvara. Caņda- kauśika, Act IV, vss. 26-27. 6Act IV, after vs. 29.

74 THE KĀPĀLIKAS AND KĀLĀMUKHAS

vara temple, and the eleventh century grant from Baroda District, which compares its priestly donee to Kapālin Śamkara. What was this Mahävrata? The best known rite by this name takes place during the last day but one in a sattra and is described in the Jaiminīya Brähmana and a few other early works. It is asso- ciated with the mysterious brotherhood, the Vrātyas, whom Hauer saw as precursors of the yogins,7 and it incorporates a number of features which seem appropriate for a Kāpālika ceremony, such as ritual reviling, obscene dialogue and sexual intercourse.8 There is little likelihood, however, that this ritual would have been resurrected several hundred years after it had to all intents and purposes died out and after its original religious and social context had disappeared. Furthermore, there is another Mahāvrata which may be identified with some certainty as the Great Vow of the Käpälikas. This is the chief penance prescribed for the removal of the sin of (accidently) killing a Brāhmaņa. The rules for this penance, with several variations, are found in most of the major law books, but it is called the Mahāvrata in only one of them, the Visnu-smrti. This work says :

1. Let a man make a hut of leaves in a forest and dwell in it; 2. And let him bathe (and perform his prayers) three times a day; 3. And let him collect alms, going from one village to another, and proclaiming his own deed; 4. And let him sleep upon grass : 5. This is called a [the] Mahāvrata (great observance). 6. He who has killed a Brāhmana (unintentionally) must perform it for twelve years.

2. He who is performing any of those penances must carry (on his stick) the skull of the person slain, like a flag.9

7J.W. Hauer, Der Vrätya. J.C. Heesterman has recently taken issue with the views of Hauer and others in his article 'Vratya and Sacrifice' (IIJ, VI [1962-63], 1-37). Heesterman sees them as 'authentic Vedic Aryas' whose rituals 'are the crude predecessors of the śrauta ritual' (p. 36). 8See Hauer, pp. 246ff .; Eliade, pp. 103-105; and Kane, HDS, II, 1243-45. 9Trans. J. Jolly, 1. 1-6, 15. Ed. V. Krishnamacharya. Compare Manu-smrti, trans. G. Bühler, xi. 73; Yājñavalkya-smrti, ed. N.R. Acharya, iii. 243; Gautama' Dharmaśāstra, trans. G. Bühler, xxii. 4-6; Baudhāyana Dharmaśāstra, trans. G. Bühler, ii. 1. 2-3; Āpastambīya Dharmasūtra, trans. G. Bühler, i. 24. 11-20; ibid., .i. 28. 21 to i. 29. I; and Kūrma Purāņa, ii. 30.

KĀPĀLIKA CULT AND DOCTRINE 75

We have quoted above the version of this penance prescribed in the Yājnavalkya-smrti iii. 243.10 There the performer is called a kapālin, but only in the sense of 'one who carries a skull.' Yajñavalk- ya implies that the penitent should carry a skull in his hand as well as on his staff. The commentators disagree about whether or not he should use the skull in his hand as a begging bowl.11 In the Āpastambīya Dharmasūtra, however, a person who has killed a learned Brāhmaņa (Bhrūnahan) is required to 'take a human skull for his drinking-vessel.'12 One who kills an ordinary Brahmana, on the other hand, is merely instructed to carry a shallow metal or clay vessel.13 Several works require the penitent to carry a skull on his staff, and this skull is generally identified as the skull of the person slain.14 Some works also require the carrying of a khat- vānga,15 the staff most often associated with the Kāpālikas. In his comments on Āpastambīya Dharmasūtra i. 29. 1, Haradatta (c. twelfth century) in fact says : 'the word khatvānga is well known in the Kāpālika-tantra.'16 Literally, khatvānga means 'limb of a bedstead,' apparently on account of its shape. Vijñāneśvara's Mitākșara on Yājnavalkya iii. 243 describes it as a 'banner made of a skull mounted on a stick (danda).' A few of the law books specify the clothes the penitent must wear. Apastamba says that a Bhrünahan 'shall put on the skin of a dog or of an ass, with the hair turned outside.'17 Baudhayana (ii. 1. 3) prescribes the hide of an ass alone. For an ordinary Brah- mahan Āpastamba (i. 24. 11) requires a plain hempen loincloth reaching from the navel to the knees. Because he is polluted by his crime, the sinner must live in a hut in the forest and avoid entering a village except to beg. According

10See above, p. 13. 11See Kane, HDS, IV, 89. Kane discusses in some detail this and other penances for the crime (ibid., pp. 87-96). 12Trans. Bühler, i. 28. 21. See also Baudhāyana ii. 1. 3 and Gautama xxii. 4. 13 Āpastambîya i. 24. 14. 14Manu xi. 73 and Āpastambīya i. 24. 11. Vijñāneśvara's Mitākșara (ed. N.R. Acharya) on Yājñavalkya iii. 243 quotes Śātātapa as saying that the guilty person should visit the tirthas taking with him the skull of the Brähmana he has killed. Vijñāneśvara adds that if the head of the person slain is not available, the head of another Brāhmana should be used. 15 Āpastambīya i. 29. 1 (penance for a Bhrūņahan); Gautama xxii. 4; and Baudhāyana ii. 1.3. 16Ed. M. Śāstrī and K. Rangāchārya. 17Trans. Bühler, i. 28. 21.

76 THE KĀPĀLIKAS AND KĀLĀMUKHAS

to Baudhāyana (ii. 1. 3) a Bhrūņahan should build his hut in a burial ground. Apastamba (i. 29. 1) suggests that he should live in an empty house or under a tree. Apart from begging, the Brahmahan's daily duties are not much discussed. The Vișnu-smrti 1. 2 instructs him to perform the usual trisamdhya ablutions. Āpastamba (i. 24. 11 and 18) requires him to tend cows and restrain his speech. Gautama (xxii. 4) says he should remain chaste. The Brahmahan must obtain all his food by begging. Āpastamba specifies that he should visit only seven houses on one day. At each he should cry : 'Who will give to an Abhisasta [guilty one]?'18 A Bhrūņahan. says Apastamba. should cry: Who (gives) alms for a Bhrunahan?' According to Baudhayana the Bhrūnahan should also follow the seven house rule.19 It is generally agreed that the penance for both the Brahmahan and the Bhrunahan should be performed for twelve years, but Apastamba (i.29.1) says that the Bhrūņahan must maintain the vow until death. Several law books list additional penances for the crime of killing a Brähmana.20 The commentators assign these different penances according to the education and wisdom of the victim and the presence or absence of intention in the slayer. Some of these alternative penances end in almost certain death and others merely require the spending of large amounts of money for Vedic sacrifices. One of them, the chief penance prescribed for a Bhrūnahan in the Vāsistha Dharmasūtra, seems to have tantric overtones. The guilty person is instructed to build a fire and offer in it eight oblations cut from his own body : hair, skin, blood, flesh, sinews, fat. bones, and marrow. The successive oblations are offered to Death with the words 'I offer my hair to Death, I feed Death with my hair' and so forth.21 At the least, this penance requires self-mutilation, and excessive diligence could easily cause death. The rite is reminiscent both of the grisly oblations that the Kāpālika in the Prabodhacandrodaya claims to offer to Bhairava and of the sale of flesh cut from their own bodies by the Mahāvratikas of the Candamārī temple in Somadeva's Yaśastilaka.22 The Mahävrata penance of the Visnu-smrti and other law books

18Trans. Bühler, i. 24. 15. 19Trans. Bühler, ii. 1. 3. 20See Kane, HDS, IV, 87-96. 21Trans. J. Jolly, xx. 26. 22See above, pp. 17 and 60.

KĀPĀLIKA CULT AND DOCTRINE 77

bears an unmistakable resemblance to the observance of the Kāpālikas. These ascetics lived in the forest, wore loincloths or animal skins, carried a khatvānga and a skull bowl, obtained their food by begging, and polluted those with whom they came into contact. Given the pervasive tantric motif of the identity or con- junction of opposites,23 the relation between the penance of the law books and the vow of the Kāpālikas is not inexplicable. The Kāpālikas, we suggest, adopted this vow precisely because it was the penance for the most heinous of all crimes, the killing of a Brähmaņa. They were at the same time the holiest of all ascetics and the lowest of all criminals. As in the case of the dombi (and the Kāpālin) of Känhapāda's songs, that which is lowest in the realm of appearance becomes a symbol for the highest in the realm of the spirit. Furthermore, if the Kāpālikas were in reality already guiltless, the performance of this penance would result in an unprecedented accumulation of religious merit and hence of magical power (siddhi). The paradoxical identity of Kāpālika saint and Brahmahan sinner finds its divine archetype in the curious myth of the beheading of the god Brahmā by Siva. This also introduces the essential ingredient of Saivism which is lacking in the law book penance. The myth occurs in a number of the major Purānas, but their accounts vary considerably. We will summarise the Matsya Purāna version since it seems to preserve most of the basic features of the story.24 One day Siva is asked by Pārvatī why he never leaves the Avimukta kşetra in Varanasi, where the Kapālamocana (Setting Free of the Skull) tirtha is located. Siva replies (clxxxiii. 84-87) :

Formerly, O Varārohā, there was an excellent fifth head of Brahmā. It arose, O Suśronī, having the same lustre as gold. When that flaming fifth head of the great-souled one was produced, O Devi, he said (to me) : 'I know (the circumstances of) your birth.' Then, filled with anger and my eyes inflamed, I cut off his (fifth) head with the tip of the nail of my left thumb. Brahma (then) said : 'When you cut off the head of me who is guiltless, you will become a Kapālin endowed with a curse. Having become burdened

23See above, pp. 70-71. 24Ānandāśrama edition, clxxxiii. 83-108.

78 THE KĀPĀLIKAS AND KĀLĀMUKHAS

with the (sin of) Brahmahatyā you should visit the tirthas on earth.'

By cutting off the head of Brahma, Siva himself becomes guilty of the crime of killing a Brähmana and must undergo the prescribed penance. The head magically attaches itself to his body, and he travels with it to the Himalayas to ask Nārāyana (Vișnu) for alms. Nārāyana lacerates his own side with the tip of his nail. A great flood of blood streams out and spreads over fifty yojanas. This great flood flows for a thousand divine years, but it cannot fill the skull. Nārāyana asks Siva about the origin of this amazing skull, and Siva tells him the story of the beheading and its aftermath. Siva is then instructed to go to 'his own place' where the skull 'will establish itself.' Śiva travels to many famous tīrthas but the skull does not 'establish itself' until he visits 'the great resting place Avimukta' and there his curse finally departs. Siva concludes his tale :

Through the grace of Vișnu, O Suśronī, the skull was there broken in thousands (of pieces). As many pieces were produced as riches are obtained in a dream. This sacred field (kşetra) I made the tirtha which removes (the sin of) Brahmahatyā. It is renowned on earth, O Devi, as the Kapalamocana of the gods ... Whoever abandons his body while abiding there will merge with me.

Every ritual has a divine model or archetype, and the penance Śiva performs is the model of the Mahvrata penance for the killing of a Brāhmaņa. The Kāpālika in the Mattavilāsa makes this identification explicit :

By strict observance of this holy course [Mahāvrata] Our Lord whose crest-gem is the crescent moon

25Ibid., 100-101, 104. For summaries of the other puranic versions of this myth, see T.A.G. Rao, Elements of Hindu Iconography, II, Part I, 295-300; M.A.D. Rangaswamy, The Religion and Philosophy of Tēvaram, Book I, pp. 372-76; and S. Das Gupta's introduction to her edition of Kşemīśvara's Caņdakausika, p. Ixx. See also Vāmana Purāņa ii. 17 to iv. 1; Kūrma Purāna ii. 30 and 31; and Canna- Basava Purāņa, trans. G. Wurth, chaps. xviii-xx. Rao identifies Siva's penance with the Bhrünahan vow in the Āpastambīya Dharmasūtra. Another version of the myth is found in Kathāsaritsāgara ii. 13.

KĀPĀLIKA CULT AND DOCTRINE 79

Was freed from guilt that sprang from cutting off The Grandsire's head ... 26

Akhough the myth is religiously prior to the legal prescription, the historical precedence is uncertain. The law books are in general much older than the Puranas, but both classes of works are based on earlier sources which are now lost. The killing of a Brāhmaņa ( Brahmahatyā) is already regarded as the worst of all sins in the Brähmanas,27 but these works do not refer to the expiatory penance. The essential features of the Siva-Brahma myth are found, however, in the story of Räma Räghava and the sage Mahodara from the Śalyaparvan of the Mahābhārata.28 According to this story Rāma once fought and beheaded a wicked Rāksasa. The demon's head attached itself to the thigh of the sage Mahodara. The sage wandered from tirtha to tirtha trying to rid himself of this burden, but he had no success until he bathed at the Auśanasa tīrtha on the Sarasvati River. This place, named after the sage Uśanas or Sukrācārya, washed away the skull and thereafter became known as Kapāla- mocana.29 There can be little doubt that the two myths are related. Even the name, if not the location, of the sacred tirtha is the same. The Mahābhārata legend, however, contains no suggestion of Brahmahatyā. The Rāksasa's skull attaches itself to Mahodara because it is itself demonic, not because of the guilt of beheading. We suggest the following historical development. The Rāma- Mahodara story, or some similar prototype,30 was borrowed to provide the basis of the myth of the beheading of Brahmā, and this myth was then used to give divine sanction or precedent to the already existing legal prescription against killing a Brāhmaņa. The relative priority of the Saivite myth and the Kāpālika

26Trans. Barnett, BSOS, V, 713. Ed. Śāstrī, vs. 17. The Kūrma Purāņa (ii. 30-31) also says that Siva had to perform the penance of a Brahmahan but does not call it the Mahāvrata. 27See Kane, HDS, IV, 10-12. 28Ed. R.C. Dandekar, xxxviii. 1-20. Another version of this myth, in which the sage is called Rahodara, appears in the Vamana Purāņa xxxix. 1-14. 29This Kapalamocana is probably identical with a tank of this name on the Sarsuti or Sarasvati River ten miles south-east of Sadhaura. See A. Cunningham, Report. of a Tour in the Punjab in 1878-79, pp. 75-78. 30The Vedic myth of Indra's destruction of Vrtra, the demon son of Tvastr, is similar insofar as the sin of killing a Brähmana was thought to attach to Indra's deed.

80 THE KĀPĀLIKAS AND KĀLĀMUKHAS

ascetics themselves is also uncertain. Did the Kāpālikas invent the myth in order to provide a divine model for their ascetic obser- vance, or did they model the observance on the myth? The evidence is inconclusive. The sources in which the myth first appears, the Purāņas, also mention human Kāpālikas, and there are no references to the ascetics significantly earlier than these works. In some respects this question is a needless one. Since both the penance for killing a Brähmana and the association of Siva, the god of death and destruction, with skulls undoubtedly antedated the Śiva-Kapālin myth, Śaivite ascetics who observed the Mahāvrata might also have antedated it. Whether or not such ascetics existed and whether or not they themselves invented this myth, it is certain that the later Kāpālikas adopted it as their divine archetype. The ultimate aim of the Kāpālika observance was a mystical identification or communion with Siva. Through their imitative repetition of Siva's performance of the Mahävrata, the ascetics became ritually 'homologised' with the god and partook of, or were granted, some of his divine attributes, especially the eight magical powers (siddhis) .31 An important aspect of this ritual communion with Śiva-Kapālin seems to have been the identification of the devotee's begging skull with the skull of Brahma. As their name indicates, this skull bowl was the Kāpälika's trademark. In the Prabodhacandrodaya (act III, vss. 12-13), the Kāpālika describes himself as one who 'eats from a human skull' and says that `the conclusion of our fast (is accomplished) by drinking liquor distributed in the skull of a Brahman ( Brahma-kapāla)' The Kāpālika in Yaśaḥpāla's Moha- rājaparājaya (act IV, vs. 23) states : 'Nara-kāpālin declares that he who invariably eats human flesh in the skull of a noble man (uttama- purușa) obtains the position of Siva ( Śiva-sthāna).' Ugra-Bhairava, the Kāpālika opponent of Samkarācārya, claims that Giriśa (Śiva) had told him that he would attain the ultimate goal of men if he would 'sacrifice in the sacrificial fire either the head of an ominscient sage or the head pf a king.'32 In the Mattavilāsa, the Kapālin's wench laments that their lost skull 'was as splendid as the skull of the Lotus-throned God,' another allusion to the Śiva- Brahmā myth.33 We have noted that some of the law books specify

31 The psychology of this type of ritual identification with gods and heroes is well analyzed by M. Eliade in his Cosmos and History (chaps. i and ii). 32 Mādhavācārya, Śamkara-digvijaya xi. 11. 33Trans. Barnett, BSOS, V, 712-13. In Rājaśekhara's Kurpūramañjarī the tantric

KĀPĀLIKA CULT AND DOCTRINE 81

that the Mahävratin should carry the skull of the Brähmana he has slan as his alms bowl. This is what Siva does with the skull of Brahmā. It is unnecessary and unlikely that the Kāpālika Mahā- vratin first killed a Brähmana in order to obtain a skull bowl, but not any old skull; it seems, would suffice. It had to be the skull of a noble man (uttama-puruşa-kapāla) or the skull of a Brahman ( Brahma-kapāla). If our hypothesis about the ascetics' identification with Siva-Kapālin is correct, the term Brahma-kapāla would equally imply the skull of the god Brahmā. There remains one other Mahävrata we have yet to discuss. According to Patañjali's Yogasūtra ii. 30-31, when the five yamas (restraints) of ahimsā (non-injury), satya (truthfulness), asteya (non-theft), brahmacarya (chastity and restraint of the senses), and aparigraha (non-acceptance of more than is necessary for bodily subsistence) are practised without exception being made for status, place, time, or occasion, the observance is known as the Mahavrata.34 Its performance is incumbent on yogins at all stages.35 This Mahavrata, we believe, is the Mahavrata of the Kālāmukhas. Although the evidence to support this contention is rather slim, there is virtually no reason to connect the Kālā- mukhas with the Mahavrata of the Brāhmanas or the Mahāvrata of the Visnu-smrti.36 On the other hand, the Kalamukhas of the Ködiya-mațha in Belagāve (Shimoga District, Mysore) are said to study the Pātañjala and other Yogaśāstras,37 and most Kālā- mukha inscriptions stress the yogic attainments and virtues of these ascetics. Furthermore, the texts of the Pasupatas, the sect most closely related to the Kālāmukhas, lay particular emphasis on the performance of the five yamas. Kaundinya's commentary on the Pāśupata-sūtra attributed to the Pāśupata-Kālāmukha saint Lakulīśa devotes no less than nineteen pages to praise of

ascetic Bhairavānanda, who might be a Kāpālika (see above, p. 49). praises the goddess Käli, who drinks the blood of demons 'from a goblet made of the skull of Parameşthin [Brahmā].' (Trans. Lanman, Act IV, vs. 19). 34'ahimsā-satyāsteya-brahmacaryāparigraha yamāḥ/ jāti-deśa-kāla-samayānavac- chinnāh sārvabhaumāḥ Mahāvratam//' Ed. J. Ballantyne. 35See Kane, HDS, V, Part II, 1420. 36If one does not accept the identification of the Kāpalika Mahavrata with the penance of the Brahmahan, one could argue that the Yogasūtra Mahāvrata was also the vow of the Kāpālikas. The insistence of the Yogasūtra on absolute ahimsā and brahmacarya, however, makes this doubly unlikely. 37See below, p. 104.

82 THE KĀPĀLIKAS AND KĀLĀMUKHAS

the five yamas and five niyamas.38 In the absence of other alter- natives, it is best to assume that the Mahāvrata of the Kālāmukhas was the same as the Mahāvrata of Patañjali's Yogasūtra.

Somasiddhānta In a number of sources the doctrine of the Kāpālikas is called Somasiddhānta.39 Śrīharșa's Naișadhacarita contains a lengthy description of the goddess Sarasvati in which the various parts of her body are said to be formed from different philosophical doctrines. Her face is Somasiddhanta.40 The commentator Cāņdū- paņdita explains this as Kāpālika-daršana-śāstra.41 The Kāpālika characters in Krsnamisra's Prabodhacandrodaya and Ānandarāya's Vidyāpariņayana are both named Somasiddhānta. Gokulanātha's Amrtodaya (act II, after vs. 25) claims that Vardhamāna, the commentator on Udayana's Nyāyakusumāñjali, fought and killed Somasiddhanta, also called Somatantra. When Somasiddhānta fell, his comrades Kāpālika, Nīlalohita, Mahābhairava, Bhūta- dāmara, and Umāmaheśvara all fled the field. As we have noted, Kşīrasvāmin, a commentator on Amarakośa, identifies Mahā- vratin, Kapalin, Somasiddhantin, and Täntrika.42 We have also seen that a few Puränas and other sources contain sect lists which seem to replace Kāpālika by Soma, Sauma or Saumya.43 Raghūt- tama's commentary on Vātsyāyana's Nyāya-bhāșya includes Sauma in a list of six heretical doctrines (sad bahyā siddhāntāḥ). The six are Cārvāka, Sauma, Saugata, Jina, Ārhata, and Digambara.44 Somasiddhānta-vādins are also mentioned in the Akulavīra- tantra.45 G. Tucci has found allusions to a philosophical school called na ya siu mo in the Chinese translations of Harivarman's Tattvasiddhisāstra (fourth century A.D.) and Asanga's Madhyān- tānugamaśāstra. This school, Tucci believes, should be transcribed in Sanskrit as Nyāyasauma or Nayasaumya and is the same as Somasiddhānta.46

38Ed. R.A. Sastri, pp. 15-34. 39Most of the Sanskrit references to Somasiddhänta have been collected by Handiqui in the notes to his translation of Śriharsa's Naișadhacarita, pp. 640-44. 401bid., x. 87 (p. 149). 41Cited ibid., p. 427. 42See ibid., p. 640. 43See above, pp. 7-11. 44Cited by G. Tucci, JRASB, n.s. XXVI, 130. 45Cited ibid. 46Ibid., pp. 129-30.

KĀPĀLIKA CULT AND DOCTRINE 83

None of the sources which refer to Somasiddhänta say much about the term apart from identifying it as the name of the Kāpālika doctrine. Several commentaries on the Prabodhacandrodaya derive the word Soma from the compound sa-Umā (with Umā, i.e. Parvatī).47 Although this etymology is not historically correct, by the time of Krsnamiśra Soma or Someśvara was a common name for Siva. The sexual implications of the derivation sa-Umā are particularly suitable for the god of the Kāpālikas. A few inscriptions briefly mention the Somasiddhänta doctrine but do not contain any significant information about it. A priest entitled Caturanana-pandita, who headed the Tiruvorriyūr matha (Chingleput District, Madras) in A.D. 1171-72, is described as a contemporary of a Somasiddhāntin named Vāgīśa Bhațța.49 The priests of the Tiruvorriyūr matha were Mahāvratins. 49 Another allusion to the Somasiddhanta doctrine is found in an inscription from Mevundi in Dharwar District dated A.D. 897.50 As we have seen,51 Somakhadduka ascetics of the congregation of Munda- śkhalika Pāśupatācārya are named as donees in a seventh century grant from Nepal.

Kāpalika Bhakti While we possess no actual Kāpālika text, we can attempt to reconstruct the basic doctrines and attitudes of the sect from the many references we have cited. The keystone of the Kāpālika faith was bhakti, personal devotion to a personal god. This god was usually identified as Siva in his terrific Bhairava incarnation. The rituals into which the Kāpālika's bhakti was channeled were either propitiatory, imitative or a combination of both. The aim of these rituals was a mystical communion of the worshipper and his god. The rewards of this communion were twofold. On the mundane plain the devotee gained suprahuman magical powers ( siddhis) while on the eschatological plain he attained final libera- tion from transmigratory existence (mukti) and dwelt in a heaven of perpetual sexual bliss. The statements of Bodholbaņa-nityānanda and his Kāpālika followers in Anandagiri's Śamkara-vijaya form the locus classicus,

47See ibid., p. 131. See also V. Pansīkar's edition of the play, pp. 111, 113-14. 48See V. Raghavan, EI, XXVII, 297. 49See ibid., p. 300 and text vs. 3. 50C.R. Krishnamacharlu (ed.), SII, XI, Part I, no. 22, line 48. 51See above, p. 30.

84 THE KĀPĀLIKAS AND KĀLĀMUKHAS

as it were, of Kāpālika bhakti. When Śamkara asks them to describe the observances (ācāra) and precepts (vidhi) of their kula, they reply :

O Svämin, our observance, which is free from karman, causes satisfaction to all beings since it is said : 'There is no salvation with (or by means of) karman.' I should worship Bhairava alone, the creator of the world, who afterwards becomes the (cause of) destruction. He who is the cause of destruction is also the cause of preservation and creation ... All the gods, each endowed with a particular authority, are merely portions of him. They carry the command of Bhairava on their heads, and their powers, which attend on his word, are each directed to a particular duty.52

Parts of this passage are ambiguous but its general purport is clear. The word Karman, for instance, may imply either 'religious ritual and good works' or 'the effects of past good and bad deeds', although the latter interpretation is clearly the more plausible one. The chief object of the passage is to proclaim Bhairava to be the creator, preserver, and destroyer of the universe and lord of all the gods. The epithet 'world-creator' (jagat-kartr) suggests the dualistic distinction between the material and instrumental causes of the universe which Rāmānuja and other Brahma-sūtra commentators attribute to the Kāpālikas and other worshippers of Paśupati. The demotion of the many gods of the Hindu pantheon to the position of portions (amsas) of one primary god is a concept frequently found in bhakti literature, particularly the Purāņas, and is a form of what has been called henotheism, a kind of halfway house between polytheism and monotheism. Bodholbaņa-nityānanda and his followers continue with an elaboration of their doctrine. Bhairava, they claim, has eight major for.ns : Asitānga, Ruru, Caņda, Krodha, Unmatta, Kāpālin, Bhīşana, and Samhāra. The first seven of these forms they identify

52'Svāminn asmad-ācāraḥ sarvva-prāņi-santoșa-karaḥ karma-hīnaḥ, karmaņā na muktir iti vacanāt/ mad-upāsyo Bhairava eka eva jagat-karttā/ tatah pralayo bhavatīti yo vā pralaya-karttā sa eva sthity-utpattyor apīti/ ... tad-amsā eva sarvve devāḥ tat-tad-adhikāra-sampannāḥ śrīmad-Bhairavājnām śirasā dhrtvā tad-ukti- pratyāsanna-śaktayaḥ tat-tat-kāryya-parā babhūvuḥ/.' Ed. J. Tarkapanchanana, chap. xxiii.

KĀPĀLIKA CULT AND DOCTRINE 85

with the gods Visnu, Brahmā, Sūrya, Rudra, Indra, Candra,53 and Yama respectively. The eighth, Samhāra-Bhairava, is Bhairava himself. The remaining gods are merely his 'portions' and are further distinguished as creation-makers (srstikartrs), preservation- makers (sthitikartrs) and destruction-makers (samhārakartrs). Taken all together, the creation-makers are his Rudra (sic for Ruru-Brahmā) portions, the preservation-makers his Asitānga (Vișnu) portions, and the destruction-makers his Krodha (Rudra) portions. The Kāpālikas conclude : 'Thus having caused the creation of the world etc., and afterwards the dissolution, he makes a contraction of seven of his forms and one eternal Samhāra-Bhairava remains who is the paramatman.'54 This omnipotent deity demands both propitiation and imitation from his devotees. In this respect the Kāpalika faith differs from other theistic religions only in the procedures adopted. Ritual propitiation is sacrifice. Externally this usually takes the form of human or animal sacrifice. To be acceptable to the deity, the victim must be of auspicious color and size, unpolluted, and, in the case of humans, morally pure. At the same time, however, he is normally regarded as a scapegoat, the repository of the trans- gressions of the sacrificers. The Kāpālikas, if their critics are to be believed, specialised in human sacrifice. As we have seen, allusions to Kāpālikas performing human sacrifices, making offerings of human flesh, or doing pūjā with the aid of corpses are numerous. In Mālatī-Mādhava (act V. vs. 25) the faultless heroine is led forward wearing the marks of a sacrificial victim. The heartless Kāpālika Aghoraghaņța raises his weapon and invokes Cāmuņdā: O divine Cāmuņdā, the offering (pūjā) placed before you was promised at the beginning of the performance of incantations. May you (now) receive it.' More often the god the Kāpälikas invoke is Bhairava. At the end of our fast, says the Kāpālika in Prabodhacandrodaya (act III, vs. 13): 'Mahābhairava should be worshipped with offerings of awe-inspiring human sacrifices from whose severed throats blood flows in torrents.' To this god, he adds, we offer oblations of 'human flesh mixed with brains, entrails, and marrow.' Samkarācārya's Kāpālika opponent Krakaca puts the matter more forcefully : 'If he (Kapālin-

53 Another word for Candra (the moon) is Soma. The equation of Kapalin-Bhairava and Candra might have something to do with Somasiddhānta. 54Ānandagiri, chap. xxiii.

86 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Śiva) does not receive Bhairava worship with liquor and blood- smeared lotuses which are human heads, how can he attain joy when his body is embraced by the lotus eyed Umā ... ?'55 Here Bhairava seems to be not only gratified by head-offerings but in some sense dependent upon them. Although little reliance can be placed on the specific details of these statements-the authors were all opponents of the Kāpālikas -it is difficult to doubt that the Kāpālikas practised human sacrifice. The purpose of the rite was to appease and gratify a wrathful and blood-thirsty deity. The idea of the victim as a scapegoat is less explicit but is inherent, in any case, in the very concept of sacrificial propitiation. Human sacrifice existed in India, as in most parts of the ancient world, from a very early date. According to a legend which first appears in the Aitareya Brāhmaņa, the Īkșvāku king Hariścandra volunteered to sacrifice his first-born son to the god Varuna.56 The purușamedha (man-sacrifice) is described in a number of Brähmanas but had become merely symbolic by the time of the Śatapatha Brāhmana.57 Human sacrifices were given a new lease on life, as it were, with the emergence of tantric cults in the early mediaeval period. In some regions, particularly Bengal and Assam, the practice became fairly common. The sixteenth century Koch king, Nar Nārāyan, is said to have sacrificed about 150 men at a single ceremony.58 A combination of British suppression and Hindu reform virtually eliminated the practice by the early nine- teenth century, but cases of alleged human sacrifice are still reported sporadically.59 The important Śākta work, the Kālikā-Purāņa, devotes an entire

55Mādhavācārya, xi. 11. 56vii. 13-18, cited by E.A. Gait, 'Human Sacrifice (Indian),' ERE, VI, 849-53. See also P.B. Joshi, 'On the Rite of Human Sacrifice in Ancient, Mediaeval and Modern India and Other Countries,' JAnSB, III (1893), 275-300; and R. Mitra, 'On Human Sacrifices in Ancient India,' JRASB, XLV (1876), 76-118. 57See Joshi, JAnSB; III, 280. 58Gait, ERE, VI, 850. 59The Indian Express, August 15, 1966, reports a case from Medak District in Andhra Pradesh. The Milwaukee Journal, Sept. 15, 1968, reports that Prime Minister Gandhi sent 1,000 rupees to the family of a twelve year old boy sacrificed 'at the laying of the foundation stone for an irrigation project' in Rajasthan. The same paper, Oct. 31, 1968, contains an account of the beheading of another twelve year old boy at a town 200 miles southeast of New Delhi. The boy was sacrificed to Siva.

KĀPĀLIKA CULT AND DOCTRINE 87

chapter to animal and human sacrifice.60 It justifies the rite with arguments similar to those attributed to the Kāpālikas :

By a human sacrifice attended by the rites laid down, Devi ... remains gratified for a thousand years; and by the sacrifice of three men, one hundred thousand years. By human flesh the goddess Kāmākhya's consort Bhairava ... remains pleased three thousand years. Blood consecrated immediately becomes abrosia and since the head and flesh are gratifying, therefore should the head and flesh be offered at the worship of the goddess. The wise should add the flesh free from hair, among food offerings.61

Before executing his victim, the sacrificer says to him: 'Thou, by gratifying Candikā, destroyest all evil incidents to the giver. Thou, a victim who appearest as a sacrifice meet for the Vaisnavi, hast my salutations.'62 The scapegoat aspect of the sacrificial propitiation -of Bhairava and Durgā is here made more explicit. The personal counterpart to animal and human sacrifice is self-sacrifice. This concept subsumes a wide range of activities from self-immolation or suicide to self-mutilation and from physical penances to simple exercises of mental discipline. The chief penance performed by the Kāpālikas was, of course, the Mahävrata. There is also some evidence that they occasionally practised various forms of self-mutilation such as cutting flesh from their own bodies for sacrificial oblations.63 The Kāpālika Ugra-Bhairava claims to have gratified Ugra (Śiva) 'with arduous and severe penances for a full one-hundred years.'64 The Kāpālika in.Candakauśika (act IV, vs. 26) claims to subsist on unrequested alms and to have control over the five senses. The king greets him as a Mahävratin who has undertaken a vow of lifelong chastity (act IV, after vs. 29). In Prabodhacandrodaya (act III, vs. 13) Somasiddhänta claims to see 'through eyes made clear by the

60This chapter was translated at the end of the eighteenth century by W.C. Blaquiere in Asiatic Researches, V (1797). 371-391. We have not found an edition or more recent translation although H. Zimmer paraphrases parts of the Purāna in his The King and the Corpse. 61 Passage translated by Gait. ERE, VI, 850. 62Ibid. 63See above, pp. 17 and 76. 64 Mādhavācarya, xi. 10.

88 THE KĀPĀLIKAS AND KĀLĀMUKHAS

ointment Yoga' and to conclude his fast (pāraņā) by drinking liquor. Several references, albeit sarcastic ones, to Kapālin tapas also appear in Mattavilasa.65 Although the above allusions to Kāpālika asceticism and Yoga are few and not very detailed, it is evident that the authors were aware that the Kāpālikas were not simple hedonists. In addition to propitiating Bhairava through various kinds of sacrifice, the Kāpālikas imitated the god by ritual reenactment of his mythological exploits. To a large extent the paths of propitiation and imitation overlap. The Mahävrata, for instance, is both a propitiatory penance and a reenactment of the penance of Śiva. In some rituals, however, the idea of propitiation is absent or insignificant. Most of these are communion rituals in which the worshipper is united with divinity through food, drink, sex, or mental ecstasy. These rituals are normally preceded by propitiatory ones which give the devotee preparatory purification. The Kāpālika in Prabodhacandrodaya does not drink until he has fasted; Ugra- Bhairava gratifies Ugra with severe penances for one-hundred years 'in order to go to Kailāsa with this body to sport with Isa.' Kāpālika rituals of food and drink are referred to in a number of sources.66 In Yaśaḥpāla's Moharājaparājaya the Kāpālika says that one obtains Siva-sthana by eating human flesh in the skull of a noble man. The lost skull bowl of Mahendravarman's Kapālin was full of roast meat. Guņaratna and the Bārhaspati-sūtra claim that the Kāpālikas are sybaritic Nāstika materialists addicted to wine, meat and illicit intercourse. Samkara's opponent Krakaca fills his skull bowl with sura through his power of meditation. After drinking half of it, he invokes the god Bhairava. Unmatta- Bhairava, another of Śamkara's Kāpālika opponents, proudly declares that his father and gradfather were liquor makers and espouses a thoroughly hedonistic code of conduct. In Mattavilāsa the Kapālin similarly advocates wine and women as the road to salvation recommended by Śiva, and the Kāpālika in Prabodha- candrodaya describes wine as the 'remedy against (transmigratory) existence prescribed by Bhairava.' In Ānandarāyamakhin's Vidyā- pariņayana (act IV, after vs. 32) the Kāpālika Somasiddhānta almost apologetically defends his use of wine and meat by main- taining that they are prescribed in the Bhairavāgamas : 'We are

65 Mahendravarman, after vss. 6 and 10 and vs. 21. 66See above, chap. ii.

KĀPĀLIKA CULT AND DOCTRINE 89

counted among heretics through divergence from the Veda by addiction to wine (madhu), meat (māmsa), etc., which are prohi- bited in the Vedas, (but in fact we hold) the doctrine of the authori- tativeness of the Veda with compliance to the Bhairavāgamas.' Since the Kāpālikas were a tantric Saivite sect, their addiction to meat and wine, as well as sex, should be associated with the five Ma-sounds (pañcamakāra) of tantric tradition and not with hedonistic materialism. The passage from Vidyāpariņayana men- tioning madhu and māmsa tends to confirm this association. In Ānandagiri's Samkara-vijaya the Kāpālika Bodholbaņa declares (chap. xxiii) that true and fearless sages are `all always dependent on knowledge (bodha) produced from substances (dravya).' These 'substances' probably represent the five Ma-sounds since the terms pañcadravya and pañcamakāra (also pañcatattva) are synon /mous in tantric texts. The context of Bodholbana's statement alsc tends to support this interpretation. In tantric practice the partaking of wine and meat has both a hedonistic and eucharistic aspect but is in no way connected with materialism. Hedonistically, the first four of the five Ma-sounds -wine, meat, fish, and grain (mudrā)-are regarded as aphrodisiac (uttejaka) preparatives to the final maithuna or sexual union between the initiated adept and his female partner. These four ingredients do not in fact possess aphrodisiac qualities although wine, of course, may help to release inhibitions. A. Bharati points out that the only substance used in tantric sādhanā which has any such qualities is vijaya or Indian hemp (Cannabis Indica).67 This is taken about an hour and a half before the five Ma-sounds. The eucharistic significance of the four preliminary ingredients is variously explained in tantric sources. Kulārņava-tantra v. 79-80 says : 'Wine (sura) is Sakti; the meat is Siva; the enjoyer of those is Bhairava himself. The bliss sprung from the union of those two (? =Siva and Śakti) is calied moksa. This bliss, which is the form (rūpa) of Brahman, is established in the body (of the worshipper). The wine makes it manifest. For that reason the yogins drink.' The reformist Mahānirvāņa-tantra states: 'Wine [surā] is Tārā Herself in liquid form, Who is the Saviour of beings, the Mother of Enjoyment and Liberation.'68 In the Kaulāvalinirnaya the goddess is worshipped as the surā which was churned from the milk-

67 The Tantric Tradition, p. 252. 68Trans. Woodroffe, xi. 105.

90 THE KĀPĀLIKAS AND KĀLĀMUKHAS

ocean and emerged from the kula-nectar : 'Having eighteen arms, lotus-eyed, born on the summit of bliss, (and thence also originated) bliss as Maheśvara. From their union come forth Brahma, Vișnu, and Siva. Therefore I drink thee with my total personality, o goddess of liquor.'69 Bharati's tantric informants variously inter- preted the term kula-nectar (kulāmrta) as 'the spiritual essence of the five makāras ... , the cosmic residuum caused by Siva's and Śakti's eternal copulation'; 'the liquid which emerges from the contact of Siva and Sakti'; and the rajas (menstrual blood) of the goddess.70 Saktisangama-tantra ii. 32. 25 says that the wine 'is produced from the rasa of Sakti.' This statement seems to have similar sexual implications although it appears in the context of a list of substitutes (pratinidhi) suitable for dakşiņācāra worship. All of the symbolic equations just cited clearly indicate the presence of an element of totemic communion in the ritual con- sumption of the first four Ma-sounds. This alimentary communion is based on the archaic maxim that we are what we eat -- man ist was er isst. The identification of the ritual foods with the body or body products of Siva and Sakti confers on the communicant consubstantiality with them. He becomes the god and shares. various divine attributes such as immortality and magical powers. Before discussing these supernatural benefits, however, we must examine the central ritual of tantric communion, sexual intercourse. Since some of our sources for the Kāpālikas are quite explicit about the significance of this ritual, it is not necessary to rely on tantric works except for confirmation. The archetypal basis of the ritual is delineated most succinctly in the traditional etymology for the term Somasiddhanta-the doctrine of Soma (Siva) united with Umā (Umayā sahitaḥ Somas tasya siddhāntaḥ).71 The human participants of the ritual mentally identify themselves with Siva and Sakti respectively. In the bliss of sexual union the human pair realize the divine bliss of Śiva and Śakti. Final salvation (mukti), on this view, is perpetual orgasm, not merely extinction of the cycle of rebirth. When asked about his conception of moksa, the Kāpālika in Prabodhacandrodaya replies (act III, vs. 16) :

69Trans. by Bharati in The Tantric Tradition, p. 259. 701bid., pp. 259-60. 71 Prakāśațīkā to Krșņamiśra's Prabodhacandrodaya, ed. Paņśīkar, p. 114. See also Candrikāvyākhyā, ibid., p. 11l, and G. Tucci, JRASB, n.s. XXVI, 131.

KĀPĀLIKA CULT AND DOCTRINE 91

Thus spoke the Lord of Mrdānī (Siva): 'Bliss is not found anywhere without sense objects. How (can) mukti be desired (when) the condition of the soul (jiva) is the condition of a stone, devoid of the awakening of bliss. One who has the appearance (vapus) of the Moon-crested (Siva) and amuses himself in the embrace of his wife, the image of Pārvati, is (truly) liberated.'

A similar view is put forward by the Kāpālika Unmatta-Bhairava in the commentary to Madhava's Samkara-digvijaya xv. 28 : 'The bliss which becomes manifest through sexual union is the (true) form of Bhairava. The attainment of that (bliss) at death is mokșa. This is the ultimate truth.'72 The Kāpālika Bodholbaņa-nityānanda praises the fearless sages who are 'always dependent on the knowledge produced from substances (dravya), whose hearts are gratified by the embrace of Kāpālika śaktis, who are addicted to drinking the excellent nectar arising from sexual union, ... and who (declare that) Bhairava is the abode (pada) in death.'73 The drinking of the nectar of sexual union is probably a reference to the yogic exercise of reabsorbing with the penis the seminal fluid discharged in coitus. The rationale for this practice, called the vajrolimudrā, is explained in the Hathayogapradīpikā: 'Having drawn up his own discharged bindu [the Yogi] can preserve (it) ... By the loss of bindu (comes) death, from its retention, life.'74 The same idea lies behind the allied practice of coitus reservatus recommended especially in Buddhist Vajrayāņa texts. Breath (prāņa), thought (citta), and semen (bindu)-the three jewels-must be simulta- neously 'immobilised' in an act which yields the perfect state of oneness in duality.75 The belief that the loss of semen causes the destruction of mental and spiritual as well as physical potency is widespread even in modern industrial societies. In India the asso- ciation of celibacy and religious or magical power has been stressed since early times. To cite just one example, a legend about the rși Dadhīca in the Salyaparvan (chap. 1) of the Mahābharata tells how the gods became imperilled by the sage's growing ascetic

72Dhanapatisūri, Dindima commentary, vs. 22. 73Ānandagiri, chap. xxiii. 74Trans. Briggs in Gorakhnāth ... , p. 334. 75See Eliade, Yoga ... , pp. 248-49, 253-54 and Bharati, p. 265.

92 THE KĀPĀLIKAS AND KĀLĀMUKHAS

power (tapas) and sent a beautiful Apsaras to earth to tempt him. When Dadhica spied the celestial nymph, he lost his semen, and consequently his sacred power, in the Sarasvati River. The Sanskrit word for religious novice, brahmacārin, quite early came to refer mainly to sexual continence although its original etymological meaning was 'moving in Brahman,' one whose mind is fixed on the absolute.76 The Kāpālika in Caņdakaušika (act IV, vs. 34) implies a sexual- or at least a sensual-conception of moksa when he praises the immoral world where the Siddhas frolic on the peaks of Meru. In Rāmānuja's Śrībhāșya ii. 2. 35-37, the Kāpālas declare that 'he who meditates on the Self as seated in the female vulva attains nirvāņa.' This statement may reflect a partial spiritualisation or sublimation of overt sexual ritual. Vämamārg, a modern tantric manual in Hindi and Sanskrit by V.S. Vaidyarāj, describes the climax of pañcamakāra-sādhanā in terms similar to those attributed to the Kāpalikas : 'Viewing the Sakti as Gauri (i.e. the spouse of Siva) and himself as Siva, he [the sdhaka] should pronounce the root-mantra of his chosen deity and should offer that father-face into the mother-face.'77 During the sexual act the sādhaka should mentally recite a mantra verse to the goddess. By this means 'he creates the attitude of the oneness of Siva and Sakti.'78 As he 'abandons his semen' he should recite the following mantra: 'Om with light and ether as my two hands, I, the exulting one, relying on the ladle, I, who take dharma and non-dharma as his sacrificial ingredients, offer (this oblation) lovingly into the fire, svāhā.'79 Here orgasm is both communion and sacrifice! The aim of the Kāpālika's religious endeavours is not simply the attainment of a state of divine bliss. On a more mundane or practical level, he seeks magical yogic powers (siddhis). These may be won either through the achievement of consubstantiality with Siva in rituals of communion or, more directly, as a gift

76See A. Bharati, The Ochre Robe, p. 99. The psychiatrist-anthropologist G.M. Carstairs found that a preoccupation with the involuntary discharge of semen, the source of bodily and spiritual strength, forms 'the commonest expression of anxiety neurosis among the Hindu communities of Rajasthan, and perhaps elsewhere as well' (The Twice Born, p. 87). 77Trans. Bharati in The Tantric Tradition, p. 264. 78Trans. ibid., p. 265. 79Trans. ibid.

KĀPĀLIKA CULT AND DOCTRINE 93

from the deity earned by penance or sacrifice. The priest-magician existed in India, as elsewhere, from earliest times. His penances, spells, and magic rituals gave him the power to perform supernatural deeds with or without the assistance of the gods. The development of the doctrine and practices of Yoga led to a systematic cultivation and enumeration of the priest- magician's magical powers.80 In the Yogasūtra of Patañjali, generally dated sometime between 200 B.C. and A.D. 300,81 a considerable number of magical powers are said to stem from the practice of samyama.82 The powers include the ability to know present, past and future, to become invisible, to become strong as an elephant, to enter another's body, to walk on water or thorns, to hear inaudible sounds, and to fly through the air. Elsewhere in the Yogasūtra Patañjali states that the siddhis may be obtained by any of five methods : birth, drugs, mantras, penance, and samādhi. Yājñavalkya-smrti iii. 202-203, a work slightly later than or contemporary with the Yogasutra, says that supra-normal powers of hearing, seeing, remembering, becoming invisible, abandoning one's body, and entering another's body are the mark ọf Yoga-siddhi. The Rājamārtanda commentary on Patañjali by King Bhoja (early eleventh century) contains a list of eight great siddhis (mahäsiddhis) which can be won by Yoga : (1) animan, the power of becoming small; (2) laghiman, the power of levitation; (3) gariman, the power of becoming heavy; (4) mahiman, the power of becoming limitlessly large; (5) īśitva, control over body and mind; (6) prākāmya, irresistible will; (7) vaśitva, control over the five elements; and (8) kāmāvasāyitva, fulfilment of desires.83 Similar lists are found in the Yogabhāsya of Vyäsa (seventh to eighth centuries), the tantric Prapañcasāra, and other works.84 In spite of the abundant textual references to various siddhis in classical Yoga texts, many modern Indian scholars, and like- minded western ones as well, have seized on a single sūtra of Patañjali (iii. 37) to prove that magical powers were regarded as

80Perhaps the oldest mention of the siddhis is in the Āpastambīya Dharmasūtra ii. 9. 23. 6-7. 81See Kane, HDS, V, Part II, 1395-99 and Eliade, Yoga ... , pp. 370-72. 82 Yogasūtra iii. 16-50. The term samyama refers to the last, and highest, three 'limbs of Yoga' : concentration (dhāraņā), meditation (dhyāna) and samādhi. 83Commentary on Yogasūtra iii. 44, cited by Eliade, Yoga ... , p. 88. Most of the English equivalents given are based on Eliade's renderings. 84See, Kane, HDS, V, Part II, 1112-13.

94 THE KĀPĀLIKAS AND KĀLĀMUKHAS

subsidiary, and even hindrances, to final liberation and conse- quently not worthy of concentrated pursuit.85 This attitude may have been operative in Vedantic and Buddhist circles and is now popular among practitioners imbued with the spirit of the Hindu Renaissance, but it was not the view of Patañjali and certainly not the view of mediaeval exponents of Hatha Yoga. Arthur Koestler has pointed out that the sūtra in question seems clearly to refer back only to the powers mentioned in the previous one or two sūtras and not to the many powers mentioned afterwards.86 He concludes that 'all disclaimers notwithstanding, the siddhis are an integral part of Yoga,' a statement that has the explicit support of no less a scholar than P.V. Kane.87 Most tantric sects were well-infused with the doctrines and practices of Hatha Yoga, and it is unlikely that the Kāpālikas were an exception. Our sources suggest that they were especially pre- occupied with magic and the siddhis. The Kāpālika Ugra-Bhairava laments to Samkara that 'the skull of an anointed king or a lord of munis is the prerequisite for my siddhi.'88 Samkara's enemy Krakaca fills a skull bowl with wine through the power of medi- tation.89 In Candakauśika (act IV, vss. 31-32) the Kāpālika offers King Hariścandra a large collection of magical skills and equipment as well as a great treasure of immortality-bestowing siddharasa ( ?= a mercurial drug). Other references to the magical powers of Kāpālika ascetics appear in the Kathāsaritsāgara stories of Madana- mañjarī, Candrasvāmin, Devadatta, and the Kāpālika spy.90 In Jambhaladatta's Vetālapañcavimśati (pp. 10-11) the Kāpālika mutters a great incantation (mahāmantra) in order to obtain siddhi. The Kāpālin-Pāśupata Aśvapāda in Kalhaņa's Rājataran- giņī displays the ability to remember his past lives and to magically transport his disciple to Kashmir.91 Krsņamiśra's Prabodhacandrodaya (act III, vs. 22) contains a particularly interesting allusion to the siddhis of the Kāpālikas.

85Even as objective a writer as Eliade partially succumbs to this view (Yoga .... pp. 88-90). He and other scholars also ignore the mention of drugs among the means of obtaining siddhis, perhaps for similar reasons. 86 The Lotus and the Robot, pp. 110-11. 87HDS, V, Part II, 1451-52. 88 Mādhavācārya, xi. 14. 89Ibid., xv. 23-24. 90See above, pp. 62-64. 91See above, pp. 66-67.

KĀPĀLIKA CULT AND DOCTRINE 95

Somasiddhänta here claims that in his doctrine the devotee gains the eight mahasiddhis without renouncing the pleasures of the senses. In other doctrines, he says, even the ordinary siddhis (prākṛtasiddhis) of subjecting (vaśya), attracting ( ākarșa); bewil- dering (vimohana), stupefying (praśamana), agitating (prakșo- bhana), and removing (uccāțana) are no more than obstacles for the learned. This list seems to be unique, but the idea that certain siddhis may be obstacles probably alludes to Yogasūtra iii. 37. The fifth act of Bhavabhuti's Malati-Madhava begins with the entrance by an aerial path of Kapālakuņdalā, the female disciple of the Kāpālika Aghoraghanta. In her opening invocation to Siva she asserts that the god's 'ätman is situated in the midst of the ten nādis and six cakras' and that he 'gives siddhis to those who know (him).' The theory of the six cakras and ten nādis forms the core of the mystical physiology of Hatha Yoga. Since this theory is well-known92 we need only note here its association with the Kāpālikas. Kapālakundalā then tells how she flies through the sky, clearing the clouds in front as she goes. She claims to perceive the ätman manifested in the lotus of the heart as the form of Siva through her power of yogic absorption (laya-vaśāt) and to fix it in the six cakras by the practice of nyāsa.93 Then she causes the drawing off of the five elements from the body by means of the swelling of the nādīs (with the breath restrained by prānāyāma) and flies up into the air.

92See Eliade, Yoga ... , pp. 236-45, and the Sat-cakra-nirūpana, ed. and trans. J. Woodroffe, The Serpent Power. 93The Hatha Yoga and tantric meditation called nyasa 'ritually projects' various divinities into different parts of the body by touch and mantra recitation.

CHAPTER IV

KĀLĀMUKHAS OF THE SAKTI-PARIȘAD

Preliminary The Kālāmukha sect of Śaivite ascetics inhabited the Karņāțaka region mainly during the eleventh, twelfth and early thirteenth centuries. The name Kalāmukha, sometimes spelt Kālamukha, may refer to a practice of marking their foreheads with a black streak.1 Judging from the large number of epigraphs recording donations to Kālāmukha temples and mathas, these ascetics must have wielded considerable influence in the region. Unfortunately few indications of their beliefs and ritual survive apart from the information which can be gleaned from these epigraphs. They reveal the existence of at least two major divisions of the Kālāmukha order-the Sakti-parişad and the Simha-parisad. Records of the · latter division have been found over a wide area including various parts of Andhra Pradesh and Mysore. The former division seems to have been limited mostly to the Dharwar and Shimoga Districts of Mysore. Nonetheless, the number of extant Sakti-parişad epigraphs is greater and they have been found at a larger number of sites. Moreover, they are generally of greater length and contain more religious information. Barring historical accident, it must be assumed that the Sakti-parisad was the more important of the two groups. Approximately sixty-five inscriptions from eighteen Sakti-parişad temples have been found and published. Its control over two of the temples is doubtful, however, and a few of the inscriptions, though found in Kālāmukha temples, date from a period either before or after Kālāmukha occupation. Four separate subdivisions of the Sakti-parisad are distinguished, and it may be assumed that others existed whose names have not survived. The most prominent division was centred in the Kedāreśvara temple at Belagāve in Shimoga District.2 The ascetics styled themselves as

IT.A.G. Rao, Elements of Hindu Iconography, II, Part I, 25. 2Most of the records found at this temple have been edited and translated by B.L. Rice in EC, Vol. VII. This line of ascetics has been discussed at some length by J.F. Fleet ('Inscriptions at Ablur,' EI, V [1898-99], 213-65) and by A. Venkata

98 THE KĀPĀLIKAS AND KĀLĀMUKHAS

members of the Mūvara-koneya-samtati (or -santāna) of the Parvatāvaļi (or Parvatāmnāya) of the Sakti-parșe (or -parișad). No less than fifty of the sixty-five epigraphs refer to this line of ascetics. About twenty-two of the records are located at the Kedā- reśvara temple itself. The rest are at five other temples in the region. The same ascetics seem to have been in charge of all six temples. At two of them, however, the connection with the Sakti-parisad is based merely on the correspondence of ascetics' names and dates. The records are slightly inconsistent about the hierarchy of the three parts of the organisation. The Kedāreśvara epigraphs of 1094 and 11033 and the duplicate Ablür epigraph of 1101-44 refer to the Mūvara-koneya-samtati of the Parvatāvaļi of the Śakti-parșe, but the Kedāreśvara record of 11135 seems to refer to the Sakti-parşe of the Mūvara-koņeya-santāna of the Parva- tāmnāya. The 1129 and 1156 Kedāreśvara records6 mention only the Mūvara-köneya-samtati of the Parvatāvaļi, while the Ablūr record of 1144 and the Kedāreśvara one of 11647 only mention the Mūvara-koņeya-samtati. The Kedāreśvara record of 11938 refers to the Parvatāvali alone. This confusion is easily resolved by comparing the names used by the other subdivisions of the Sakti-parisad. An inscription from Hombal in Dharwar District praises some ascetics belonging to the Parvatāvali and Belleya-santāna.9 Another from Gogga in Shimoga District mentions the ascetics of the Sakti-paridhi of the Parvatāvali and A .. ka-santati.10 One from Matțikoțe in Shimoga District eulogises ascetics of the Sakti-parse of the Bhujangāvali of the Ittige-samtati.11 If all these terms are collated only one order of precedence is possible-samtati of āvaļi of parișad.

Subbiah ('A Twelfth Century University in Mysore,' Quarterly Journal of the Mythic Society [Bangalore], VII [1917], 157-96). The architecture and sculpture of this and many other of the Kālāmukha temples are described by H. Cousens ( The Chāluk yan Architecture of the Kanarese Districts) and in Mysore Gazetteer (ed. C. Hayavadana Rao, Vols. Il and V). 3Ed. and trans. Rice, EC, VII, Sk. 94 and 98. 4Ed. and trans. Fleet, EI, V, Nos. A and B. 5Ed. and trans. Rice, EC, VII, Sk. 99. 6Ibid., Sk. 100 and 104. 7Ed. and trans. Fleet, EI, V, No. C and Rice, EC, VII, Sk. 108 respectively. 8Ed. and trans. Rice, EC, VII, Sk. 105. 9Ed. P.B. Desai, SII, XV, no. 73. 10Ed. and trans. Rice, EC, VII, Sk. 316. 11Ed. and trans. Rice, EC, VII, Sk. 292.

KĀLĀMUKHAŚ OF THE ŚAKTI-PARIȘAD 99

These various terms clearly denote organisational divisions, although they may have encompassed some doctrinal differences as well. A few of the terms derive from Kannada, the language of most of the inscriptions, while the others derive from Sanskrit. Parșe is a Kannada variant of parișad (group, assembly, council). Āvaļi (row, line, lineage, dynasty), āmnāya (sacred tradition or texts, instruction), and samtati or santāna (continuation, lineage, offspring) are common Sanskrit words, but their use in this context is rare. We may translate the Sakti-parisad as the Assembly-of-the- Goddess and the Parvatävali as the Mountain-Lineage. The latter term probably refers to the sacred mountain Srīparvata or Śrīśaila in Kurnool District. A priest of the Parvatavali named Rāmeśvara was presiding over the Mallikārjuna-silā-matha at this site in A.D. 1090.12 Sriparvata was an important pilgrimage center for the Kālāmukhas and is frequently mentioned in their epigraphs. Parvata might also refer to the holy Himalayan mountain Kedāra- näth commemorated in the name of the Belagāve temple or to the goddess Pärvatī, who was herself of the lineage of the Mountain. The Bhujangāvali or Serpent Lineage may be an allusion to the association of Siva with the cobra. The Kannada term Mūvara- kōņeya is obscure. J.F. Fleet notes (EI, V, 219) :

Muvara must be the genitive of mūvaru, 'three persons', unless it can be connected with mū, = mudu, 'advanced age'. For köne, of which köneya is the genitive, the dictionary only gives the meanings of 'a pitcher; an inner apartment or chamber, a room'.

Belleya seems to be the genitive of belli, a Kannada word meaning 'silver.' Ittige, 'a brick,' is still current in Kannada. It is derived from the Sanskrit istakā or ișțikā.

The Mūvara-köņeya-samtati The Mūvara-koneya-samtati of the Parvatavali had its head- quarters at Belagäve, but its control extended to about five additional sites in the surrounding region. These are the Brahmeśvara temple

12See the inscription ed. and trans. by P. Sreenivasachar. HAS, XIII, Part II, no. 7. Sreenivasachar mistakenly read 'Appaparv(v)atv= Aļiya Rāmēśvara-pam- ditarg(ge)' for 'Appaparv(v)atāvaļiya Rāmēśvara-pamditarg(g)e.' The correction was made by N. Venkata Ramanayya in G. Yazdani (ed.), The Early History of the Deccan, II, 705.

100 THE KĀPĀLIKAS AND KĀLĀMUKHAS

at Ablūr, the Mallikārjuna temple at Haļe-Nidnēgila, the Trikū- teśvara temple at Gadag, and the Nagareśvara temple in Südi-all in Dharwar District-and the Kotīśvara temple at Devasthāna- Hakkalu near Kuppațūr in Shimoga District. The last two sites, however, cannot with certainty be said to belong to this line. The earliest inscription at the Kedāreśvara temple in Belagāve is dated by Rice at c. A.D. 1078.13 It is a grant made to the priest Vālmīki-muni, the second in descent from Kedāraśakti-munipati. Another priest second in descent from Kedāraśakti was Someśvara- pandita-deva. He is the donee in four grants : three dated A.D. 1094, 1103 and 1113 from the Kedāreśvara temple and one dated 1101 from Ablur.14 Subtracting about twenty-five years for each priest, Kedāraśakti must have headed the monastery between about 1025 and 1050. If the Belagāve-Ablūr Someśvara is identical with the ascetic by that name teaching at Sūdi as early as 1060,15 Kedā- raśakti may be placed slightly earlier. The name Kedāraśakti suggests that he may have been the founder of the Kedāraśvara temple and priesthood. The latest inscription of this priesthood found at the temple is dated 121516 although a collateral line at Gadag has left a record dated 1225.17 If we can identify the priest- hood at Devasthana-Hakkalu as another collateral line, the period may be extended up to the twelfth year of Rāmacandra-rāya of the Seunas, or 1280.18 This is one of the latest dates of all Kālā- mukha epigraphs. The majority of the Mūvara-koņeya-samtati records are dated in the second half of the twelfth century. The diagram of the genealogy of this priesthood is on next page. The full name of the form of Śiva who presided over the Belagāve temple was Dakşiņa-Kedāreśvara, Lord of the Southern Kedāra. This implies a comparison with the northern Kedarasvara, the god of the famous and holy Kedara Mountain in the Himalayas. The Belagäve temple was built in the southern portion of the town

13Ed. and trans., EC, VII, Sk. 107. 14Ibid., Sk. 94, 98, 99, and Fleet, EI, V, No. A-B respectively. 15See L.D. Barnett, 'Inscriptions of Sūdi,' EI, XV (1919-20), 73-112 (No. F). A.V. Subbiah (QJMS, VII, 184) claims that the Kedāreśvara temple did not yet exist in A.D. 1054. This statement is based on the absence of any mention of the temple in a record dated in this year (ed. and trans. Rice, EC, VII, Sk. 118), which lists most of the other temples in Belagäve, but not the Kedāreśvara. The portion of the record in which this list appears, however, seems to date from about a century later. 16Ed. and trans. Rice, EC, VII, Sk. 95. 17Ed. Desai, SII, XV, no. 609. 18Ed. and trans. Rice, EC, VIII, Sb. 275.

KĀLĀMUKHAS OF THE SAKTI-PARIȘAD 101

KEDĀRAŚAKTI-munipati

RUDRABHARANA ŚRIKAŅȚHA-paņdita I

VALMĪKI-muni SOMEŚVARA-paņdita- (c. 1078) deva (1094, 1103, 1113) (1101-4 Ablūr)

VIDYĀBHARAŅA (1129)

VĀMAŚAKTI-munīśvara I JÑĀNAŚAKTI I (Ibid. 1129) (1130, 1144 Ablūr)

GAUTAMA-muni CANDRABHŪȘAŅA (Ibid. 1129, 1139, paņdita-deva 1149)- (1191, 1192, 1199 Gadag) VĀMAŚAKTI-paņdita- deva II (1156, 1159, KRIYĀŚAKTI- c. 1160, 1162, 1164, 1168, 1171, ?1181, paņdita (1213, 1225 1181, 1193) (1165 Gadag) Haļe-Nidnēgila)

JÑĀNAŚAKTI-deva II (?1181)

ŚRIKANTHA-deva II

VĀMAŚAKTI-deva III (1215)

102 THE KĀPĀLIKAS AND KĀLĀMUKHAS

on the bank of a tank called Tāvaregere or Tāvareyakere, 'the tank of the water lilies.'19 The priests of the temple also controlled another temple or shrine at this site dedicated to the god Nakhareśvara or Nagareśvara.20 In A.D. 1139 a third shrine was constructed at the site by two sculptors who set up an image of the god Kusuveśvara and presented the `temple of the god ... as attached to the god Kedāreśvara. '21 The inscription of c. 1078 contains the following genealogy :

In the world-renowned Sakti-parşe, in the Mūvara-kōņeya- santati (? of the Parvvatāvaļi), shone Kedāraśakti-munipati. His disciple, an ornament to the Lākula-samaya, was Rudrābharaņa. His disciple was Vālmīki-muni (his praise, including)? a hand to Lākula.22

The remainder of the inscription is damaged. It records a gift to a temple, presumably the Kedāreśvara. This line of ascetics seems to have died out with Välmiki-muni since both he and his predecessor, Rudräbharana, are mentioned only in this record. The term Lākula-samaya, 'doctrine of Lakula,' and other references to this Saivite saint frequently appear in Kālāmukha epigraphs and will be discussed below. The special significance, if any, of the phrase 'a hand to Lakula' is not known. The main line of Kedāreśvara pontiffs passed from Kedāraśakti through Śrīkantha-pandita to Someśvara-pandita-deva. Ignoring some minor variation, the four grants written during Someśvara's reign describe Kedāraśakti as follows :

In the line named Parvatāvali, which was esteemed to be greatly (i.e. undoubtedly) the leading (division) of the sect, celebrated in the world named Sakti-parse, there became famous the eminent ascetic Kedāraśakti, an ornament to the succession named Mūvara-köņeya-samtati.23

The grant of A.D. 1113 adds the important information that this priest and his disciples were included 'among the Kāļamukhas,

19Fleet. EI, V, 221. 20Ibid. 21Ed. and trans. Rice, EC, VII, Sk. 112. 22Ibid., Sk. 107. 23Ed. and trans. Fleet, EI, V, 219.

KĀLĀMUKHAS OF THE SAKTI-PARIȘAD 103

who ... had caused themselves to be spoken of as the very burst of the rainy season for the cātaka-birds that are disciples.'24 Someś- vara is called a 'Kāļamukhācārya' in the Gadag record of 1192 and his disciple's disciple Siddhānti-candrabhūșaņa is said to have 'sprung from the lineage of Kālamukha ācāryas.'25 Several inscriptions of other branches of the Sakti-parisad establish the connection between it and the Kālāmukha order beyond any doubt. The matha (cloister or college) of these priests is called the Ködiya-matha in the Kannada grants and the Kōti-matha in a Sanskrit passage from the record of 1215.26 It is also referred to as the Kedāra-matha and the Kedāra-sthāna. On the basis of the inscription of 1159 Fleet suggested that the matha was built by the hergade Vennamarasa,27 but the passage which he translates as 'the Ködiya-matha of the Hergade Vennamarasa' is ambiguous. Rice in fact connects this Vennamarasa with the Tripurāntaka temple in Belagāve, not the Ködiya-mațha.28 More plausible is Fleet's suggestion that the matha was so named because 'it stood somewhere near the ködi or outlet of the Tavaregere tank.'29 Among the several descriptions of the Kedāreśvara temple and matha, the following from the record of A.D. 1162 is the most striking :

There is the Ködiya-matha, which has become the abode of the god Kedära of the South,-a very field charming with a crop which is the standing erect of the hairs of the body that is induced by doing worship to the linga of Siva, -a place devoted to the observances of Saiva saints leading perpetually the life of celibate religious students,-a place for the quiet study of the four Vedas, the Rc, Yajus, Sāman, and Atharvan, together with their auxiliary works,-a place where commentaries are composed on the Kaumāra, Pāņinīya, Śākațāyana, Śabdānušāsana, and other gramma- tical works,-a place where commentaries are composed

24Ibid., 221. 25Ed. H. Lüders, 'Gadag Inscription of Vira-Ballala II,' EI, VI (1900), 96-97. 26Ed. and trans. Rice, EC, VII, Sk. 95. See Fleet, EI, V, 221-22. 27 EI, V, 221-22. 28Ed. and trans. Rice, EC, VII, Sk. 123. The text reads : 'Śrī-Tripurāntakad ācāryya Jñānaśakti-paņdita-dēvara mattam alliya herggade Veņņamarasa Śrī- Kōdiya-mațhad ācāryya Vāmaśakti-paņdita-dēvara ... ' 29EI, V, 222.

104 THE KĀPĀLIKAS AND KĀLĀMUKHAS

on the six systems of philosophy, namely the Nyāya, Vai- śeşika, Mīmāmsā, Sāmkhya, Bauddha, etc.,-a place where commentaries are composed on the Lākuļa-siddhānta, and the Pātañjala and other Yogaśāśtras,-a place for (studying) the eighteen Purānas, the law books, and all the poetical compositions, the dramas, the light comedies, and the other various kinds of learning,-a place where food is always given to the poor, the helpless, the lame, the blind, the deaf, and to professional story-tellers, singers, musicians, bards, players, and minstrels whose duty it is to awaken their masters with music and songs, and to the naked and the crippled [nagna-bhagna], and to (Jain and Buddhist) mendicants [kşapaņakas], to (Brāhmaņa) mendi- cants who carry a single staff [ekadandins] and also those who carry a triple staff [tridandins], to hamsa and parama- hamsa ascetics, and to all other beggars from many countries [nānā-deśa-bhikșuka-jana],-a place where many helpless sick people are harboured and treated,-a place of assurance of safety for all living creatures.30

The description continues with an elaborate and uninspired series of metaphors and similes which compare the temple and its matha to various mythological places and personages. These metaphors and similes also appear in the records of 1129 and 1156.31 The 1129 record adds an attractive comparison of the temple and the Himalayan mountain Kedāra :

Moreover the course of the sacred bathing streams there at the temple is like that of the Ganges at Kedāra, the lofty tower of the Siva temple piercing the sky rises up like the peak of Kedära, and the holy ascetics performing penance there are like holy ascetics at Kedāra whose minds are bent on the performance of the most difficult penances,- thus this is a new Kedära, the standing crops of its fertile fields resembling the horripilation arising from the Siva- linga worship, its temple the abode of Parameśvara. The god Kedära therein, who, thinking with supreme benevolence on his faithful worshippers,-afraid of the

30Ed. and trans. Fleet, ibid. 31 Ed. and trans. Rice, EC, Sk. 100 and 104.

KĀLĀMUKHAS OF THE SAKTI-PARIȘAD 105

cold and unable to make the distant pilgrimage (to Kedära) -frees them from all sins (here) ...- may he protect you, the wearer of the crescent moon.

These descriptions are remarkably different from what one would have expected on the basis of the statements of Yamunācārya and Rāmānuja. Certainly these Kālāmukhas do not seem to uphold a doctrine 'in conflict with the Vedas (Veda-virodha),' unless by this Ramanuja merely means in conflict with his own Vedic exegesis. The list of subjects studied at the matha includes nearly the whole of traditional Sanskrit learning with the addition of two slightly less orthodox subjects, Lākula-siddhānta and Pātañjala Yogaśāstra. The eclecticism in the choice of alms reci- pients is astonishing and testifies to the charity and tolerance of the directors of the matha. The sectarian affiliations of all the various classes of ascetics given alms cannot be determined exactly, but the list probably includes Jains, Buddhists, Saivas, Vaișnavas, and perhaps even Ājīvikas. The term kșapaņaka usually refers to naked Jain ascetics although it occasionally refers also to Buddhists. Hamsa and paramahamsa ascetics may be either Vaișņava or Śaivite since these terms seem to denote levels of spiritual advance- ment rather than sectarian affiliation.32 There is some confusion regarding the titles ekadandin and tridandin. G.S. Ghurye believes that the former are Saivite and the latter Vaisnava while A.L. Basham seems to suggest the opposite.33 K.K. Handiqui has clearly shown that the original sources themselves are ambiguous or contradictory.34 Basham further suggests that the compound nagna-bhagna, 'naked and crippled,' may refer to a class of ascetics, possjbly the Ajivikas, who were 'naked and crippled' owing to ritual austerity and initiatory mutilations.35 The comparison of the Southern-Kedära with the original Himalayan mountain points to some connection between these ascetics and the North-west. Several Kālāmukha priests in Mysore bore the name Kāśmīra-pandita, and this fact led A.V. Subbiah to assert that the Kālamukha sect originated in Kashmir.36 This

32See G.S. Ghurye, Indian Sadhus, pp. 72-78, and A.L. Basham, History and Doctrines of the Ājivikas, p. 114. 33Ghurye, pp. 71-72, and Basham, History ... , p. 105. 34See Śrīharșa, pp. 586-88. 35 History ... , p. 105. 36QJMS, VII, 176.

106 THE KĀPĀLIKAS AND KĀLĀMUKHAS

is too sweeping a generalisation, but evidence has since been discovered which explicitly confirms that at least some of these ascetics did migrate from the Kashmir region. This will be discussed below. The successor of Kedāraśakti in the main line of Kedāreśvara pontiffs was Śrikantha. In the 1094 grant he is called Kedāraśakti's 'chief disciple ... of whom what more can be said than that he was himself Läkulīśa in the world, and farther, shone as the equal of omniscience.'37 The Kedäreśvara grant of 1103 and the duplicate grant of 1101-04 from the Brahmeśvara temple at Ablūr contain the following additional verses in praise of Śrīkantha (with some minor variation) :

Of that great ascetic Kedāra, the disciple praised indeed throughout the world, was Śrikantha, abounding in extre- mely pure virtues, of spotless behavior, a very cuckoo (or ring-dove) in the grove of mango-trees that are learned men. Amidst great applause, Śrīkanthadeva, abounding in great virtue, an ornament of great saints, a forehead- ornament of learned people, a very ocean of the science of logic [tarkka-vidyā], firmly fixed his thoughts on the water-lilies that are the feet of the god Hara (Siva), and made the beauty of the goddess of eloquence abide in the charming water-lily that was his mouth, and maintained purity in all his behaviour, and established to the ends of all the points of the compass a brilliant fame like that of (Airävata) the elephant of (the east which is) the quarter of Indra.38

In the Kedāreśvara grant of 1113, as we have noted, Śrīkaņțha is included 'among the Kaļamukhas.' The record continues :

Praised by the learned, the son of Kedārasakti, ever cheri- shing Śrīkantha (Siva) in the lotus of his heart, his holy throat ( śrīkantha) retained the blessed words uttered by the munipati. Understanding the paramātmāgama, skilled in overpowering eloquence, like the purest gold if it had acquired perfume, having placed the lotus feet of Iśvara

37Ed. and trans. Rice, EC, VII, Sk. 94. 38Ed. and trans. Fleet, EI, V, 219, This is the Ablür version.

KĀLĀMUKHAŚ OF THE ŚAKTI-PARIȘAD 107

on his head, self-chosen husband of the wife severe penance, distinguished by all the ācārya qualities, was Śrīkaņțha- yogīśvara.39

The Kedāreśvara grant to Gautama dated 1129 asserts that Śrī- kanțha was 'like a pearl necklace to the throat of Sarasvati, a touchstone for testing the gold of learning, reverenced by the world.'40 Although these descriptions, like those of the Kodiya-matha, tend to run counter to some of the accusations made against the Kālāmukhas by Rāmānuja, there are also several points of agree- ment between his account and the epigraphs. First, both the Kālāmukhas of Rāmānuja and the priests of the Kedāreśvara temple are worshippers of Siva. Furthermore, however great the learning of the Kodiya-matha priests, the essential feature of their faith seems to be personal devotion or bhakti to Siva rather than metaphysical speculation or a' religion of sacrifice and ritual observance. In this respect these priests bear resemblance to their famous opponent who at this time was preaching his bhakti- yoga at Śrīrangam some 250 miles to the south-east. Rāmānuja identifies the Kāpālas, Kālāmukhas, Pāśupatas, and Saivas as the four Saivite orders which follow the doctrine of Paśupati. The association of the Kālāmukhas with the Pasupatas is well-documented. Many Kālāmukha teachers, including Śrī- kaņțha, are identified with Lakulīśa, the famous Pāśupata saint to whom the Pāsupata-šūtra is traditionally ascribed. Lākulasid- dhänta, the Doctrine of Lakula, is one of the chief subjects studied at the Kodiya-matha, and most of Srikantha's successors are either identified with Lakula or said to follow the Lākulasiddhānta or Lākulāgama. Other South Indian inscriptions attest to the importance of Lakulīśa to the Kālāmukhas and consequently to the close relation between the Kālāmukhas and Pāsupatas. A nearly identical verse contained in two records from Belgaum District-one from Sirasangi dated A.D. 1148 and one from Nesargi dated 1219-2041-seems to identify the Kālāmukhas as Mahāvratins and Mahāpāśupatas. The names Jñānaśakti and Kriyāśakti,

39Ed. and trans. Rice, EC, VII, Sk. 99. 40Ibid., Sk. 100. 41Ed. Panchamukhi, Karnatak Inscriptions, I, no. 24 of 1939-49, and ed. and trans. J.F. Fleet, JBBRAS, X, 167-298 (No. VI).

108 THE KĀPĀLIKAS AND KĀLĀMUKHAS

which are frequently adopted by Kālāmukha ascetics, are also technical terms for various mystic powers in Pāśupata texts.42 There are a number of post-Gupta statues from northern India which have been identified as representations of Lakulīśa. These usually portray him as a naked yogin with a staff (lakuta) in his left hand and a citron in his right, with his penis erect, and either standing or seated in the padmäsana. At about the beginning of the eleventh century, however, the Lakulīśa cult seems to have shifted its activities to southern India, especially to the Mysore region. The number of statues in northern India declines and the name Lakulīśa suddenly appears in a large number of Kannada epigraphs. Some of the donees in these epigraphs are identified as Kālāmukhas and others as Pasupatas. It is likely that there was an actual migra- tion of Lakulīśa devotees to the Karņāțaka region from various parts of north-western India. As we have noted, V. Subbiah suggested that the Kālāmukhas originated in Kashmir. In support of this theory, however, he could cite only a few inscriptions which mentioned Kālāmukha ascetics named Kāśmīra-paņdita. A recently edited inscription from Muttagi in Bijāpur District dated A.D. 1147 helps to give the theory some added weight.43 The record eulogises a line of Saivite priests who had migrated from Kashmir and had settled at Bijāpur. This priesthood is positively identified as a Kālāmukha one in a grant of A.D. 1074-75 from Bijäpur itself.44 It appears, in fact, that the priests belonged to the Bhujangāvali, another branch of the Sakti-parisad. One of them was named Kāśmīra and another Lakulīśvara. The reason or reasons for the migration from the North of Lakulīśa devotees are uncertain. Missionary zeal, loss of patronage, unsettled political conditions, and famine are all possible factors. Several Kālāmukha ascetics are known to have been peripatetic teachers, but missionary activity in the South cannot explain why the Lakulīśa cult seems to have lost its power in the North. Famine or loss of patronage by one or more royal dynasties are quite strong possibilities but virtually impossible to confirm. Many parts of north-western India were being rocked by the incursions of Mahmud of Ghazni at about this time and this may also have

42The date and teachings of Lakulisa are discussed in detail below, chap. vi. 43Ed. Desai, SII, XV, no. 32. 44Ed. and trans. J.F. Fleet, 'Sanskrit and Old-Canarese Inscriptions,' IA, X (1881), 126-31.

KĀLĀMUKHAS OF THE SAKTI-PARIȘAD 109

been a factor in the move to the South. Two northern sites connected with the worship of Lakuliśa and attacked by Mahmud were Mathura and Somnath. In A.D. 380 the former city was the home of the group of ascetics who traced their descent from Kuśika, a disciple of Lakulīśa. In A.D. 1287 Somnāth was the home of a line of Pasupatas who traced their descent from Gärgya or Garga, another of Lakulīsa's disciples. Kashmir itself, however, was not conquered by Mahmud although he plundered the Kashmir valley in A.D. 1014 and again attacked the place, this time without success, in the following year. In about A.D. 1030 the great Muslim scholar Al-Bīrūnī noted that 'Hindu sciences have retired far away from those parts of the country conquered by us, and have fied to places which our hand cannot yet reach, to Kashmir, Benares, and other places.'45 If many Hindu scholars had fled to Kashmir, many of the more prudent Kashmiri scholars must have decided that the time was ripe to move elsewhere. Among them may have been the Kashmiri Kālāmukhas who travelled to the South where some of their co-religionists had established themselves as early as the end of the eighth century. Several interesting similarities exist between the Somnāth Pāśupatas and the Belagāve Kālāmukhas in addition to their association with Lakulisa. The Cintra praśasti of 1287 records the consecration of five linga temples in Somnāth.46 The Pañcalinga temple in Belagāve belonged to the Kālāmukhas. The five lingas at Somnāth were consecrated by a priest named Tripurāntaka. One of the Kā āmukha temples in Belagäve was dedicated to the god Tripurāntaka. The Somnāth Tripurāntaka's preceptor was Vālmīki-rāśi, a name also found among the early priests of the Mūvara-köņeya-samtati. The Somnāth record describes a pilgri- mage undertaken by Tripuräntaka during which he visited two sites with important Kālāmukha associations-Kedāra in the Himalayas and Śrīparvata in Kurnool District. These similarities show that the Pasupatas and Kālamukhas continued to share a large body of common traditions in addition to having a common base in the teachings of Lakulīśa. A third point of agreement between the statements of Rāmānuja and the Kālāmukha epigraphs is their references to the āgamas. Rāmānuja seems to state that the doctrines of the Saivas and other

45Trans. E.C. Sachau, Alberuni's India, I, 22. 46Ed. and trans. G. Bühler, EI, I, 271-87.

110 THE KĀPĀLIKAS AND KĀLĀMUKHAS

worshippers of Paśupati are 'set forth in the Śaivāgamas.' Many Kālāmukha epigraphs refer to the Lākulāgama. The Mūvara- koņeya-samtati priest Vāmaśakti II, for instance, is called 'an ornament of Lākulāgama,' and the earlier priest Śrīkaņțha is said to understand the Paramātmāgama.47 The Saivāgamas seem to have originated mainly in South India. Tradition enumerates twenty- eight of these texts, but the actual number is much larger. They are usually associated with the doctrine of the Saiva sect proper, Śaiva-siddhānta, but other Śaivite schools developed their own ägamas. Most of these are now lost although there are several extant āgamas of the Vīraśaiva or Lingāyat sect. The term āgama is sometimes used simply as a generic term for Śaivite religious texts or for the Tantras. A fourth point of agreement is the connection with Yoga. Śrīkaņtha is called a 'Lord among Yogins (Yogīśvara),' while Rāmānuja's Kālāmukhas practise various Yoga type rituals. The emphais on Yoga is better exemplified by some of the epithets of Śrīkantha's successors, but nowhere do the inscriptions suggest quite such unusual measures as those listed by Rāmānuja. We have noted above the description of a Kālāmukha priest as a typical Śaivite ascetic in the 1252-53 record from Munavaļli. 48 One additional feature to emerge from the descriptions of Śrīkantha is their emphasis on his knowledge of the science of logic, or tarka-vidyā. This emphasis becomes more explicit in the epithets of some of his successors who are called Naiyāyikas and Vaiśeşikas. The Nyāya-Vaiśeșika system of philosophy is preeminently the science of logic and is closely associated with the Pāśupatas.49 We might also compare the predilection for logical, or rather casuistical, argument of the Kāpālika in Mahendravar- man's Mattavilāsa. Someśvara-pandita-deva, the third in the main line of Kedāreś- vara pontiffs, is the reigning pontiff in three records from Belagāve and in two identical records from Ablür, all dated between 1094 and 1113.50 He may also be identical with the Someśvara who

47Ed. and trans. Rice, EC, VII, Sk. 123. 48See above, p. 6. · 49This association of Nyāya-Vaiśesika philosophy with Pāśupata Śaivism has been convincingly documented by R.G. Bhandarkar, p. 117, and by S.N. Dasgupta in his A History of Indian Philosophy, V, 143-45. 50Ed. and trans. Rice, EC, VII, Sk. 94, 98 and 99; Ed. and trans. Fleet, EI, V, no. A-B.

KĀLĀMUKHAŚ OF THE SAKTI-PARIȘAD 111

was presiding over the Nagareśvara temple in nearby Sūdi between 1060 and 1084.51 The inclusion of verses eulogising him in many records of his successors indicates that he was an important member of this priesthood. The inscription of A.D. 1094 is the earliest of the dated records from the Kedāreśvara temple.52 It records a grant made by the whole town of Belagāve to Someśvara-paņdita-deva, 'the ācārya of the god Nakhareśvara of Tävaregere in the southern quarter' of the town, 'for the service and decorations of the god, for repairs to the temple, for gifts of food to the students and ascetics there.' Somevara is called the pupil of Śrikantha and the possessor of the yogic virtues of yama (restraints), niyama (disciplines), svādhyāya (repetition of the scriptures to one's self), prāņāyāma (control of respiration), pratyāhāra (ability to free sense activity from the domination of external objects), dhyāna (meditation), dhāraņā (concentration), maunānusthāna (constant silence), japa (incantation or murmured prayer), and samādhi (yogic ecstasy).53 These are more or less standard yogic virtues and exercises, and the list is similar to that given in Patañjali's Yogasūtra ii. 29.54 We have seen that Patañjali's sūtras and other Yogaśāstras were included in the curriculum of the Ködiya-matha. The present record further claims Someśvara proficient in siddhānta (doctrine), tarka (logic), vyākaraņa (grammer), kāvya (poetry), nāțaka (drama), Bharata ( ?= music), and 'many other branches of literature and learning.' The term siddhānta may refer simply to philosophy in general; to Saiva-siddhänta, the doctrine of the Saiva sect; or, most likely, to the Lākula-siddhānta taught at the Ködiya-mațha. The duplicate inscription of A.D. 1101-04 found at the Basaveśvara temple at Ablür reveals that the original name of the temple was Brahmeśvara, after the name of the official who built it.55 According to this record, the village of Muriganahalli was given to the temple by the dandanāyaka Govindarasa in the year 1101. On this occasion Govinda washed the feet of Someśvara, the disciple of Śrīkaņtha

51Ed. and trans. Barnett, EI, XV, nos. F-I. 52Ed. and trans. Rice, EC, VII, Sk. 94. 53 Most of these terms are analyzed in detail by M. Eliade in his Yoga ... , pp. 47- 100. We have used his translations for some of the terms. 54He lists yama, niyama, āsana, prāņāyāma, pratyāhāra, dhāraņā, dhyāna, and samādhi. 55Ed. and trans. Fleet, EI, V, no. A-B.

112 THE KĀPĀLIKAS AND KĀLĀMUKHAS

who was the disciple of Kedāraśakti. Someśvara is described as follows :

Some people are learned in logic [tarka], and some can impart the knowledge of well-chosen speech; some are acquainted with the dramas, some are conversant with good poetry, and some know grammar [vyākaraņa] : there are none (others) who know all of these; but the learned Someśvara, indeed, the sinless one, the leader of the Naiyā- yikas, knows them all. A very season of Caitra (i.e. a very month of spring) to (develop the fruit of) the mango-tree that is Akalanka,-a very cool-rayed moon to (bring the full tide to) the ocean that is the Lokāyatas,-a very guardian elephant of that quarter of the region which is the Sāmkhya-doctrine,-a very pearl-ornament glittering on the white throat of the woman who is the Mīmāms,-a very hot-rayed sun to (close) the water-lilies (blooming at night) that are the Buddhists,-the logician [tārkika], the learned Someśvara, the leader of the Naiyāyikas, attained greatness.56

These two verses are repeated in reverse order and with some minor variation in the Kedāreśvara temple record of A.D. 1103.57 The references to Akalanka (probably the famous Jain logician by that name), the Lokāyatas, the Sāmkhya-doctrine, and Mīmāmsā, as Fleet notes, are confusing since Someśvara would be expected to oppose rather than to support them. Fleet thinks there may be 'some hidden second meanings,' but we feel they are probably correct as rendered. If they are correct, then the reference to the Buddhists must also be a positive one. The water-lilies (nīrējāta) of this passage are, in fact, probably day-blooming, not night- blooming. Someśvara would then be the sun that opens the water- lilies that are the Buddhists. The object of the whole passage is to announce Someśvara's mastery of all philosophical doctrines. We have noted that commentaries to the 'six systems of philosophy' including the Bauddha system were said to have been composed at the Ködiya-matha. Vācaspati Miśra provides a precedent for this. He wrote treatises on each of the orthodox systems of philosophy

56Ibid., pp. 219-20. 57Ibid., p. 219.

KĀLĀMUKHAS OF THE ŚAKTI-PARIȘAD 113

with the exception of the Vaiśesika. Similarly, Sāyaņa-Mādhava, in his Sarvadarsana-samgraha, temporarily adopts the views of his opponents for the purpose of explication. This rather lengthy eulogy to Someśvara continues with a series of rhyming-compound epithets which also appear in the Belagve record of A.D. 1103. The first of these epithets merely repeats the yogic virtues mentioned in the 1094 grant with the exception of prāņāyāma and pratyāhāra. His praise continues :

He who is gracious to learned men; he who is a very sun to (open) the great cluster of water-lilies (blooming in the daytime) that is the Nyāyaśāstra, and who is a very autumn- moon to bring to full tide the ocean of the Vaiseşikas; he who is a very ruby-ornament of those who are versed in the Sāmkhyāgama, and who is a very bee on the water-lilies that are the feet of his teacher; he who is a very spring to the grove of mango-trees that is the Sabdaśāstra, and who has given new life to the Lākulasiddhānta by the development of his wisdom; he who is a very stream of the river of the gods in unequalled reasoning, and who has made the assembly of his disciples to prosper by the favour of the counsel given by him; he who is a very ocean to (receive) the stream of the great river that is the Sāhityavidya, and who has quite satisfied the god Parameśvara (Śiva) with the unbroken flow of his devotion; he who is the sole abode of the virtues of blameless and spotless penance, and who has delighted the whole circuit of the earth with the moonlight that is his fame.58

The Belagäve record of 1103 is a grant to the temple of Daksiņa- Kedāreśvara made by the same dandanāyaka Govindarasa for incense, lights and offerings to the god while washing the feet of Someśvara.59 In addition to the above epithets, he is also called 'the äcärya of the temple of the southern Kedāreśvara of the Tāvaregere of Baļļigāve.' Thus it appears that Someśvara was at the same time the ācārya of the Nagareśvara temple, the Brahmeśvara temple, and the Daksina-Kedaresvara temple (unless he had by this time given up the former positions).

58Ed. and trans. Fleet, EI, V, 220. 59Ed. and trans. Rice, EC, VII, Sk. 98.

114 THE KĀPĀLIKAS AND KĀLĀMUKHAS

In A.D. 1113 Govinda made another grant to the temple of Dakşiņa-Kedāreśvara for 'sandal, flowers, incense, lights, offerings and all manner of services, and for the food of the ascetics and others there.'60 The inscription opens with an invocation to Siva as Lakulīśa, who is 'the heart of Brahma shining as a stone on which is inscribed the säsana of the Vedas which extol the abode of Viśvanātha.' This inscription, as we have mentioned, identifies the ascetics Kedāraśakti, Śrīkaņțha and Someśvara as 'Kāļamukhas.' Someśvara is called not merely the disciple, but the son of Śrīkaņtha, and is extolled in another series of rhyming compounds which includes one or two epithets found in the earlier records. Most of the epithets are of little interest save one, listed twice, which identifies him as a 'distinguished Särasvata.' In all likelihood the term Särasvata designates the caste of Someśvara although the generic sense, 'a learned man,' may also be implied. The Sārasvatas were and are a Brähmana caste resident chiefly in Punjab and Sind but also prominent in both Kashmir and Mysore. The identification of a Kālāmukha priest as a member of this caste shows that at least some, and probably most, of the Kālāmukha priests claimed Brähmana status and also tends to confirm the connection of the Kālāmukhas with the Northwest and Kashmir. It seems probable that a good number of the present day Särasvatas of Mysore are descended from northern migrants including the Kālāmukhas. A famous member of the Sarasvatas, the poet Bilhana, migrated from Kashmir to the court of Vikramāditya VI, a patron of the Kälāmukhas, in the second half of the eleventh century.61 It is not inconceivable that Bilhana himself was in some way allied with the Kālāmukhas. An ascetic bearing the name Someśvara was attached to the gods Nagareśvara and Acaleśvara in Sūdi, a village in Dharwar District not too far from Belagäve, at about this time. He is mentioned in grants dated A.D. 1060, 1069-70, 1075, and 1084.62 He may plausibly be identified with the Mūvara-koneya-samtati priest, but this identification cannot be confirmed since none of the Sūdi records mention any of his preceptors or the Śakti-parișad. The Nagareśvara temple is known to have contained Kālāmukhas, however, from this unusual statement from the record of 1060 :

60Ibid., Sk. 99. 61See V.S. Pathak, Ancient Historians of India, pp. 56-57. 62Ed. and trans. Barnett, EI, XV, nos. F-I.

KĀLĀMUKHAŚ OF THE ŚAKTI-PARIȘAD 115

If the Goravas [Saivite ascetics of the monastery] who are Kālāmukhas should not be devout, if they should be so neglectful that the company of fair women [vara-kāntā- samkulam] should not come for three days for enjoyment, or if all the students should fail to study actively always, the worthy superintendents must never allow them to stay.63

Here its seems that the Kālāmukhas were responsible for the upkeep and management of the temple under the overall supervision of some sort of board of directors who were perhaps government officials. The company of fair women must be the 'public women' (sūļeyar, i.e. devadāsīs), who figure prominently among the donees in this grant. The Sūdi records praise the ascetic called Someśvara in terms which closely mirror the praises of the Ablūr and Belagāve inscrip- tions. The record of 1060 says :

'O thou whose lotus-feet are scarred by the rubbing of the crest-jewels of all monarchs, crest-jewel of Vaiśeşikas, sun to the lilies of Naiyāyikas, excellent in mastery of Sāmkhya, a Brahman in grammatical science, who is peer to thee?' On this account the great ascetic Somēsvara, a worthy Gotra-trāsi [family- or mountain-shaker = Indra] to Mīmām- sakas, has become renowned on earth.64

The record of 1075 adds :

A primal Buddha to the Buddhist, a primal Jina to an Akalańka, an Akșapāda (Gotama) to the student of logic [pramāņa-mārga], a Kaņāda skilled in discrimination of all meanings to the student of (the science of) the soul, and likewise a Jaimini indeed to the student of (scriptural) texts, a Brhaspati to the student in the realm of grammar: thus was the master of (the temple of) Nagareśvara renowned.65

He is further said to practise the yogic virtues of yama, niyama, svādhyāya, dhyāna, dhāraņā, maunānușthāna, japa, and samādhi

63Ibid., p. 93. 64Ibid .. p. 92. 65Ibid .. p. 99.

116 THE KĀPĀLIKAS AND KĀLĀMUKHAS

and to favor the Lākulāgama. He was 'a royal swan in the lake of Sāmkhya doctrine, an ear-jewel of the lady of Nyāya doctrine, a crest-jewel of Vaiśeșika doctrine.' The similarities between this description and that in the 1101-04 grant from Ablür are too many to be ignored, especially his characterisation as the master of all the rival religious doctrines. We are inclined, therefore, to accept his identity with the Someśvara at Ablūr and Belagave. Someśvara is further praised in several records of his successors. The A.D. 1129 Kedāreśvara epigraph lists the entire line of ascetics from Kedäraśakti to Gautama but claims that 'the fortune of the Kedāra temple was planted, as if a tree of plenty for the world, through Someśvararya.'66 The 1156 epigraph of Vāmaśakti II, disciple of Gautama, contains a nearly identical statement.67 The much defaced grant of c. 1164 mentions Someśvara as a disciple of Śrīkaņtha.68 Someśvara-deva begins the list of ācāryas in the Trikūteśvara temple inscriptions from Gadag of the years 1191 and 1192.69 The inscription of 1192, as mentioned above, adds the significant title 'Kāļamukha-ācārya' to his name. The main line of the successors of Someśvara seems to have passed from Vidyābharaņa, also called Vādividyābharaņa, to Vāmaśakti I and Gautama-muni. These three as well as their three predecessors all appear in the 1129 Kedāreśvara inscription. The description of Vidyābharaņa makes clear what must have been the true attitude of these ascetics to the rival creeds of Bud- dhism, Mīmāmsā, and Syādvāda or Jain scepticism :

After that, the equal of the celebrated Bhārabhūti [unidenti- fied], was celebrated that fortunate munipa's [Someśvara's] younger brother, Vidyābharaņa, a faultless ornament of learning, an ornament of the lady fame. A thunderbolt in splitting the great boulders the Bauddhas, a lion in tearing open the frontal lobes of the elephant the Mimāmsa creed, a sun to the cluster of water-lilies the Syādvāda,- shines

66Ed. and trans. Rice, EC, VII, Sk. 100. 67Ibid., Sk. 104. 68Ibid., Sk. 108. 69Ed. F. Kielhorn, 'Gadag Inscription of the Yadava Bhillama,' EI, III (1894-95), 217-220, and ed. H. Lüders, EI, VI, 89-97. The second inscription was earlier edited and translated by J.F. Fleet, 'Notes on Inscriptions at Gaddak in the Dambal Tālukā of the Dharwad District,' IA, II (1873), 296-303.

KĀLĀMUKHAS OF THE ŠAKTI-PARIȘAD !17

Vidyābharaņa, a true ornament and muni of the Naiyyāyikas. 70

The record then states that Vidyābharana 'made over the business of the matha' to his senior disciple Vāmaśakti-munīśvara in order to devote all of his time to the cultivation of learning. Nonetheless, it was to Vidyābharaņa that the Cālukya king Someśvara III came to make a grant to the temple in 1129. After the details of the grant-a gift of the village of Tadavanale for repairs to the temple and for worship of the god-the inscription continues with an unusual passage which relates how 'Vidybharana, despising it [the gift] as being a cause for the destruction of the various pleasures of learning and the happy state of yoga, made it over to his own world-renowned senior disciple Gautama-muni, with the headship of the matha.' The section closes telling how the tree which is the Kedāreśvara temple was planted by Someśvara, 'threw out branches, was filled with sprouts, blossomed and spread into all the world' under Vämasakti I and bore fruit through the great Gautamācārya. Vidyābharaņa is not mentioned. The invo- cation, however, declares that Vidyābharana has commanded the god Kedäreśvara to protect Gautama, 'a present manifestation of the ancient Gautama-muni.' Evidently something unusual must have happened in the suc- cession of the matha at about this time. Fleet (EI, V, 224) notes that there is nothing in the inscription to explain why both Vāma- śakti and Gautama are called the chief disciples of Vidyābharaņa and nothing to explain 'why Vidyābharana "censured" or came to regret the happiness of having devoted himself to the various delights of learning because it had proved "destructive of stability," and on that account, appointed Gautama to the office of Matha- pati.' The former question is a bit puzzling, but the latter one, at least in the terms given by Fleet, is based on an inaccurate transcription of the text. It is not happiness which is destructive of stability, but the gift which is destructive of Vidyābharaņa's happiness. The main problem seems to be why Vāmaśakti was passed over in favor of Gautama when it came to choosing the successor of Vidyābharaņa. Vidyābharaņa, it should be emphasised, only transferred the 'business of the matha' to Vämasakti. The record does not state that Vāmasakti ever received the actual

70Ed. and trans. Rice, EC, VII, Sk. 100.

118 THE KĀPĀLIKAS AND KĀLĀMUKHAS

title of mațha-pati. Vidyābharaņa seems to have kept this honour for himself, since it is to him that Someśvara III of the Cālukyas went in 1129 to make the grant. Vämaśakti was probably appointed only executive director of the monastery and may have either died before the final appointment of Gautama or else simply have been passed over for some unknown reason. Fleet thinks that the ins- cription was drawn up at some time considerably after 1129 but we cannot see any significant reason why this should be the case. In the 1149 Kedāreśvara grant, Gautama is called the disciple of Vādividyābharaņa-paņdita-deva.71 Vidyābharaņa is given the same expanded title in a grant from the Brahmeśvara temple at Ablūr dated A.D. 1130 and 1144.72 This grant introduces us to a new member of this priesthood named Jñānasakti-paņdita-deva who was 'the disciple of Vādividyābharaņa-paņdita-deva of the Mūvara-koņeya-samtati.' The inscription records how a certain Bammagāvuņda was reminded that the shrine of Brahmeśvara at Ablür had prospered under the protection of his father and grandfather and that he too should make donations to this temple. Bammagāvuņda accordingly became 'inflamed more than ever with a desire for union with the passionate woman that is devotion to the god Siva.' He then mounted a horse and promised to donate as much land as the horse could cover while running at top speed. After this unusual miniature aśvamedha he washed the feet of Jñānaakti and presented the land he had promised. This was in 1130. In 1144 a daņdanāyaka named Mallibhāvarasa made another grant to the Brahmeśvara temple. Both grants were 'preserved' by Bammagāvuņda and the great saint Jñānasaktideva. A Jñānaśakti is named as the priest of the Tripurāntaka temple in Belagäve in two grants dated c. 1150 and 1159, but it is unlikely that he is the same person.73 Gautama again appears in the 1139 Kedāreśvara temple ins- cription.74 This grant records the establishment of an image of the god Kusuveśvara and the donation of its temple, 'as attached to the god Kedāreśvara,' to Gautama-deva by two sculptors named Bāvaņa and Rāvaņa 'in order to clear an aspersion on their own race of the sculptors.' What this aspersion or fault of their guild

71Ed. and trans. Rice, EC, VII, Sk. 103. 72Ed. and trans. Fleet, EI, V, no. C. 73Ed. and trans. Rice, EC, VII, Sk. 118 and 123. 74IBid., Sk. 112.

KĀLĀMUKHAŚ OF THE SAKTI-PARIȘAD 119

was the record does not say. The two sculptors claim that 'Gauta- mārya, reverenced by a multitude of munis, and the others who were ācāryas of the Kodi-matha were their religious teachers.' In response to this gift, Gautama is said to have himself allotted some land for this temple. The Kedāreśvara inscription of 1149 announces the arrival in Belagāve of a Sāntara feudatory of the Cālukya Jagadekamalla for the purpose of granting the village of Kundūr in the Kodanād Thirty of the Santalige Thousand for the Kedāresvara temple. The feudatory made the grant while washing the feet of 'Vādividyā- bharaņa-paņdita-deva's disciple Gautama-paņdita-deva :

To describe the qualities of the great ācārya of that Kedāra- sthāna,-Gautamārya :- Like bright lamps many munīn- dras, abodes of the highest good qualities, illumined that mațha; after whom the muni Gautama, a pure jewel lamp like a young bud, ever shone in it with world-wide fame, while all the world, folding their hands, addressed him as Jīya. 75

Gautama is also mentioned in a few of the numerous grants of his successor Vāmaśakti II, namely the Kedāreśvara grants of 1156, 1162, 1168, 1179, and 1193,76 but these grants add nothing of importance, except those of 1179 and 1193 which call Vāmaśakti the son rather than the disciple of Gautama. It is not clear whether this means spiritual or actual son. Before discussing Vāmaśakti II we must refer to another important disciple of Vidyābharaņa named Siddhānti-candrabhū- şana-paņdita-deva, alias Satyavākya. This priest is the donee in three late 12th century grants found in the Trikūteśvara temple at Gadag in the Dharwar District. These record donations to this temple by the Yädava king Bhillama in A.D. 1191, by the Hoysala king Vīra-Ballāla II in 1192, and again by the latter king in 1199.77 The inscriptions of 1191 and 1192 are of considerable importance for the political history of the area since they show that Vira-Ballāla II defeated Bhillama and his general Jaitrasimha sometime between these two dates. The 1191 inscription of Bhillama is the only

75Ibid., Sk. 103. 76Ibid., Sk. 104, 102, 92, 123, and 105. 77Ed. Kielhorn, EI, III, 217-220; Lüders, EI, 89-97; and Desai, SII, XV, no. 214.

120 THE KĀPĀLIKAS AND KĀLĀMUKHAS

extant reference to his support of these ascetics, but at least one inscription from Belagäve issued during the reign of Ballāla records a donation to the Kedāreśvara temple. These two grants from Gadag and another dated A.D. 121378 are the only Sakti-parişad records written mainly in Sanskrit. One of the most remarkable features of these Gadag grants is their location so far from Belagäve. Gadag is situated some seventy- five miles north-north-east of Belagāve. Ablūr and Sūdi are only about fifteen to twenty miles from Belagäve. This long distance raises the possibility that Candrabhūsana of Gadag was not in fact a member of the Belagäve line. His rather late dates also suggest this. In the 1129 Kedāreśvara grant Vidyābharaņa must already have been quite old since he had seemingly turned over management of the matha to first one and then a second senior disciple. If Candrabhūșaņa was still alive in about 1199 and was, say, about eighty years old, he would have been only ten years old in 1129. In addition, the Gadag inscriptions mention neither the Sakti- parişad, the Mūvara-kōņeya-samtati, nor the Parvatāvali. None- theless, the evidence in favor of Candrabhūșana belonging to the Belagāvé line is too strong to be dismissed. Most important is the designation of his teacher and teacher's teacher as Vidyābharaņa- deva and Someśvara-deva respectively. The odds against these two being persons other than the Kedāreśvara ascetics are high on the basis of the identity of the names alone. Siddhänti-candra- bhüşana-pandita-deva's name, particularly the ending 'paņdita- deva,' and the explicit mention of his belonging to the Kālāmukha lineage render the identification nearly certain. Although Ablür and Sūdi are much nearer than Gadag to Belagāve, they provide a precedent for the extension of the priesthood to other temples. The large gap in years between the 1129 inscription of Vidyābharana and the 1199 inscription of his disciple Candrabhūsana is rather difficult to account for, but it is not impossibly large if Vidyābharaņa lived until about 1140 or 1150. The fact that he is given the expanded title Vādividyābharaņa-pandita-deva in the 1144 grant to his Ablūr disciple Jñānasakti and in the 1149 grant to his Belagāve disciple Gautama suggests that he acquired the title sometime after 1129. This could have happened any time up to 1144 and he may have been alive even after this date. Vīra-Ballāla's grant of A.D. 1192 includes several Sanskrit verses

78Ed. Desai, SII, XV, no. 159.

KĀLĀMUKHAS OF THE ŚAKTI-PARIȘAD 121

in praise of Candrabhūșana, one of which calls this priest a jamgama, a term later used for the Lingayat priesthood :

There is in the village named Kratuka ([Gadag] the god) Svayambhū called Trikūteśvara. (He is) Siva whose pleasing seat is decorated with the radiance of the crest-jewels of all kings. The ācārya of the sthāna of that (god) is the muni named Siddhānti-candrabhūșaņa-pandita-deva born in the lineage of Kālāmukha ācāryas. (People) regard that same god, (who is known as) Tri- kūțeśvara on account of his three stationary (or mountain) 'lingas (i.e. Kāleśvara, Śrīśaila and Bhīmeśvara), as Catuș- küteśvara on account of that jamgama (priest or moving lińga). (That priest) today becomes indifferent to women and appears as if a brahmacārin like Siva, owing to his close union with Gauri who eternally occupies half of his body. Even when the kula-mountains tremble and the rivers overflow their banks, he 'whose second name is Satyavākya does not abandon the truth. - There is no equal of him not only in (the knowledge of) kāvya, nāțaka, Vātsyāyana, Bharata, rājaniti, etc., but also in all the kathā-siddhāntas. At some time there might be seen the cessation of the waves in the ocean but never (a cessation) in offerings being given to those who eat them because of (his great) share of compassiơn. There is no limit of men whom he continually gratifies not only with food but with gold. medicines, water. cloth. etc. In that sthāna he renovated everything which was ruined and built a new and pleasing pura. To the vicinity of the (temple of the) god he brought a street of public women (veśyā-vīthī) which had been situated elsewhere (formerly). (He then) built a lotus pool filled with water which resembles nectar and a grove of trees which resembles Nandana (the garden of Indra) covered with various flowers and creepers. But why tell more? Whatever was already full here

122 THE KĀPĀLIKAS AND KĀLĀMUKHAS

outside the rampart's stony wall has been rendered perfect. (It is all) his work.79

The grant proceeds to register Vira-Ballala's donation of the village of Hombālalu in the Belvola Three Hundred to the temple. The above passage is written in a pleasing kāvya style and contains several points of interest. First, the priest is said to command the support and respect of royalty. Although he does not claim the rank of rājaguru like other Kālāmukha priests, the list of his intellectual attainments, particularly the inclusion of rājanītī or polity, is well-suited for a royal advisor. The term Vātsyāyana might refer to the famous fourth century commentator on the Nyāya-sūtra, but the context favors an identification with the author of the Kāma-sūtra. Second, Candrabhūșaņa seems to have carried on the tradition of his Belagāve preceptors in acts of charity and munificence and to have instituted much building and expansion of the temple and the area about it. Third, he is shown to have supported the practice of temple prostitution. This settles any doubts about whether or not devadāsīs were employed at Kālāmukha temples. They obviously were, and this practice may have been partially responsible for Rāmānuja's dislike of the Kālāmukhas. Several other epigraphs have been found at the Trikūtesvara temple in Gadag, but most are fragmentary or contain little infor- mation about the priesthood. The earliest, dated A.D. 1002, registers a land grant to the temple of Svayambhū in Gadag.80 The gift was entrusted to the priest Kāļajñāni-vakkhāņi-jīya, the disciple of Koppina-vakkhāņi-deva, a disciple of Pūliya-paņdita. Although there were no priests by these names at the Kedāreśvara temple in Belagäve, the Svayambhu temple is evidently the same as the temple of Svayambhū Trikūțeśvara. A later Koppina-vakkhāņa-deva appears in a genealogy of Parvatāvali ascetics at Rōn, also in Dharwar District.81 Another priest unknown to the Belagave records, Kriyāsakti-pandita, was the head of the matha of Svayam- bhū Trikūțeśvara in A.D. 1102 when a feudatory of the Cālukya king Vikramāditya VI made a grant to the temple.82 This priest

79Ed. Lüders, EI, VI, 96-97. My translation. 80Ed. N.L. Rao, SII, XI, Part I, no. 48. 81Ed. and trans. L.D. Barnett, 'Two Inscriptions from Ron, of Saka 944 and 1102,' EI, XIX, 222-236. 82Ed. N.L. Rao, SII, XI, Part II, no. 15.

KĀLĀMUKHAŚ OF THE ŚAKTI-PARIȘAD 123

must also have belonged to the Mūvara-koneya-samtati. The Belagäve priests evidently never took direct control over the Gadag temple, probably because its distant location made this impracticable. A fragmentary grant to the Trikūteśvara temple made in 1184-85 contains the name Vidyabharana but must have been made to his disciple Candrabhūșaņa.83 Candrabhūșaņa's own disciple Kriyāśakti-paņdita is the donee in a grant to the temple made in A.D. 1213.84 The grant is badly damaged and fragmentary. In 1225 two golden banners were given to the temple by Candauvve, the 'daughter,' i.e. disciple, of Siddhānti-Kriyāśakti-paņdita.85 This must have been the same priest with an expanded title like that of his preceptor. The prestige of the Belagave matha probably was greatest during the reign of Gautama's main line successor Vāmaśakti (II) who claims the exalted rank of rājaguru in seven grants dated between 1159 and 119386 during the reigns of the Kalacuris Bijjala, Someśvara Deva, and Ahavamalla Deva, and of the Hoysala Vira-Ballāla II. That Vämasakti should have held such a post is not inherently unlikely, since there are several other examples of Kālāmukhas assuming this title. Rudraśakti-deva claims it is a grant made in 1249 to the Koțīśvara temple at Kuppațūr in the Shimoga District.87 An earlier priest from this temple named Sarveśvaraśakti-deva is called rājaguru in a grant dated 1070.88 Another Kālāmukha priest named Sarveśvara-deva takes the title in the 1252 grant from Munavaļli in Belgaum District.89 It is by no means certain if any of these priests were ever rājaguru to a more exalted official than local feudatory rulers and officials. Vāmaśakti II first appears in a grant to the Kedāra-mațha dated A.D. 1156 during the reign of the Cālukya Taila III.90 It was issued by a minister of Taila III's feudatory, the Kalacuri Bijjala- devarasa, at the request of 'learned men and attendants' to provide for offerings to the god, for food for the temple ascetics, and for

83Ed. Desai, SII, XV, no. 547. 84Ibid., no. 159. 85Ibid., no. 609. 86Ed. and trans. Rice, EC, VII, Sk. 123 (A.D. 1159), 92 (1168), 150 (1171), 96 (1181), 101(?1181), 97 (1186-87), and 105 (1193). 87Ed. and trans. Rice, EC, VIII, Sb. 270. 88Ibid., Sb. 276. 89Ed. Panchamukhi, I, no. 31 of 1939-40. 90Ed. and trans. Rice, EC, VII, Sk. 104.

124 THE KĀPĀLIKAS AND KĀLĀMUKHAS

repairs to the temple buildings. Following the previously quoted description of the matha, it tells how the fortune of the Kalpa vine which was the Kedara-sthana was raised on the soil of Someś- varāryya's penance and :

well nourished and covered with branches and blossoms through Gautamärya, till its fame has spread over all the world through Vāmaśakti-munīndra. His face a pleasure- house for Vāņī [Sarasvatī], his true heart a pure jewelled house for the Destroyer of Madana [Siva], of worldwide fame was Vāmasakti-paņdita-deva.

He had acquired the ascetic virtues-yama, niyama, svādhyāya, dhyāna, dhāraņā. maunānușthāna, japa and samādhi-and was the 'ācārya of the temple of the god Daksiņa-Kedāreśvara of the royal city Balipura.' In A.D. 1159 Bijjala's feudatory or officer Keśirāja-daņdādhīśa, Kēsimayya, or Keśava-deva, constructed a temple to Keśava 'in the southern quarter of Balipura' on land obtained from Sarveś- vara, priest of the Pañcalinga temple.91 He also established a 'quarter' (pura or Brahmapura) of the town named Vira-keśava-pura and donated it to a band of Brahmans. In trust to these brahmans and others in the town of Belagave he gave :

to the god Jagadēkamallśvara, 2 shares; to the Pañca-Linga god, 2 shares; to the god Kēdāra, 2 shares; to the Brahmans, 38 shares; to the pūjari, 1 share; to the garland maker, 1 share; -altogether 46 shares, in the village of Belvaņi.

This eclectic gift was made in the presence of the officials and prominent people of the 'royal city' Balipura or Belagäve, including Vāmaśakti-pandita-deva and the heads of other temples in the town. All of these priests appear to have been Saivite. One of them named Dharmaśiva-deva, however, is said to be the head priest of the five mathas or Pañca-matha, which, according to the record of A.D. 1129, comprised the temples of Hari, Hara, Kamalāsana (Brahmā), Vītarāga (Jina) and Bauddha. It seems probable then that not only were rival creeds tolerated in Belagave, but that their monasteries were administered by a Saivite priest. Although the

91Ibid., Sk. 123.

KĀLĀMUKHAS OF THE SAKTI-PARIȘAD 125

Ködiya-matha does not appear to have been one of the 'five mathas,' we have noted above the highly syncretic character of the knowledge attributed to the Kodiya-matha pontiff Someśvara. Not all of the Kedāreśvara inscriptions display such tolerance, however, and it is impossible to determine what exactly were the relations between the Mūvara-koneya-samtati priests and their rivals. Perhaps they constantly altered in accordance with the personalities of the reigning priests. Bijjala's 'great minister Kēsimayya-daņdanāyaka' had established and endowed a temple and a Brahmapuri dedicated to Keśava- Vișņu. One would naturally suppose him to be a staunch Vaișņava. The inscription of 1159 claims, to the contrary, that none other than Vāmaśakti, priest of the Kedāreśvara temple, was his ārādhya or family priest and that it was to him that the superintendence of the new temple and Brahmapuri was entrusted :

. Washing the feet of his ārādhya possessed of ... [the ascetic virtues yama, niyama, etc.], kind to the learned, patron of the assemblies of good poets, delighting in gifts of food, gold, virgins, cows, lands, shelter, medicine and many other gifts, an ornament of the Lākuļāgama, skilful in his investigation of all the śāstras and āgamas, son of Gautama- muni, worshipper of the lotus feet of the god Daksiņa- Kēdārēśvara of Balipura,-the rājaguru Vāmaśakti-deva, [Kēsimayya] gave to him that place and the superintendence of the Brahmapuris.92

This again suggests a degree of syncretism in the religion professed by the priests of the Kodiya-matha. Otherwise this passage adds no new information. The title ārādhya is interesting, however, since it was also attributed to the legendary founders of the Virasaiva sect. In 1162 A.D., according to another Kedāreśvara inscription,93 Bijjala came to 'Baļļigāve' in order to subdue the southern region. His feudatory Kasapayya-nāyaka then petitioned him to make a donation to the Kedāreśvara temple and its chief priest, Vāmaśakti II, who is compared to a tree of plenty (kalpa-vrksa) and said to restrain the actions of Desire (Käma) in the world through the

92Ibid., Sk. 123. 93Ibid., Sk. 102.

126 THE KĀPĀLIKAS AND KĀLĀMUKHAS

severity of his penance. In a verse identical to one describing Gautama in the record of 1149, it is claimed that the whole world addressed Vāmaśakti as Jīya. A supplementary grant appended to the same record states that Bijjala donated the revenue of several villages 'for the decorations of the gods Dakşiņa-Kēdārēśvara of the Kodiya-matha, Somanātha, and Brahmēśvara of Abbalūr ... ' As we have seen, the Brahmēśvara temple at Ablūr was staffed by ascetics of the Ködiya-matha. The Somanätha temple cannot be identified, although it too must have belonged to this matha. A much defaced grant to Vāmaśakti which Rice dates c. 1164 A.D.94 briefly mentions Vāmaśakti and the earlier priests Kedāra- śakti, Śrikantha, and Someśvara as well as two unidentified ascetics named Kedāraśaktīśvara and Devavrata. The donor was Mahā- deva-daņdanāyaka, a feudatory or officer of Bijjala and ruler of the Banaväsi province. One day, the record states, 'the famous muni Kedāraśakti delivered a discourse on dharma, his text was this,- "Whoso sets up but one linga, obtains a myriad-fold all the merit described in the āgamas".' Mahādeva-daņdanāyaka then washed the ascetic's feet and set up an image of the god (? Ma) lapeśvara, presumably in the form of a linga. The emphasis on linga worship has already been met in some of the earlier Mūvara-köneya-samtati epigraphs. It connects these Kälāmukhas both with their ancestors, the Lakulīśa-Pāśupatas, and their successors, the Vīraśaivas. In 1168 A.D. the feudatory or officer Kēśava-daņdanāyaka was administering 'all the countries attached to the treasury of the south' under the new Kalacuri king Rāyamurāri-Sovi-Dēva or Someśvara. An inscription in the Kedāreśvara temple states that on this date Keśava made a visit to Belagāve and was impressed by the temple and its chief priest :

[He approached] the ācārya of the temple, the rājaguru-deva; he noted for a long time his pre-eminence in all learning. In grammar, Pāņini pandit; in polity and discernment, Śrībhūșaņācārya; in drama and the science of music, Bharata-muni; in poetry Subandhu himself; in siddhānta Lakulīśvara; in Śiva devotion Skanda ;- thus in the world is he truly styled the rājaguru, the yati Vāmaśakti.95

94Ibid., Sk. 108. 95Ibid., Sk. 92.

KĀLĀMUKHAŚ OF THE SAKTI-PARIŞAD 127

Keśava then petitioned the king and obtained from him a copper śāsana donating a village for the temple. A nearly identical passage appears in an 1181 A.D. grant to the temple by the Kalacuri king Sankama-Deva. Vāmaśakti must have already been a fairly old man by 1171 A.D. since a warrior memorial in Belagäve dated in that year mentions an attack by some robbers on 'the rājaguru Vāmaśakti-deva's grandson.'96 He was still alive in 1193, however, after nearly fifty years as head of the Ködiya-matha. The reference to his grandson shows that not all the ascetics of the monastery were celibate. In about 1181 A.D. Vāmaśakti and his disciple Jñānaśakti-deva (II) undertook to maintain some land, money and three houses in Belagäve, all of which had been granted 'to the dancing girl ... Mallave and the drummer Mādiga as a temple endowment.'97 Temple dancers and musicians are mentioned in other grants to this and other Kālāmukha temples. Together with the devadāsīs they apparently provided entertainment for the lay supporters of the order. Vāmaśakti's disciple Jñānaśakti appears only in this record. The rājaguru Vāmaśakti was still head of the matha in 1193 A.D. when a feudatory of the Hoysala king Vira-Ballāla II donated land to the god, Dakşiņa-Kedāreśvara.98 The grant includes a lengthy eulogy of this priest from which the following excerpt is taken :

The glory of the penance of the priest of that temple, the rājaguru Vāmaśakti-deva,-that great one's possession of all the ascetic virtues ... [yama, niyama, etc.], his being surrounded with disciples devoted to the astānga-yoga which he expounded to them; his lotus feet covered with clusters of bees the large sapphires set in the crowns of friendly kings bowing before him ...; a portable tree of plenty for giving joy to poets, declaimers, orators, con- versationalists and other manner of learned men; able in giving decisions on the meaning of the vedānta, siddhānta, āgama, the six systems of logic, all branches of grammar, pure dharmaśāstra, and all other sciences; skilled in

96Ibid., Sk. 150. 97Ibid., Sk. 101. 98Ibid., Sk. 105.

128 THE KĀPĀLIKAS AND KĀLĀMUKHAS

splitting, as with a thunderbolt, the pētana of the mountains opponent speakers; ... devoted to gifts of food, gold, virgins, cows, lands, and gifts of freedom from fear, of medicine, and all other benefactions; ... worshipper of the holy lotus feet of the god Daksiņa-Kedāreśvara of the immemorial city, the royal city (rājadhāni) Balipura; master of all kinds of spells [visişta-nāna-mantra-sādhakar]; ... His commands on the heads of kings, his fame in the dwellings of the learned, his mind at the feet of the lord of the life of Pärvatī, ...- long may he live, the world- renowned bhratindra [sic] Vāmaśakti. A mountain for the rising sun of logic, an ocean for the jewels good poems, clever in investigating the principles of grammar, foremost in formulating prosody, an only treasure to those who desire instruction in such learning, an expounder of princi- ples ... With those who with cheeks puffed out play all manner of tunes on the flute, with singing women who give forth enchanting songs with clear modulation of the seven notes, and with those who play sweet sounds on drums? bound to their waists,-is he the most skilled in the world in daily performing pleasant dances,-Vāmaśakti-bhra- tīndra. One man composes the aphorisms of a science, another analyses the words, and yet another makes the commentary; but the marvel here is that Vāmaśakti occupies himself alone in both composing, analysing, commenting, and even instructing those who do not understand.

This verbose recitation adds little to our knowledge of Vāmaśakti and the other priests of the Mūvara-koneya-samtati, but it does forcefully underline the main themes of the earlier epigraphs; the priest's possession of the usual yogic or ascetic virtues; his formidable learning in a vast array of subjects, particularly philo- sophy, logic, poetry, and grammar; his influence over kings; his debating and teaching skill; his charity; and his devotion to Śiva. The reference to him as a master of various mantras suggests tantric influence. The unusual term astānga-yoga, may refer to the eight 'ascetic virtues' or to eight parts of the body. It is highly unlikely that Vamasakti himself was 'most skilled in the world in daily performing pleasant dances,' since he must have been an exceedingly old man by this date. Some sort of ritual dancing may have been practised by the priests of this temple, however,

KĀLĀMUKHAS OF THE ŚAKTI-PARIȘAD 129

since Sāyaņa-Mādhava includes song (gīta) and dance (nrtya) among the six oblations (sadangopahāra) to be performed by the followers of Nakulīśa,99 but it is somewhat difficult to reconcile this with the frequent inscriptional emphasis on penance and asceticism. This statement apart, the record gives an attractive picture of the singing and dancing at the temple. Vāmaśakti's name appears in one other epigraph, a grant from the village Hale-Nidnegila in Dharwar District dated 1165 A.D. during the reign of the Kalacuri king Bijjala.100 On this date a Sinda mahāmandaleśvara feudatory of this king made a gift to the Mallikärjuna temple in the village after washing the feet of Vāmaśakti-paņdita-deva, the temple priest of the god Dakșiņa- Kedāreśvara of the town 'Balipura.' The name of the effective head of the Mallikārjuna temple is not given but it seems certain that he was another member of the Mūvara-koneya-samtati. Vāmaakti is not called a rājaguru in this record, a fact which supports the theory that he was never royal advisor to a higher official than the local rulers of the region about Belagāve. After 1193 A.D. nothing more is heard of Vāmasakti or his shadowy disciple Jñānaakti II. The last record of the Mūvara- köneya-samtati at Belagäve is dated 1215 A.D. during the reign of Simhana-Deva of the Devagiri Yadavas. On this date a grant was made to the Kedāreśvara temple by an official named Hemayya- Nāyaka and his wife Ruppabäyi, and was given in trust to the temple ācārya, 'Śrīkaņțha-deva's disciple, the mahābrati Vāmaśakti- deva (III).'101 We have discussed above the possible implications of the term Mahāvratin.102 There remains one other site which may have been connected with the Mūvara-köneya-samtati. This is the Kotīśvara temple at Devasthāna-Hakkalu near or in Kuppațr in the Shimoga District. About nine inscriptions have been found in this place which belong to the period of Kālāmukha occupation. They range from 1070 A.D. to 1280 A.D. In the year 1231 A.D. an official and a general of the Yädava king Singhana-Deva, on orders from their sovereign, donated two villages to the temple in care of the Kālāmukha priests Rudraśakti-deva and his younger brother

99Ed. U S. Sharma, p. 311. 100Ed. A.M. Annigeri, Karnatak Inscriptions, IV, no. 13. 101Ed. and trans. Rice, EC, VII, Sk. 95. 102See above, pp. 73-82.

130 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Sarveśvara-śakti-deva.103 At this time several temples were attached to the Kotīśvara temple including the Siddhanätha temple of Kabbina-Sirivür, the Svayambhū temple of Mulugunda, the Rāmanātha temple of Kiruvade, the Grāmeśvara temple of Abbalūr (Ablūr), the Mūlasthāna Vosavanteśvara of Tiļuvaļļi, the Caitrā- pura of Devangiri, Mūlasthāna of Hānungal, and the Rāmanātha temple of Kuppatūr. If Grāmeśvara is a scribal error for Brahmeś- vara, the Ablür temple of the Mūvara-köneya-samtati, then we might assume that all these temples belonged to this organization. Although there does not appear to be any record of a Grāmeśvara temple at Ablür, this hypothesis is a dubious one. It is best to assume that the Kuppatūr temple of Kotīśvara and its subsidiaries formed a separate complex. The arrangement into a central archdio- cese with various parishes in the surrounding region may well have been a regular feature of the Kālāmukha church. The efficiency of this type of organisation may help to explain the sect's rapid rise to prominence in the region. Similar tactics were successfully employed by Śamkarācārya as well as by the Christian church.

Other Divisions of the Parvatăvaļi The most important of these was the Belleya-santāna. Inscriptions of this group have been found at Hombal and Lakshmēśvar in Dharwar District. The latter town, under its ancient name of Puligere or Huligere, was the capital (rājadhāni) of the province known as the Puligere Three-Hundred. In A.D. 1118 an officer of the Kalyāņa Cālukya king Vikramāditya VI made a gift of income from certain taxes to Sāmavedi-paņdita, the ācārya of the Rāmeś- vara-deva temple in this town. Sāmavedi-pandita is said to belong to the 'Belleya-devara-santāna.'104 In 1123 another feudatory or officer of the same king gave some land to Agastya-pandita-deva who was teaching Kaumāra-vyākarana to the students of the matha attached to this temple. Agastya-pandita-deva was the disciple of Samavedi-pandita-deva of the 'Belleya-samtana.'105 The temple no longer stands. Another epigraph from the same place states that in 1161 during the reign of Bijjala of the Kalacuris, Devarāśi- paņdita, disciple of Amrtarāśi-paņdita and ācārya of the temple of 'Rāmaidēva' or 'Rāmaiya-dēva,' bought some land and donated

103Ed. and trans. Rice, EC, VIII, Sb. 275. 104Ed. G.S. Gai, SII, XX, no. 78. 105Ibid., no. 83.

KĀLĀMUKHAS OF THE SAKTI-PARIȘAD 131

it for worship of the god Muttinakeyya-Indreśvara.106 Devarāśi is said to have belonged to the Kālāmukha-samaya, but it is not certain whether this is the same temple or same line of ascetics. The Grammar of Kumāra (Kaumāra-vyākaraņa) taught by Agastya-pandita-deva was also part of the curriculum at the Ködiya-mațha in Belagāve. The name Sāmavedi-paņdita undoub- tedly indicates that this priest was especially devoted to that Veda, but this fact does not seem to be unduly significant. A single record from Hombal in the Gadag Taluk of Dharwar District contains a good deal more religious information than the laconic grants from Lakshmēśvar.107 It introduces a famous teacher named Bonteyamuni and two generations of his disciples. Bonteyamuni is called a 'Kāļamukha-munīśvara' and a member of the Belleya-santāna and Parvatāvali. His chief immediate disciple was Avadhuta who had the following junior colleagues : Kēdāra- śakti, Mallikārjuna, Mūrujāvi, Nirvāņayōgi, Vāmadēva or Vāma- śakti, Siddhēśakti, Rudraśakti, and Kriyāśakti. Vāmaśakti had three disciples : Bonteyaguru, Mallikārjuna, and Rudraśakti. In 1189 A.D. this last-named priest purchased some land from the local officials of Hombal for the temple of the god Kumāra- Bontēśvara which he had built in memory of his teacher Vāmaśakti. Other gifts were made to the god by local artists and merchants on the same occasion. The most important feature of this lengthy record is the following unique recitation of a series of miracles performed by Bonteyamuni :

Of the powers of Bonteyamuni, the outstanding were: In summer, when the burning heat was attacking him from all the four directions, he stood unperturbed on a slab of Sūryaśiļā in Śrīgiri with his uplifted pleasant face and hands : seeing this the Sun granted him omniscience and told him 'Preach Kartrvda to whomsoever you meet whether they be dēvas, manuşyas, yakşas or rakşasas.' Having received this favour from the sun, Bonteya, full of all powers and dedicated to Sivadharma, returned from the mountain and performed a linga-pratisthāpana during which there was a höma whose fires burned in the skies, and he made many scholarly Brahmans fold their hands in respect.

106Ibid., no. 137. 107Ed. Desai, SII, XV, no. 73.

132 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Moreover, as he was (once) coming to Karahada begging alms, an arrogant man on the way drew his dagger out and waved it in front of him saying 'Receive the alms!' and acted as if he was going to stab him. The dagger melted and collected like water in his bowl. He drank it and went on his way as everyone was amazed at his great powers. Further, taking a round in Kataka he came to the emperor Jayasimha's house and stood in front of him. At that time famous and well-versed logicians of other systems of philo- sophy were there and questioned him how the Kartr he defended could be formless. He stood invisible (became formless) amidst the hundred logicians for a while and made them speechless (answerless), and expounded the philosophy of Isvara-kartrvāda. Thus, by his negation of the other schools of philosophy, he got the title Kartrvāda- cakravarti. Furthermore, (once) as Bonteya was coming on a round in the capital, he was seen by a man who was riding an (intoxicated) elephant in rut and who said : 'Hey! Did you see a bonte (a jumble bundle-a pun on the name of Bonte- yamuni)?' Hearing that the sage threw it (the bonte) on the ground. The elephant came and lifted it up and collapsed to the ground. [This section is obscure]. Further, as he was going round different countries for pilgrimage, one day he was begging alms in Kollāpura and a jōgini (sorceress) offered him molten metal as alms and he received it without evading it and drank it; the jogini was in flames. Further, once when he was attacked by fever, as if to illustrate the moral that even great ones get rid of the effects of their past deeds, he placed the fever apart in a bag and was busily engaged in the meditation of God.108

Several of the places visited by this peripatetic teacher can be identified. Śrīgiri is probably the same as Śrīśaila or Śrīparvata, the famous pilgrim center in Kurnool District. At about this time the Mallikārjuna-śilā-mațha of Śrīparvata was headed by a Kālā- mukha priest of the Parvatāvali named Rāmeśvara-paņdita.109 108Ibid., 11. 16-26. We thank Professor H.S. Biligiri of Deccan College, Poona, for this translation. 109Ed. and trans. Sreenivasachar, HAS, XIII, Part II, no. 7.

KĀLĀMUKHAŚ OF THE ŚAKTI-PARIȘAD 133

Karahada must be the same as Karahāța, the modern Karād or Karhād on the river Krishņa in the southern part of Sātāra District in Maharashtra. During this period Karahāta was the capital of the province known as the Karahāta Four (or Ten) Thousand. It is not clear whether Kataka is to be taken as a proper noun. The word generally means simply 'royal camp' and in this sense might refer to a number of places. If a proper name is meant, the most likely possibilities are Cuttack (Kataka) in Orissa and Dhānyakațaka, sometimes called simply Kadaka and better known as Amarävati in Andhra Pradesh. The latter site is a better choice since it is known to have contained a Kālāmukha temple in the 10th century 110 and is not very far from Śriparvata. Nonetheless, there was no king named Jayasimha ruling over either town at this time. Jayasimha seems certain to be the younger brother of the Kalyāņa Cālukya Vikramāditya VI. As early as 25 June 1077 he was the de facto ruler of the regions known as the Belvola Three Hundred and the Puligere Three Hundred under the nominal overlordship of his older brother. He rapidly extended his control to the Kandur One Thousand, the Banavasi Twelve Thousand and the Santalige One Thousand. His name disappears from ins- criptions after A.D. 1083. Bilhaņa's Vikramānkadeva-carita tells of a quarrel between the two brothers the consequence of which was the defeat of Jayasimha. This must have taken place in about 1083.111 Jayasimha's kataka or royal camp was probably located in this region, which roughly corresponds to the present day Shimoga and Dharwar Districts. The 'capital' (rājadhāni) mentioned in the next section of the inscription might refer to Kalyāņa, the main Cālukya capital; to Belagāve, the capital of the Banavāsi Twelve Thousand and Santalige One Thousand; or to any one of several regional capitals in the Cālukya empire. Kollāpura is, of course, the same as the modern city Kolhäpur in southern Maharashtra. Kollāpura was the center of a Sākta cult of Pāśupata ācāryas dedicated to the goddess Kollāpura-mahālakșmī.112 This may well account for the presence of a jögini there. The dates of Jayasimha's viceroyalty, c. 1077-1083, give the best clue to the period of Bonteyamuni's priesthood. This is slightly

110See B.V. Krishnarao (ed.), 'Tandikonda Grant of Ammaraja II,' EI, XXIII (1935-36), 161-70. l11See G. Yazdani (ed.), The Early History of the Deccan, I, 356-59. 112Ibid., pp. 441-42.

134 THE KĀPĀLIKAS AND KĀLĀMUKHAS

earlier than would be expected by calculating backwards from the date of his great-great-disciple Rudraśakti so we may assume that the great debate took place early in Bonteyamuni's career. The Sun (Sürya) instructed Bonteyamuni to preach Kartrvāda or Iśvara-kartrvādā, the doctrine of Iśvara as Creator. This is, in essence, the doctrine which Rāmānuja attributed to the Kālāmukhas and other worshippers of Pasupati-the dualistic view of Siva as the instrumental but not the material cause of the universe. The term Īśvara-kartrvādā, in the form issarakāranavādi, first occurs in the Māhābodhi Jātaka where an adherent of this doctrine appears as one of the five heretical councillors of King Brahmadatta of Benares.113 Śamkarācārya, in his Brahma-sūtra-bhāșya ii. 2.37, discusses the views of the Iśvara-kāranins. These persons are allied to or identical with the Mäheśvaras who 'maintain that the five categories, viz. effect [kārya], cause [kāraņa], union [yoga], ritual [vidhi], the end of pain [duhkhānta] were taught by the Lord Paśupati (Siva) to the end of breaking the bonds of the animal (i.e. the soul); Paśupati is, according to them, the Lord, the operative [instrumental] cause.'114 These Māheśvaras must be Pāsupatas since the same five categories appear in Kauņdinya's bhāșya on the Pāśupata sūtra and, in a disjointed form, in Sāyana-Mādhava's discussion of the Nakulīśa-Pāśupata system. The extant Pāśupata texts do not make any special effort to give an ontological analysis of the material world, but they do maintain an essentially dualistic world view. In at least one important respect, therefore, the Kālāmukhas appear to have followed the philosophical doctrines of their close spiritual relatives, the Pāśupatas. Śamkarācārya further associates his Māheśvaras with the Vaiseșikas, who teach 'that the Lord is somehow the operative cause of the world,' and with the Naiyayikas.115 We have noted above the frequent epigraphical references to Kālāmukhas as Naiyāyikas and Vaiseșikas. There is other evidence connecting the Päsupatas with these two closely related philosophical systems. Bhäsarvajña, the author of the well-known Nyāya-sāra, also wrote a commentary on the Pāśupata work, the Gaņakārikā. Guņaratna, the author of a commentary on the Saddarsana-samuccaya of Haribhadra, describes the adherents of both the Nyāya and

113 Jātaka, ed. V. Fausboll, V, 228, 238 and 241. 114Trans. G. Thibaut. 115Ibid.

KĀLĀMUKHAS OF THE SAKTI-PARIȘAD 135

Vaiśeşika systems as typical Śaivite yogins and claims that the Naiyāyikas especially adore the eighteen avatāras of Siva beginning with Nakulīśa. The Naiyāyikas, he says, call themselves Śaivas and the Vaiśeşikas, Pāśupatas. Rājaśekhara, in his Saddarśana- samuccaya, speaks of the Nyaya sect of Pasupatas.116 In all likelihood the philosophical position of the Kālāmukhas did not differ a great deal from that of these Nyāya-Vaiseşika Pāśupatas. After receiving his commission from the Sun, Bonteyamuni's first act was to set up a linga. The establishment of lingas was a characteristic Kālāmukha activity, but the worship of Sūrya was not. Sūrya's command to preach Kartrvāda to gods, men, and demons is best interpreted to mean that the doctrine was to be taught to men irrespective of caste. This also agrees with Rāmā- nuja's statements and helps to confirm the historical link between the Kālāmukhas and Vīraśaivas. Bonteyamuni's miracles themselves need little comment. He performed a great penance and a god then rewarded him with divine knowledge and a commission to preach this to the world. He magically turned back attacks by men, animals, witches, and disease. He converted his opponents in a great debate with the aid of a miracle. These are all typical motifs of religious folklore and occur as frequently in western traditions as in eastern although the contexts may differ. These exploits point to a significant 'magical' element in Kālāmukha belief, but this does not necessarily exclude a high degree of philosophic sophistication as a comparison with mediaeval Catholicism easily shows. The curriculum of the Kodiya-matha could not have been mastered by a group of illiterate witch-doctors. There is one other Kälāmukha epigraph which may belong to the Belleya-santana, a grant found in the Harihareśvara temple at Satēnahalli in Dharwar District.117 It was issued in A.D. 1204 during the reign of the Hoysala king Vīra-Ballāla II. Settikavve, the chief lady of Konavatti, is introduced along with her spiritual advisor Haraśakti. His genealogy is given as follows : Pinākapāņi, a Kālāmukha of the Billa-matha; his disciple Sivarāśi, a devotee of Hara; his disciple Amrtarāśi; and his son or disciple Haraśakti- deva, a devotee of Siva' and follower of the doctrine of Lakula. Haraśakti worshipped the god Bhāyilēśvara of the agrahāra village Sūrili but also received a grant of land for the temple of

116S.N. Dasgupta, History of Indian Philosophy, V, 143-45. 117Ed. Annigeri, IV, no. 1.

136 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Harihareśvara in Sātēnahaļļi. If the Billa-mațha can be connected with the Belleya-santäna, these ascetics must also have belonged to this organisation. One other samtati of the Parvatāvali is mentioned in a frag- mentary grant from Gogga in the Shimoga District dated A.D. 1117. A local official made a grant of land to an unnamed temple after washing the feet of 'Rudraśakti-paņdita, disciple of Kriyāśakti- paņdita, promoter of the Kālāmukha-samaya, of the Śakti-paridhi of the Parvvatāvali, and A .. ka-santati.'118 Unfortunately the full name is lost, Rudraakti is given the usual list of ascetic virtues but no other significant epithets. Several other epigraphs mention ascetics of the Parvatāvali but omit the name of the samtati or santāna. The most important of these registers the gift of a village in A.D. 1090 to 'Rāmēśvara-paņdita of the Appa-Parvatavaļi, the head of the famous Mallikārjuna- śilā-mațha of Śrīparvata, an ascetic of the Kālāmukha (creed), and to the succession of masters, his disciples,' by King Dugga- Tribhuvanamalla, the son of the Kākatīya Bētarasa.119 Bētarasa was a feudatory of the Cālukya king Vikramāditya VI. As mentioned above, this grant furnishes the best clue to the meaning of the term Parvatāvali and confirms that Śriparvata was an important Kālā- mukha center. It was found on a pillar at Käzipet in the Warangal District of Andhra Pradesh. Another grant found at Hanam-Konda or Anamkonda in the same district seems to mention the same teacher.120 This states that the father (?) of king Tribhuvanamalla (Bētarasa II) granted Vaijanapali alias Śivapura to Rāmeśvara- pandita. The father of Tribhuvanamalla was Prola I. King Prōla is called 'the best pupil of that Rāmēśvara-paņdita.' The teacher is described as follows :

This Rāmēśvara-paņdita, who pervaded the quarters (i.e. was well-known) with the moonlight of his fame, who was a Mēru mountain for the gems of qualities, the greatest of Siva's devotees, compassionate, the giver of food to the poor, to the wretched, to the mendicants and to the brahmins, and who was well known for his tapas, conducted service at these temples with great interest ...

118Ed. and trans. Rice, EC, VII, Sk. 316. 119Ed. and trans. Sreenivasachar, HAS, XIII, Part II, no. 7. 12Ibid., no. 12.

KĀLĀMUKHAS OF THE ŚAKTI-PARIŞAD 137

[This priest] knew the nectar of the essence of the great cult of the Lakulēśvara āgama and ... spread its practice in the world.121

The record has been provisionally dated at c. 1050. The Sivapura of this last grant appears to have been the name of the settlement at Śriparvata. A grant from Sivapura dated A.D. 1069 states that king Someśvara II of the Kalyāņa Cālukyas granted a village to Sureśvara-paņdita, disciple of Gangarāśi- bhațtāraka, for the satra in the temple of the god Mallikārjuna at Śivapura at the request of his chief queen Kañcaladēvī.122 These priests are described as residents of Śrīparvata or Śrīśaila, posses- sors of the usual ascetic virtues, followers of the Kālamukha-samaya and Lākula-siddhānta, and Naiyyāyikas. At the request of the queen Mailaladevi the king also granted another village to the same priest and to a priest called Devaśakti-paņdita. An earlier grant, dated 1057 and found at a temple near Kottapalle not far from Sivapura, states that Someśvara I, the father of the previous king, came to Śrīśaila with his queen Maiļaladēvī and an official named Ballavarasa and in the presence of the god Mallikār- juna donated a village for the god Svayambhū-Bhairavadeva at Kolla near the western gate of Śrisaila after washing the feet of the same Sureśvara-pandita.123 The wording of the praises of Sureśvara and his preceptor is nearly identical to that in the Sivapura grant. The relation of these priests to Rämeśvara, 'the head of the famous Mallikārjuna-śilā-mațha of Śrīparvata,' is not known. If Rāmeśvara's doubtful date of c. 1050 is correct, Sureśvara cannot have been his preceptor and may have belonged to a rival or subsi- diary mațha. A 1075 record found at Kop in the Bijapur District registers the gift of a village to Tatpurusa-pandita-deva, disciple of Tribhuvana- śakti-pandita-deva of the Kālāmukha Parvatāvali, for the main- tenance of the Svayambhū-Nagareśvara mațha at Vikramapura.124 This must be the town of that name used by Vikramāditya VI as an occasional residence, the modern Arasībīdi in the same district. The donor was Ballavarasa, a feudatory or official of the Cālukya

121Ibid., pp. 55-56. 122Ed. R.S. Sastry and N.L. Rao, SII, IX, Part I, no. 134. 123Ibid., no. 119. 124Ed. N.L. Rao, SII, XI, Part I, no. 116.

138 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Someśvara II. This is probably the person who accompanied Someśvara I to Śrīparvata. A supplementary grant to the god Gövardhaneśvara of Sivapura is attached, but the place mentioned here is probably not the town at Śrīparvata. In 1136 during the reign of the Hoysala king Vişnuvardhana a priest named Kalyāņaśakti-paņdita, a descendant of Īśānaśakti- paņdita-deva of the Kālāmukha Parvatāvali, resided at the Hoysala capital Dorasamudra, the modern Haļebidu in Hassan District.125 Another Iśānasakti from the same line is mentioned in a grant provisionally dated c. 1185.126 The priestly genealogies of these two teachers are confused. In 1179 Vikramāditya, a Sinda feudatory of the Kalacuri Sankama II, made a series of gifts to the sanctuaries of Cāmeśvara and Mäleśvara at Roņ in Dharwar District.127 The donee was Guru- bhakta-deva, a priest of the Parvatāvali of the Kālāmukhas. He was the pupil of Jñānaśakti-deva, who was the disciple of Rudra- śakti-deva, who was the disciple of Koppina-vakhäņa-deva. This is the last of the known Parvatāvali sites.

The Bhujangāvaļi A record from Mattikoțe in Shimoga District dated A.D. 1077 registers a gift to Vareśvara-pandita-deva, disciple of Trilocana- pandita-deva, and chief of the Kālāmukhas of the Sakti-parșe in the Bhujangavali and Ittige-santati.128 The grant was issued by some minor officials, while they were visiting Belagäve, to provide for offerings to the god Mallikärjuna whom they had established in Mariyase (? Mațțikoțe). It is not inappropriate that these officials were visiting Belagāve at the time of issuing this grant. The Kālāmukhas Vareśvara and Trilocana were heads of the Tripuräntaka temple in that town. In A.D. 1096 Sarvadeva, a daņdādhipa of Vikramāditya VI, donated some land to a temple of Sarveśvara which he had built in Belagāve 'as an ornament to the famous Tripurantaka.'129 He had established the temple through the teaching of Vareśvara-munīndra, or Vareś- vara-deva, the disciple of Trilocana-munīndra. Trilocana was

125Ed. and trans. Rice, EC, V, B1. 117. 126Ibid., B1. 119. 127Ed. and trans. L.D. Barnett, EI, XIX, 222-36 (no. B). 128Ed. and trans. Rice, EC, VII, Sk. 292. 129Ibid., Sk. 114.

KĀLĀMUKHAS OF THE SAKTI-PARIȘAD 139

descended 'in the line of the emperor of Kalāmukha [sic] munis, the heavenly seer Kāśmīra-deva.' An earlier, undated inscription mentions a land donation to a dancing girl (sūļege) of the Tripurän- taka temple made by the priest Trilocana-pandita, who must be the teacher of Vareśvara.130 Of the several temples in Belagāve the Tripurāntaka is artistically inferior only to the Kedäreśvara temple. Both are now protected monuments. Around the base of the Tripurantaka is a sculptured frieze illustrating scenes from the Pañcatantra.131 Scattered between these scenes are erotic figures similar to those in the frieze around the plinth of the Laksmana temple at Khajuraho. Various theories have been propounded to explain the presence of erotic sculpture in Indian temples. One of the most prevalent views is that it reflects the influence of tantric ideas. Despite the testimony of Rāmānuja, however, there is little evidence that Kālāmukha worship was in any sense tantric. The peculiar scorn-producing ascetic practices (dvāras) of the Pāśupatas do include a miid form of sexual exhibi- tionism called śrngārana,132 but this does not seem sufficient to account for the sculpture, especially since the commentator on the Pāśupata-sūtra commends celibacy in no uncertain terms.133 More significant, we feel, is the evidence of temple prostitution at many Kālāmukha temples including, as the donation of Trilocana shows, the Tripurāntaka in Belagāve. In our opinion the erotic sculpture of this temple, and the Khajuraho temples as well, is basically profane in character. Like the devadāsis-for whom it might have been a type of advertisement-the sculpture was simply one of the many semi-secular entertainments formerly provided by the temple. This view does not debase the undoubted beauty of the sculptor's art, it simply puts it in a different light.134 It would be useful to learn if erotic sculpture is found on other Kālāmukha temples, especially those which are known to have maintained devadāsīs. The Hoysala style Mallikārjuna temple at Kalsi in the same district contains similar sculpture and may well have been staffed originally by Kālāmukha priests. 135 130Ed. ARMAD 1929, p. 130. 131See Mysore Gazetteer, V, 1282. We visited the temple in March 1966. 132See below, p. 185. 133Ed. R.A. Sastri, pp. 19-21. 134Much the same evaluation has been persuasively argued by N.C. Chaudhuri, The Continent of Circe, pp. 217-20. For a different view see P. Chandra, 'The Kaula- Kāpālika Cults at Khajuraho,' Lalit Kalā, Nos. 1-2 (1955-56), pp. 98-107. 135Visited by us in March, 1966.

140 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Some of the successors of Vareśvara are mentioned in grants made to other temples in Belagäve. In A.D. 1098 the chief priest of the Tripurantaka temple was Caturanana-pandita,136 in 1113 Kriyāśakti-paņdita,137 and between about 1150 and 1180 Jñāna- śakti-paņdita-deva.138 The founder of this priesthood, Kāśmira-deva, may be the donee in a grant made by a local chief to the Mallikārjuna temple at Bēgür-agrahāra in the same district. 139 This chief issued his donation after washing the feet of Kāśmīra-pandita-deva, a supporter of the Kālāmukha-samaya and a member of the Sakti-parise and Bhujanga ... The remainder is defaced. A 1074-75 inscription from the city of Bijäpur mentions a line of Kālāmukha ascetics founded by one Bhujanga-devācārya.140 His immediate disciple was Trilocana and a later member of the line was called Kāśmīra. These names suggest some relation with the ascetics at Mattikote, Belagāve and Bēgūr-agrahāra, but there is no way to confirm it. This Bijapur priesthood also appears in two grants from Muttagi, a village in Bijāpur District.141 Two eleventh century epigraphs from Dharwar District which mention only the Sakti-parse should also be noted here. In 1067 a local official of Adür donated three-hundred palm trees to the Kāļēśvara temple and its ācārya Bāļacandra-paņdita of the Śakti- parșe.142 In 1058 Someśvara I's feudatory Indrakēsiyarasa made a gift to the temple of Jogesvara at Kuyibal headed by the priest Lōkābharaņa-paņdita of the Śakti-parșe.143 Both ascetics are given the usual ascetic virtues but no other information is provided.

136Ed. and trans. Rice, EC, VII, Sk. 106. 137Ibid., Sk. 99. 138Ibid., Sk. 118, 123 and 119. 139Ibid., Sk. 206. 140Ed. and trans. J.F. Fleet, 'Sanskrit and Old-Canarese Inscriptions,' IA, X (1881), 126-32. 141Ed. Desai, SII, XV, nos. 32 and 97. 142Ed. Gai, SII, XX, no. 285. 143Ibid., no. 38.

CHAPTER V

OTHER KĀLĀMUKHA PRIESTHOODS

The Simha-parisad The second of the two known parişads of Kālāmukhas is the Simha-parișad or Lion Assembly. Grants to temples of this parișad have been found in the Guntür District of Andhra Pradesh and in the Bellary, Bijäpur and Gulbarga Districts of Mysore. Although the temples are spread over a large area, they are only five in number and contain a total of only eight relevant inscriptions. It is probable, therefore, that this group was less influential than the Sakti-parişad, or at least received less royal and official support. The Simha-parişad is first mentioned in the undated Tāņdikoņda grant of the Eastern Cālukya king Ammarāja II, who ruled over Vengī and parts of Kalinga between A.D. 946 and 970.1 The grant is written in Sanskrit and registers the donation, by the king, of Tāņdikoņda and three other villages for the god Umāmaheśvara in the temple (devālaya) called Samasta-bhuvanāśraya. The temple was located in the city Vijayavāțī, the modern Vijayawāda or Bez- wäda on the Krishna River about sixty miles from the river's mouth. According to the inscription, the temple was originally established by Vijayāditya Narendramrgarāja, who must be Vijayāditya II (c. A.D. 799-847) of the same dynasty. On the occasion of the summer solstice (uttarāyaņa) Ammarāja II made a gift for the increase of his country, lineage, life, health, and supremacy, in order to provide for temple repairs, bali, naivedya, music (ātodya), and a free feeding house (satra). After delineating the boundaries of the four villages, the inscription praises a line of 'Kālāmukha' priests belonging to the Simha-parisad. The diagram of their spiritual genealogy is given on next page.2 If we calculate twenty years for each generation of teachers, Lakaśipu-Paśupati must have taught over 100 years earlier than

1Ed. B.V. Krishnarao, EI, XXIII, 161-170. 2The editor B.V. Krishnarao interprets the genealogy slightly differently. He makes Kālāmukhendra an alias of Paśupati II and identifies Paupati II and III. The interpretation turns on the meaning of ādi in the expression Kālamukhendrādya- paranāmā.

142 THE KĀPĀLIKAS AND KĀLĀMUKHAS

(1) Lakaśipu alias Paśupati I

(2) Prabhūtarāśi-paņdita I (5) Bhuvanarāśi-muni

(3) Vidyeśvara I (4) Vāmeśvara (6) Paśupati (7) Prabhūtarāśi II II alias Kālamukhendra

(8) Paśupati III

(9) Vidyeśvara II

(10) Prabhūtarāśi III

Prabhūtarāśi III, a contemporary of Ammarāja II. This would be about the time of Vijayäditya II, the founder of the Samastabhu- vanāśraya temple. It is likely, therefore, that Paśupati I was the first head priest of this temple. This also makes him one of the earliest known Kālāmukha priests and an approximate contemporary of the Kalāmukha priest Iśvaradāsa of Nandi Hill in Kolar District, Mysore, who is mentioned in a grant dated A.D. 810.3 According to the text of the inscription, in various ages of the world numerous munīśvaras beginning with Śrī-Lakulīśvara ap- peared, who were self-made forms of Rudra (svīkrta-Rudra- mürtayaḥ). They became self-incarnate (svayambhuvah) on earth as teachers of the path of dharma. In that succession came the 'Kālamukhas,' who were proficient in the Vedas (śruti-mukhyāḥ), Svayambhūs on earth, and worthy of the homage of kings. Today (iha), the record adds, those munīśvaras are the beneficent lords of this sthāna of the Simha-parisad. In the lineage (santati)4 of those 'Kalamukhas,' who were residents of many ancient temples such as that of Amaravateśvara, there arose the munipa Lakaśipu or Paśupati (I), who was the husband of Śri and who understood completely all the āgamas. He fed his holy body (dharma-śarīra) only on water, vegetables, milk, fruits, and roots (text 11. 51-57).

3See below, pp. 160-61. 4This term may be used in a technical sense such as in the term Mūvara-kōņeya- santati. If so, however, the santati is not named.

OTHER KĀLĀMUKHA PRIESTHOODS 143

This passage repeats and confirms many of the facts known about the religion of the Sakti-parişad. The members of the Simha-parișad are 'Kālāmukhas'; they trace their descent from Lakulīśa; they worship Siva; they are proficient both in the Vedas and in the ägamas; and they receive the royal homage owed to world-renowned teachers. Their severe asceticism is emphasized in the reference to Lakaśipu's grainless vegetarian diet, resembling the diet of a vānaprastha ascetic. The temple of Amaravateśvara must be the one located at the famous city Amarāvatī which is about twenty-five miles upstream on the Krishna from Vijayavāțī (Bezwāda). We have noted above that the peripatetic Kālāmukha teacher Bonteyamuni visited a place called Kataka which might be the same as Dhānyakațaka, another name for Amarāvatī. Amarāvatī, Bezwāda and Śrīparvata are all located on or near the Krishna River and are natural stopping points for pilgrims, traders and travellers. The last of Lakaśipu-Paśupati's successors was Prabhūtarāśi III, the heir (yuvarāja) to the fame and prosperity of his guru Vidyeśvara II and a mighty lord for those who seek refuge (prabhur āśritānām, a typical poetic conceit). He is said to have built, in his guru's presence, a beautiful stone devakula and matha of three stories. For this he received from the king three villages and a thousand she-goats. The record ends saying that the king was the donor, the guru Vidyeśvara (II) the composer, and Kațakanāyaka the executor of the grant (text 11. 61-68). This may imply that Vidyeś- vara was the rajaguru of Ammarāja II, but we know from other records that this king was not a patron of Siva alone since he also made donations to some Jain temples in Bezwāda.5 Sometime about the end of the tenth century, or possibly earlier, members of the Simha-parisad established themselves in the Bellary District of Mysore. In A.D. 1045 the Lenka One Thousand, a military clan led by the dandanāyaka Tikaņņa, set up lingas of Nolambeśvara and Leńkeśvara and won permission from King Someśvara I of the Kalyāņa Cālukyas to grant some land for worship of the two gods.6 The gift was made while washing the feet of Someśvara-pandita of the Simha-parișad, who was in charge of a matha in Kögaļi, the capital of the Kögaļi Five Hundred. The inscription registering the grant was found in the Uddibasa-

5R.C. Majumdar (ed.), The Struggle for Empire, p. 139. 6Ed. Sastry and Rao, SII, IX, Part I, no. 101.

144 THE KĀPĀLIKAS AND KĀLĀMUKHAS

vaņņa temple at Morigeri, a village near Kögaļi. In another grant from this temple dated the same year, this clan donated some more land to the god Nolambeśvara while washing the feet of the same teacher in the presence of the god Virūpākșa (Śiva).7 Both grants identify Someśvara as the disciple of Jñāneśvara-paņdita and disciple's disciple of Maleyāļa-paņdita-deva. These priests are given the usual list of yogic virtues, and in the former grant Maleyāla is said to know the true meaning of all the sāstras which issued from the lotus-mouth of Śrī-Lakulīśa.8 This Maleyāla seems to have been an important and well-known religious leader of Simha-parișad since he begins the priestly genealogies in many grants found at other sites in this region. In A.D. 1093 Gangarasa, a mahāmaņdaleśvara of the Cālukya king Vikramāditya VI, donated some land for the god Baleśvara in a temple of the god built by Gangarasa's father at the nearby village of Halagondi.9 The grant was made after washing the feet of the teacher Khaleśvara-paņdita. Khaleśvara's teacher was Śāntarāśi- pandita, whose teacher was Someśvara-pandita, whose teacher was Jñāneśvara-paņdita, whose teacher was Maleyāļa-paņdita. Maleyāļa, Jñāneśvara and Someśvara are obviously the same teachers who appear in the two grants from Morigere. The des- cription of Maleyāļa is in fact copied virtually verbatim from the first Morigere grant. The Halagondi record adds the information that Maleyala was attached to the god Rämeśvara whose temple was presumably located in Kōgaļi. Khaleśvara-paņdita is given the usual yogic virtues. Twenty-two years earlier, in A.D. 1071, Vikramāditya VI donated a village to the god Kalideva of Huvina-hadagalli, another village in Bellary District, at the request of the mahajanas of the place.10 The grant was for service to the god and for feeding the students, servants, singers, and ascetics of the mațha of Lakuleśvara-paņdita, a priest who belonged to the lineage (santati) of Maleyāļa-pandita- deva of Rāmeśvara. This same Maleyāļa seems to head the priestly genealogies of the donees in two grants found at Yewür, a village in Gulbarga

7Ibid., no. 104. 8Another grant from this temple dated A.D. 1064 (ibid., no. 127) records a gift made to a Siva temple while washing the feet of ' ... śvara-pandita-deva of Moringere.' 9Ibid., no. 163. 10Ibid., no. 135.

OTHER KĀLĀMUKHA PRIESTHOODS 145

District of Mysore. On the occasion of a lunar eclipse in A.D. 1077, Ravideva, a Brāhmaņa minister of Vikramāditya VI, petitioned the king to grant some lands for a temple of Svayambhu which Ravideva had built in Yewür.11 The lands were given in trust 'to the ācārya of that place, the fortunate Iśānarāśi-paņdita, a disciple's disciple of Cikkadēva of Miriñje, a disciple of Maleyāļa- pandita-dēva, of a branch-body of the congregation of Eļemela- Simha [Eļemela-Simha-parșan-mandaliya].'12 The list of items for which the income from these lands was to be used gives a pleasing picture of the daily activities of the temple :

[These lands are given] for homage with perfumes, incense, lights, oblations, etc .; for the restoration of things broken ... ; for the set of procession-cloths; for the food and clothing of student-ascetics and scholars reading and hearing [lectures]; for the professors lecturing to them; for the Caitra festival and the festival of the sacred thread, and the enter- tainment of visitors and other such acts of worship; for the homa at the parva of a samkränti, an eclipse, etc., and for bali- sacrifices, etc .; and for the entertainment of poor and destitute Brāhmans and others.13

The record closes with some rules advising celibacy for the inhabi- tants of the monastery : 'Whether they are cāryas of this establish- ment or ascetics, it is not open to any persons except such as observe strict celibacy to abide in the monastery : the villagers, the burghers, and the king in concert, shall expel those who do not observe celibacy.'14 The second record from Yewur was issued in A.D. 1179 during the reign of the Kalacuri king Sankamadeva II and registers a gift of land to Jñānarāśi, the ācārya of the monastery (sthāna) of the god Svayambhu-Somanätha in the town, for maintenance of his establishment.15 Jñanaraśi is said to belong to the spiritual lineage ( samtāna) of Cikkadeva of Miriñje, the disciple of Maleyāļa-

l1Ed. and trans. L.D. Barnett, 'Inscriptions at Yewur,' EI, XII, 268-98 and 329-40 (no. B). 12Trans. ibid., p. 290. The name of Isānarāśi's guru is not given. 13Trans. ibid., pp. 289-90. 14Trans. ibid., p. 290. This may be compared with the rules prescribed for the 'Goravas who are Kālāmukhas' in the A.D. 1060 record from Sūdi quoted above. 15Ed. and trans. ibid., no. G.

146 THE KĀPĀLIKAS AND KĀLĀMUKHAS

deva of the Srīmad-Eļemela-sri-Simha-parșan-mandala. These can only be the two teachers mentioned in the A.D. 1077 grant. This Jñānarāsi seems to be again mentioned in an inscription found at Managoli in Bijapur District.16 This grant, the details of which are lost, was issued during the reign of the Yädava king of Devagiri, Jaitugi I (c A.D. 1191-1200). It says that a munipa named Gauladeva appeared in a lineage of ācāryas. His best disciple was the vratīśvara Maleyāļa-Jñānarāśi, whose son was the munipa Dharmarāi. The name Maleyāļa-Jñānarāśi should probably be interpreted to mean Jñānarāśi of the lineage of Maleyāļa. It is also likely that the same Jñānarāsi is the priest named in a grant dated A.D. 1176 found in a temple of Somanātha at Ingalēśwar in the same district.17 According to this record the Kalacuri king Someśvara donated a village to this temple and entrusted the gift to Jñānarāśi-paņdita-deva, the ācārya of the god Svayambhu- Kedāreśvara of Vijayāpura (Bijāpur) and a member of the Simha- parișad. Here again Jñānarāsi traces his descent from Maleyāļa of the Elemēla lineage and the mandala of the Simha-parișad. Apparently Jñānarāśi was in control of the Svayambhū temples at both Yēwūr and Bijāpur. A collation of all the inscriptions of the Mysore branch of the Simha-parișad yields the priestly genealogy given on next page.

Other Kālāmukha Inscriptions In addition to the records left by the Sakti- and Simha-parișads, there are a large number of Kālāmukha epigraphs which cannot with certainly be said to belong to either organization. These epigraphs are approximately contemporary with and are spread over approximately the same regions as those of the two known parișads. There are an even greater number of similar inscriptions which mention priests or ascetics, who, by their names or by the doctrines they profess, may also have been Kālāmukhas or at least Lakulīśa-Pāśupatas. It would be impractical to examine all of these records, but we will give a rapid survey, by districts, of those in which the donees are specifically identified as Kālāmukhas.

Shimoga District Belagāve, headquarters of the Sakti-parișad, contained at least

16Ed. J.F. Fleet, 'Inscriptions at Managoli,' EI, V, 9-31 (no. D). 17Ed. Desai, SII, XV, no. 129.

OTHER KĀLĀMUKHA PRIESTHOODS 147

Maleyāļa

Jñāneśvara Cikkadeva

Someśvara unnamed (1045 Morigeri)

Śāntarāśi Lakuleśvara Iśānarāśi (1077 (1071 Huvina- Yēwür) Khaleśvara (1093 hadagalli) Halagondi)

Gauladeva

Maleyāļa-Jñānarāśi = ? Jñānarāśi (1179 Yewur, 1176 Dharmaraśi (c. 1195 Ińgalēśwar) Managoļi)

two other Kālāmukha temples besides those dedicated to Tri- purāntaka and Dakșiņa-Kedāreśvara. An inscription found in the temple of the god now known as Kallēśvara registers two grants to the temple of Kalideveśvara-Svayambhu-deva made during the reigns of the Kalyana Cālukya kings Jayasimha II and Vikramäditya VI respectively.18 In A.D. 1024 the former king donated some land, two shops and a flower garden to the sthānācārya of, this god, Śivaakti-pandita, at the request of Kundarāja, the deśādhipati of V(B)anavasa. Sivaśakti is called the foremost of the Kālamukhas and given the usual yogic virtues. The second grant was made by Tambarasa, a governor of Vikramāditya VI. In A.D. 1081 Tambarasa gave some land to the temple in care of Rudraśakti-paņdita, the disciple of Śivaśakti-paņdita. These seem to be the only records of this priesthood, although a Müliga- Śivaśakti-paņdita of the temple of Mūlasthāna Nandikeśvara is the donee in a grant dated A.D. 1019 found in the town.19 This priest, who is also given the usual yogic attributes, may well be

18Ed. and trans. ARMAD 1929, pp. 131-140 (no. 65). 19Ed, and trans. Rice, EC, VII, Sk. 125.

148 THE KĀPĀLIKAS AND KĀLĀMUKHAS

the one mentioned in the Kallēśvara grants. The Pañcalinga temple in Belagäve was also controlled by Kālāmukhas. In A.D. 1036 King Jayasimha II of the Kalyāņa Cālukyas granted some land to Lakulīśvara-paņdita, also called Vādi-Rudraguņa, 'for repairs of the temple of the Pañcalinga set up by the Pāņdavas, the Kāļāmukhi [sic] Brahmacāri-sthāna of Baļļigāve, ... for sandal, incense and offerings for the god, for food and cloths for the students and ascetics.'20 This priest is described as a master of logic and other sciences, an able supporter of the Naiyayikas, and 'a submarine fire to the Bauddha ocean, a thunderbolt to the Mimamsaka mountain, a saw for cutting down the Lokāyata great tree, a great kite to the Sāmkhya serpent, an axe to the tree Advaita speakers, ... a noose of Yama to hostile proud panditas, to Digambara speakers a falling star.' Some of the individual opponents whom he defeated are also named including Tripura Akalanka, Vādi-gharațța, Mādhava-bhațța, Jñānānanda, Viśvānala, Abhayacandra, Vādībhasimha, Vādirāja, and Ayavādi. Several of these can be identified. Vādībhasiņha must be the Digambara Jain Odeyadeva Vādībhasimha, pupil of Puşpasena and author of the Kşatracūdāmaņi and the Gadya- cintāmaņi.21 Vādirāja was another Digambara Jain who wrote his Pārśvanātha-carita in A.D. 1025 during the reign of Jayasimha II.22 Abhayacandra might be the Jain author of the Padmānanda Mahākāvya.23 Mādhava-bhațța might be the Kāvirāja who compos- ed a śleşa-kāvya called Rāghava-Pāndaviya under the patronage of Kāmadeva. A.B. Keith identifies this Kāmadeva with a Kadamba king ruling c. A.D. 1182-97 but notes that R.G. Bhandarkar puts him at the beginning of the eleventh century.24 Vādībhasimha, Vādirāja, Abhyacandra, and Mādhava-bhatța were probably all contemporaries of Lakulīśvara-paņdita. Tripura Akalanka is either the well-known eighth century Jain logician or some later namesake. The names of many of these theologians, as Handiqui points out (pp. 10-11), reflect the fondness for philosophical debate and polemic which characterised the period. A few of Lakuliśvara's successors at the Pañcalinga temple are

20Ibid., Sk. 126. 21See Handiqui, p. 9, and B.A. Saletore, Mediaeval Jainism, pp. 49-54. Saletore attempts to prove that this teacher also bore the names Vadi-gharatta and Ajitasena. 22See Handiqui, p. 9, and Saletore, pp. 43-50. 23See R.C. Majumdar (ed.), The Struggle for Empire, pp. 301-302. 24 A History of Sanskrit Literature, p. 137.

OTHER KĀLĀMUKHA PRIESTHOODS 149

mentioned in grants to other temples in Belagāve. In A.D. 1098 the Pañcalinga was headed by Śrīkanțha-paņdita-deva,25 in 1113 by Kriyāśakti-paņdita,26 between about 1150 and 1159 by Sarveśvara- paņdita-deva,27 and in 1181 by Rudraśakti-deva.28 One of the most noteworthy features of the record of Lakulīśvara is a concluding verse in support of Mahādeva (Siva) and varņāśrama- dharma. Rice translates :

Mahädeva is god, his feet worthy of worship by all the world. The rule enjoined in the three Vedas for the order of castes and āśramas is dharma. Who casts aspersion on these two (statements), on his head will I place my foot in the king's assembly.29

It is difficult to estimate the weight to be given to this defence of social orthodoxy. It is fairly certain that most if not all Kālāmukha priests claimed Brähmana status. This we gather from the 1113 inscription which calls Someśvara of Belagāve a Sārasvata,30 from a few scattered references to the gotras of Kālāmukha priests, and from the common ending to many of their names, '-pandita- deva.' On the whole, however, Kālāmukha inscriptions are re- markably silent on the subjects of caste and class, and in the case of Bonteyamuni of Hombaļ a Kālāmukha priest is instructed to preach to anyone who would listen.31 Furthermore, the extant texts of the Päsupatas, the sect most closely associated with the Kālāmukhas, have virtually nothing to say about the subject, and the Virasaivas, the sect which succeeded the Kālāmukhas, were openly hostile to caste consciousness. Without more specific information it is impossible to determine the dominant attitude of the Kālāmukhas to varņāśrama-dharma. In these circumstances the above verse should perhaps be regarded as little more than a stock imprecatory formula. At least two other religious establishments in Belagāve, the Pañca-mațha and the Senior- or Hiriya-matha, may have belonged

25Ed. and trans. Rice, EC, VII, Sk. 106. 26Ibid., Sk. 99. He seems to be different from the Tripurāntaka priest by this name. 27Ibid., Sk. 118 and 123. 28Ibid., Sk. 119. 29Ibid., Sk. 126. Rice says this verse is from Kumārila-bhatța. 30See above, p. 114. 31See above, p. 131.

150 THE KĀPĀLIKAS AND KĀLĀMUKHAS

to the Kālāmukha order since the names of their priests end with the titles deva and pandita-deva in typical Kālāmukha fashion. 32 One or two other villages in Shimoga District contained Kālā- mukha temples which cannot definitely be connected with the Śakti-parișad. An inscription found in Belagāve itself registers a grant to the god Siddheśvara of Benakanakola.33 The donor issued the grant in A.D. 1039 after washing the feet of the Kālāmukha priest Kriyāśakti-paņdita-deva. A grant of A.D. 1163 from Bandalike commemorates the construction of a stone tower (prāsāda) and a golden pinnacle (kalaśa) for the town's Śiva temple by an official named Māceya-nāyaka.34 This official also set up a linga named Someśvara-after his feudal overlord, the Kadamba Soma-and donated some land for its temple. Māceya's guru was Devaśakti- bratindra, 'an ornament to the face (mukha) of the celebrated Kālāmukhas.' This priest is said to be expert in Vedänta, the eight branches of Yoga, Siddhanta, and the Saivāgamas and to possess the usual yogic virtues. It is also claimed that he received a boon from Aghora, the 'ācārya of the celebrated Hiriya-matha of Bam- maküru.' The chief priests of the Brahmacāri-matha of the Someś- vara temple were named Someśvara-paņdita and Bīreya-jīya. About eleven years later, in A.D. 1174, Māceya built another Siva temple in the town-called Boppeśvara after Kadamba Soma's father Bopparasa-and donated some land to the Mūlasthāna ācārya, the Kālāmukha Kalyāņaśakti-paņdita.35

Belgaum District In this district of northern Mysore the ancient town of Pūli, modern Hūli, seems to have been an important Kālāmukha center. A composite inscription found there registers several grants to a temple of Andhāsura (Siva) which was controlled by a line of Kālāmukha priests.36 The first grant, dated A.D. 1104, was to the ācārya Tatpurușa-paņdita, a disciple of Jñānasakti-paņdita-deva. Jñānaśakti is praised for his knowledge of logic and grammar and given the second name Ekākșara. Many monarchs are said to have offered him homage. His spiritual ancestors were the 'eminent

32See ibid., Sk. 125, 151, 106, 99, 118, 123, 119, and 168. 33Ibid., Sk. 153. 34Ibid., Sk. 242. 35Ibid., Sk. 236. 36Ed. and trans. L.D. Barnett, 'Inscriptions of Huli,' EI, XVIII (1925-26), 170-218 (no. E).

OTHER KĀLĀMUKHA PRIESTHOODS 151

saints of the Kāļāmukha order,' who were noted 'for exalted majesty of learning (and) for severe austerities.' They are named as follows : 'Pūliyadeva ...; after him, Lakulīśadeva; after him, Vakkhāņideva excelling in virtues and the great Yogin Vidyeśāna, versed in all arts and sacred tradition [sarva-kalāgama]; so after him, the distinguished saint Somadeva.' Apparently Somadeva was the teacher of Jñānasakti. A second grant, dated A.D. 1162, registers a donation by Jñānaśakti-deva, the sthānācārya of the god Andhā- sura, for the god's baths and oblations. This may be the same priest or, more probably, a successor. A third grant to the god, made by some leading citizens of Pūli in 1184, does not mention any priest. In 1224 the weavers' guilds of Püli, worshippers of the god Tri- kūțeśvara, made a final gift to Vāmaśakti-deva, the sthānācārya of the Andhāsura temple. An undated record found in Hūli refers to a Jñānaśakti who must be identical with one of the Jñānasaktis of the previous inscriptions.37 He is called 'an rādhya adored by bowing monarchs of demons and men,' and 'an excellent mirror of Kālāmukha (doctrine).' His disciple was Nāgarāśi, whose lay disciple Mādi- Gauda is mentioned in another record.38 This Jñanaśakti and his disciple Nāgarāśi may well be the same as the priests Rirapūli-jñānaśakti and Nata-nāgarāśi included in the genealogy of the Kālāmukha priest Honnayya who taught at Nēsargi in the same district.39 In A.D. 1219-20 an official of Kārtavīrya IV of the Rattas of Saundatti and Belgaum erected temples of Habbeśvara, Māņikeśvara and Siddheśvara in Nēsargi. Honnayya was the priest of these temples. His teacher was Vāma- śakti, the elder brother of Nata-nāgarāśi, who was the disciple of Rirapūli-jñānaśakti. Vāmaśakti's own teacher was Rudraśakti, the disciple of Risīsenga. An important verse in praise of Honnayya seems to identify, or at least connect, the Kālāmukhas with Mahā- vratins, Mahāpāśupatas and Śrotriyas. It may be translated as follows : 'Among the Mahavratins who have become famous, among the Mahāpāupatas who have become famous, among the Srotriyas -among the unlimited groups (who) have become famous. I cause the most just chief of the Kālāmukha (order) to be praised.'40

37Ibid., no. I. A Jñānasakti is also mentioned in undated record no. K. 38Ibid., no. J. 39Ed. and trans. Fleet, JBBRAS, X, 167-298 (no. VI). 40Trans. H. Ullrich in letter to author dated 10 January, 1967. We have discussed some of the implications of this verse on p. 19.

152 THE KĀPĀLIKAS AND KĀLĀMUKHAS

This same verse is found in an earlier grant of A.D. 1148 from nearby Sirasangi, the ancient Risiringapura or Pirisingi.41 The Kālāmukha donee of the grant was Rudrasakti-deva, the ācārya of the town's Grāmeśvara-deva temple. This priest may be the same as Honnayya's teacher's teacher in the Nesargi grant. The ancient name of Sirasangi, Riśiśringapura, is very reminiscent of the earliest priest at Nēsargi, Riśīśenga. An inscription from Munavaļļi introduces a line of Kālāmukha priests who were in charge of several temples in the surrounding region.42 The temples included those of Jagadīśvara in Munīndra- vaļļi (Munavaļļi) itself, Malleśvara of Śrī-Veļugrāme (Belgaum), Kalideva of the great agrahāra Nēsarige, Balleśvara of Gōkāge, Vijayameśvara of Koțțumbāgi in the Halasige Twelve Thousand, and Kalideva of Goļiyahaļli. In A.D. 1252 several plots of land were given to the priest of the Jagadīśvara temple by various prominent citizens of Munīndravaļļi. This priest Sarveśvara, his son Kriyāśakti, and grandson Someśvara are all elaborately extolled. Sarveśvara is said to have gained similarity of form (sāmya-rūpa) with the god Śamkara and to have 'kept himself apart from passion, anger, pride, wealth, error, fear, and avarice.' He bore the distinguished title of Holy Royal Preceptor (rāyarājaguru) and possessed the usual yogic virtues.43 Furthermore, he was a priest :

who was intent upon the six duties of offering sacrifices, conducting the sacrifices of others, studying, imparting instruction, giving presents, and receiving gifts; who delight- ed in all the learning of the Rg-veda, the Sāma-veda, the Atharva-veda, the Yajur-veda, the Vedānta, the six systems of philosophy [sattarka], Grammar, Prosody, the collection and explanation of Vedic words and names, poetry, and the drama; who practised the observances of Vyāsa, Agastya, Durvāsa, Viśvāmitra, Nārada, and other holy saints; whose body was sprinkled with ashes; who wore a small piece of cloth around the loins [kaupīna], and the hairy skin of an antelope; who carried a rosary of Rudrākșas; who

41Ed. Panchamukhi, I, 33-38. 42Ed. and trans. Fleet, JBBRAS, XII, 1-50 (no. IV). Re-edited by Panchamukhi, 1, 71-74. 43They are listed as yama, niyama, svādhyāya, dhyāna, dhāraņā, maunānușthāna, tapas, and samādhi.

OTHER KÄLĀMUKHA PRIESTHOODS 153

preserved [the observances of the Lākulāgama;44 and] who was a very incarnation of the Jangama-linga ... 45

The subjects studied by Sarveśvara are much the same as those taught at the Ködiya-matha in Belagäve. As we have noted above, 46 this passage contains the only significant physical description of a Kālāmukha priest. His costume is that of a typical Saivite ascetic. One other Kālāmukha priesthood in Belgaum District existed at the village of Hadli, ancient Paldala. A Kālāmukha priest named Nyānaśakti (Jñānaśakti), a pupil of Devaśakti-paņdita, donated some land to the god Gavareśvara there in the year A.D. 1084.47 The temple of Mallikārjuna at Saundatti must also have been staffed by priests of this sect since it is connected with the Mallikār- juna shrine at Śriśaila. In about A.D. 1230 a local feudatory named Kesiraja, having three times visited the Śrisaila shrine, built the Saundatti temple for a linga which he had brought back with him from that famous site.48 The priest of the temple was Vāmaśakti, also called Lingayya and Lingaśiva, who was the pupil of Devaśiva, the pupil of an earlier Vāmaśakti.

Bellary District Several inscriptions found in this district record donations to temples staffed by members of the Simha-parisad.49 Other Kālā- mukha temples existed at Chinnatumbalam, Kurgod and Sindigeri, and perhaps at Gudihalli, Kuruvatti and other places as well. At Chinnatumbalam a grant was made in A.D. 1068 to Candra- bhūșaņa-paņdita, a disciple of Anantaśakti-paņdita, the disciple of the Kālāmukha ācārya Nirañjana-paņdita.50 It was for service to the god Dakșiņa-Someśvara of Tumbuļa (Chinnatumbalam). An inscription found at Kurgod registers several grants to a temple of the god Svayambhu there, which had been built by a minister of Racamalla I, the Sinda feudatory of the Kalyāņa

44 L:ākuļāgama same(ma)ya samuddha(ddhā)rarum. Fleet's translation, which is based on the reading ākuļāgama for Lākuļāgama, is less satisfactory. 45Trans. Fleet, JBBRAS, XII, 40. The important term Jangama-linga will be dis- cussed below. pp. 171-72. 46See above. p. 6. 47Ed. G.S. Gai, SII, XX, no. 57. 48Ed. J.F. Fleet, JBBRAS, X, 167-298 (no. VII). 49See above, pp. 143-44. 50Ed. Sastry and Rao, SII, IX, Part I, no. 133. See also ibid., no. 218.

154 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Cālukyas.51 In A.D. 1173 Rācamalla I made a donation for service of the god to the sthānācārya Bāla-Sivācārya, who maintained the Lākulīśvarāgama and the Kālāmukha doctrine (daršana) and practised the usual yogic virtues. Several years later, in 1181, Räcamalla gave a village to the same temple and trustee (here called Bāli-Siva-deva). This gift was 'for the god Svayambhū's personal enjoyment, theatrical entertainment, offerings of food, restoration of worn-out (buildings), the Caitra and pavitra, scrip- tural study, lectures on the Vaiśeșika, class-reading of the Śiva- dharma-purāņa, and charitable gifts of food.'52 The inscription ends with a third grant by the two wives of the minister who had built the Svayambhu temple. They donated some land to the temple while mounting the funeral pyre of their dead husband.53 Another inscription from Kurgod mentions a Kālāmukha priest named Amṛtāśi-deva (? =Amṛtarāśi-deva).54 Two inscriptions dated A.D. 1144 and found near the Malleśvara temple in Sindigeri register gifts for feeding houses for pilgrims.55 Both gifts were entrusted to the priest Nirvāna-deva, who was descended from Vämadeva, also called Erköți-cakravarti, the ācārya of the Svayambhū temple at Muļugunda. Vāmadeva is described as the supporter of the Kālāmukha doctrine (samaya), the possessor of all the usual yogic virtues, and a master of a great many śāstras including grammar, logic, Siddhānta ( ?= Lākula- siddhanta), poetry, two types of drama, Vedic names, rhetoric, śruti (?), smrti (?), Purāņa, itihāsa, Mīmāmsā, and nītišāstra.56 Vāmadeva's disciple was Trilocana-deva, whose disciple was Kumāra-deva, whose disciple was Nirvāņa-deva. In A.D. 1065 an official named Bijjaladeva granted a village to the temple of the god Nolambeśvara at Gudihalli while washing the feet of Divyaśakti-pandita-deva of the Lākula sect, who belonged to the matha of the lineage (santati) of Agastēśvara ( ?= Agastyeśvara)

51Ed. and trans. L.D. Barnett, 'Two Inscriptions from Kurgod,' EI, XIV, 265-84 (no. A). 52Trans. ibid., p. 277. The text reads : 'śrī-Svayambhu-dēvar= amga-bhögam(ga)- ramga-bhoga-naivēdya-jīrņņ-oddhāra-Caitra-pavitra-svādhyāya-Vaiśēșika-byākhyā na-khaņdika-Śiva-dharmma-purāņa-pațhanav= anna-dānav ... ' It is not possible to identify the Śiva-dharma-purāņa. 53Ibid., pp. 273, 277-78. 54See V. Rangacharya, Inscriptions of the Madras Presidency, no. 108. 55Ed. Sastry and Rao, SII, IX, Part I, nos. 235 and 237. 56Ibid., no. 235.

OTHER KĀLĀMUKHA PRIESTHOODS 155

of Śrīparvata.57 In 1111 another gift was issued to the Noļambeśvara temple while the donor washed the feet of Vareśvara-pandita, the disciple of Vamasakti-pandita, who was the disciple of the same Divyaśakti-pandita.58 It seems almost certain that these priests were Kālāmukhas. Another line of priests who were probably Kālāmukhas con- trolled the Abhinava-Someśvara temple in the village Kurivatti.59 The founder of this line was Kaśmira-deva, whose pupil was Someśvara-paņdita, whose pupil was Jñānaśakti-paņdita-deva, whose pupil was Lakulīśvara-paņdita. This last priest was in charge in A.D. 1099 when the two-hundred mahājanas of Kuruvatti and their chief, the Brahman Kālidāsa, made a grant for the service of the god. The official governing the district in which the village was situated was himself a priest named Sureśvara-pandita-deva, the disciple of Vādideva-pandita-deva.

Bijāpur District This district has yielded Kālāmukha inscriptions of both the Simha- and Sakti-parișads. A record in Sanskrit and Kannada from Bijäpur itself contains a lengthy Sanskrit eulogy of the spiritual lineage of the Kālāmukha ācārya Yogeśvara-paņdita-deva.60 In A.D. 1074-75 a dandanāyaka of Someśvara II of the Kalyāņa Cālukyas built a temple of the god Śrī-Svayambhū-Siddheśvara for Yogeśvara and his pupils and donated some land for its upkeep. Yogeśvara is given the following genealogy :

Bhujanga-devācārya or -munipa - Bhaujanga, also called Bhuvana and Trilocana

Bālasūrya-munipa, also called Trailocana Kāśmīra-sūrīśvara Śrī-Vādimahāpralaya-Kālabhairava-paņdita-deva, also called Tatpurușa-munipati and Bhairava-muni Yogeśvara- or Yogīśvara-paņdita-deva

57Ibid., no. 128. 58Ibid., no. 186. 59See ibid., no. 165. 60Ed. and trans. Fleet, IA, X, 126-32.

156 THE KĀPĀLIKAS AND KĀLĀMUKHAS

We have already mentioned the possibility that Bhujanga was the founder or a member of the Bhujangāvali of the Sakti-parișad.61 He is described as 'the tilaka on the face (mukha) of the Kālāmukha (sect),' as 'the leader of the Kālāmukhas,' as 'the crest-jewel of yogins,' and as 'the possessor of the mantra for subjugating the beautiful woman Liberation (mukti).'62 By means of Yoga he assumed a hundred different forms and established twelve lingas of the god Jhañjheśvara, including one at Bijāpur. In his lineage (santäna) many excellent munis appeared who possessed the yogic powers (gunas, = siddhis) of becoming small (animan) and so forth. His disciple Trilocana is said to have been expert in the Vedas and Vedāngas and in the āgama received from Śrī- Lākula. Yogeśvara's preceptor Kālabhairava is depicted as an eclectic sage of the same type as Someśvara-paņdita of the Kedā- reśvara temple in Belagāve :

Through his intelligence that Tatpurusa-munipati assumes the status of being a Bhairava to opponents. His terrifying trident is the Mimamsa. He agitates the hearts of his proud opponents with the sound of his drum which is Sugata (Buddhist doctrine). He has the battle drum (bhaya-krt) of Triņayana, which is Viseșa (i e. the doctrine of the Vaiśeşikas), and the upraised skull of Kāpila (the Sāmkhya doctrine of Kapila). He (causes even) the inner parts of space, the sky and the earth to be deafened by the sound of his huge bell which is Nyāya.63

His disciple was the donee Yogeśvara, 'whose form was purified by actions which were capable of washing away the mud of the Kali age.'64 This mighty ascetic uprooted the tree of Love itself after reflecting on the loathsomeness (bibhatsa) of the net of samsāra. He is further described, in Kannada, as the possessor of the yogic virtues of yama, niyama, āsana, prāņāyāma, pratyāhāra, dhyāna, dhāraņā, and samādhi.65 He was a veritable rājahamsa

61See above, p. 140. 62Fleet, IA, X, 127-28 (my translations). 63Ibid., p. 128 (my translation). Much of Fleet's translation of this passage seems to be incorrect. 64Ibid., (Fleet's translation). 65This is the list given in Yoga-sütra ii. 29, It is slightly different than the one given in most of these inscriptions.

OTHER KĀLĀMUKHA PRIESTHOODS 157

among the clusters of lotuses of the Kālāmukha family (kula) and an ear-ornament of the goddess Sarasvati. He had obtained the excellent grace (prasāda) of the god Trilocana and had capti- vated the mind (citta) of the woman Liberation (mukti). The same line of ascetics beginning with Kāśmīra is praised in two grants from Muttagi in the same district, dated A.D. 1147 and 1158.66 Both register gifts to the temple of Sivalinga-deva in the town, made with the approval of the Kalacuri Bijjala, first as a subordinate of Jagadekamalla II of the Kalyāņa Cālukyas and second as emperor in his own right. The donee of these gifts was the priest Lakulīśvara-vratīśvara or Lakulīśa-vrati. He is said to be proficient in the Lākulāgama and a veritable crest-jewel among those who observe perpetual chastity (naisthikas). His preceptor was the muni Yogīśvara-deva (II), whose preceptor was Vareśvara- deva, whose preceptor was the Yogīśvara-deva, mentioned in the Bijapur grant. The inscription of A.D. 1147 adds the important information that this line of ascetics came to Bijāpur from Kashmir.67 If the first priestly migrant was Bhujanga, he must have left this northern region about the middle of the tenth century.

Dharwar District This district is located directly south of Bijäpur and Belgaum Districts and north of Shimoga District. Not surprisingly it also was a center of Kālāmukha activity. The Sakti-parișad controlled temples at Ablūr, Ādūr, Gadag, Haļe-Nidnēgila, Hombal, Kūyibāl, Lakshmēśvar, Rōņ, and perhaps also at Sūdi and Sātēnahaļļi. Other Kālāmukha temples existed at Kalkēri and Sāmsi. In A.D. 1076 a governor of the Cālukya king Someśvara II donated a village for the Kālāmukha temple of the god Svayambhū Someśvara in Kalkēri.68 The priest in charge of the temple was Devasinga-jīya. In 1144 some leading citizens of the Savasi (Sāmsi) agrahāra made some gifts to the Kālāmukha priest Iśānaśakti- paņdita-deva, the ācārya of the local temple of the god Gavarēśvara.69

Chikmagalur (Kadur) District This district is located just south of Shimoga District. Only two

66Ed. Desai, SII, XV, nos. 32 and 97 respectively. 67See above, pp. 108-109. 68Ed. Gai, SII, XX, no. 49. 69Ibid., no. 112.

158 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Kālāmukha sites have been identified. In about A.D. 1108 a feudatory of the Hoysala mahāmandaleśvara Ballāla I granted some land for a Siva temple at the village Bāņūru. The donee was the Kālā- mukha priest Gīrvāņaśakti-paņdita-deva.70 In 1139 some local officials of Berațiyakere, modern Belțikere, donated some land to the local Kālāmukha priest Dharmarāśi-paņdita.71 Another twelfth century Kalämukha priesthood may have existed at Jammāpura. 72

Chitradurga (Chitaldrug) District This district is situated immediately east of Shimoga District. Kālāmukha inscriptions have been found at Asagoda and Chaduru- goļa. In A.D. 1054 an official of the Pallava feudatory of the Cālukya king Someśvara I granted a village for the temple of the god Svayam- bhū in Asagoda.73 The temple establishment is described as 'a Kālāmukha-sthāna, the Naisthika-vedi-karttara-matha.' The full implications of the latter term are unclear although naisthika obviously refers to the sexual continence of the priests of the matha. The temple is said to have been built by Karttāra, who was evidently a spiritual ancestor of the priestly donee, Trailokya-karttāra- bhattāraka. An official of a feudatory of the Cālukya king Vikra- māditya VI made another grant to the temple in A.D. 1108 while washing the feet of Trailokya's disciple Bhuvana-karttāra-paņdita- deva.74 Three years later another grant was made to this temple and was entrusted to Kälabhairava-deva, a disciple of Dharma- karttāra-paņdita-deva. 75 A fragmentary inscription found at Chadurugoļa records a donation made in the year A.D. 1166 'for the god ... deśvara.' The donee was 'the Kālāmukha-vratin Tejonidhi-pandita-deva's son Sarveśvara-paņdita-deva.'76

Hassan District In addition to the Sakti-parișad priesthood at Haļebidu,77

70Ed. ARMAD 1925, pp. 56-57. 71Ed. and trans. Rice, EC, VI, Kd. 80. 72See ARMAD 1943, pp. 91-99. 73Ed. and trans. Rice, EC, XI, J1. 10. 74Ibid., J1. 12. 75Ibid., J1. 9. 76Ibid., J1. 8. 77See above, p. 138.

OTHER KĀLĀMUKHA PRIESTHOODS 159

there were Kālāmukha temples at Arasikere, Jājūr, Halkūr, Kaņikațțe, and Rājana Sirivūr. At Arasikere some local officials gave several plots of land for the temple of the god Gojēśvara in A.D. 1183.78 The donee was Amrtarāśi-paņdita, the son of Dharmarāśi-paņdita, who was a pupil of Aghoraśakti-paņdita. Aghoraśakti is described as a supporter of the doctrine (samaya) of the Lākulāgama and a member of the Kālāmukha order. An inscription of about A.D. 1195 found in the Kallēdēva temple at Jājūr praises two Kālāmukha priests named Candrabhūșaņa and Śivaśakti, who were ruling a place called Rājavür in connection with the Saiva-sthana of Arasikere.79 They are given the following genealogy :

Gańgarāśi-paņdita

Tribhuvanaśakti Amṛtarāśi

Trailokyaśakti Śivaśakti

Candrabhūșaņa Kalyāņaśakti

Candrabhūșana was expert in the proper characteristics of images and temples and in the rules for the performance of Iśa (Siva) worship. He uprooted opposing doctrines and energetically propagated the Kālāmukha doctrine. A line of Kālāmukha ascetics which extended over at least seven generations was located at Halkūr.80 The first priest was Kēta-jīya whose disciple was Devendraśakti-paņdita. Devendraśakti had a female lay disciple named Dekavve as well as a regular disciple named Rāmaśakti. Rāmaśakti's disciple was Kalyāņaśakti, whose disciple was Vāmaśakti, whose disciple was Mahādeva-jīya, whose disciple was Cikkakavi-jīya, who was alive in A.D. 1177. These ascetics are described as Kälāmukhas who uphold the doctrine (samaya) of the Lākulāgama and worship the feet of the god Rāmanātha. Several grants found at the village of Kaņikatte entrust gifts of land and money to Kālāmukha ascetics. In about A.D. 1158

78Ed. ARMAD 1928, pp. 26-8. 79Ed. and trans. Rice, EC, XIV, Ak. 216. See also ARMAD 1911, p. 45. 80Ed. and trans. Rice, EC, V, Ak. 62.

160 THE KĀPĀLIKAS AND KĀLĀMUKHAS

a donation seems to have been made to two priests called Kālā- mukha-dīkșita and Jagateśvara for service of the god Kamma- țēśvara.81 In about 1189 various prominent citizens and officials gave lands and taxes to a temple of Jagateśvara which they had earlier built in Kaņikatțe.82 The donee was Kalyāņaakti-paņdita, a disciple of Śivaśakti-deva, who was a disciple of the Kālāmukha teacher Nāgarāśi-paņdita. Śivaśakti was the donee in a grant dated A.D. 1152.83 A fragmentary Hoysala inscription from the village of Rājana Siruvūr records some gifts to a temple of Dharmeśvara there.84 A Kālāmukha priest of the Atri-gotra named Rudraśakti gave some money. References to the gotras of Kālāmukhas are rare. A Hoysala record of Vira Ballāla II found at Rämapura registers a grant to what may have been another Kālāmukha temple.85 The donee was Somarasi's son Bammarasi, the head of the town's Mūlasthāna Śiva temple and a follower of the Lākųlāgama.

Kolar and Tumkur Districts Very few Kālāmukha inscriptions have been found in these two districts of south-eastern Mysore. Two grants from Nandi Hill in Kolar District are of considerable importance, however, since they are by far the oldest Kālamukha inscriptions yet discovered. In A.D. 810 Ratnāvali, a queen of the Bāņa chieftain Vidyādhara- rāja, gave some land to a Siva temple that she had built at Nandi Hill, now a well-known hill station.86 The head of the matha on the hill was Iśvaradāsa, the chief disciple of the Kālāmukhya (sic) teacher Kāļaśakti. Iśvaradāsa is described as compassionate towards all beings, devoted to performing good deeds and endowed with the virtues of observing vows (vrata), fasting (upavāsa) and niyama. Four years earlier, in 806, the Rāstrakūța king Govinda III donated a village to this Iśvaradsa, 'the lord of the sthāna on Nandi Hill,' for incense, lamps, perfume, bali, and caru in the temple of Śiva.87 These two records indicate that at least a few Kālāmukha ascetics

81Ibid., Ak. 42. 82Ibid., Ak. 48. 83Ibid., Ak. 52. 84Ed. ARMAD 1940, pp. 143-44. 85Ed. ARMAD 1937. pp. 135-42. 86Ed. ARMAD 1914, pp. 29-30, 35-37. 87Ed. ibid., pp. 30-32, 39-41.

OTHER KĀLĀMUKHA PRIESTHOODS 161

had established themselves in Karnätaka by the end of the eighth century. The ascetics must have originally migrated from somewhere in the North, the home of Lakuliśa and his disciples, but we do not know precisely when they arrived. The early presence of Kālā- mukha monasteries in the Mysore region was probably an important factor in the later migration of Kālāmukha priests from Kashmir during the tenth and eleventh centuries.88 Ratnāvali's grant of A.D. 810 concludes with the following unusual imprecation : 'May he who destroys this incur the sin of having turned Śriparvata upside down, of having cut off the heads of the sages there, of having cut off the heads of a thousand tawny cows and a thousand Brāhmaņas at Bāraņāsi (Benares) and of having killed in Jambu-dvīpa sages and Brāhmaņas versed in the 4 Vedas, 18 pramānas and siddhāntas.'89 The prominent mention of Śriparvata indicates that this site was already an important holy center for the Kālāmukhas. The praising of Brāh- maņas versed in the Vedas, cows and the city of Varanasi emphasizes the relative orthodoxy of these priests' beliefs. A grant of A.D. 1169 found at Karadālu in Tumkur District regis- ters a gift to the temples of Sobbēśvara, Mācēśvara, Bammēśvara, and ' ... śvara.'90 The donee was Bamma-jīya's son Gangarāśi-jīya, 'a moon to the Kälämukhas' and 'the obtainer of a boon from the goddess Śaradā.'

The Kriyāśaktis of Vijayanagar A priesthood the heads of which each bore the name or title Kriyāśakti played an important part in the religious life of the early Vijayanagar empire.91 Many Kālāmukha and Pāśupata priests called themselves by this name and there is little doubt that the Kriyāśaktis of Vijayanagar also belonged to one of these two related sects. The term kriyāśakti-like jñānaśakti, another common Pāśupata-Kālāmukha name-denotes an important concept in Pāśupata theology.92 One of these Kriyāśaktis is said to have

88See above, pp. 108-109. 89Trans. ARMAD 1914, p. 36. 90Ed. and trans. Rice, EC, XII, Tp. 91. 91The most detailed accounts of this line appear in ARMAD 1941, pp. 168-70, and Mysore Gazetteer, ed. C.H. Rao, II, Part III, 1442-44, 1474-78, 1650-54. 92These two saktis are also found in the theologies of Kashmir Saivism (Trika) and Vīraśaivism, but the names do not seem to have been common among the followers of either system.

162 THE KĀPĀLIKAS AND KĀLĀMUKHAS

induced his disciple Mädhava-mantrin to give a village to eighty learned Brähmanas from Kashmir, another fact which suggests a connection with the Kālāmukhas.93 It must be admitted, however, that a few sources imply the existence of a close relation between these Kriyāsaktis and the advaita gurus of the famous Śringeri matha founded by Samkarā- cārya. Vidyāraņya, the famous scholar and Vijayanagar guru, was one of the heads of this matha. A Sanskrit work called Vidyā- raņya-kālajñāna actually claims that Kriyāśakti was the disciple of Vidyāranya and states that these two were revered by the first thirteen kings of Vijayanagar, who were worshippers of the god Virūpākșa.94 An inscription of A.D. 1390 seems to record a grant by Immadi Bukka, son of Harihara II, to a shrine of Vidyāśamkara erected in memory of the guru Kriyāśakti, who had died the previous year.95 Vidyāśamkarācārya was the title of the guru Vidyāraņya's predecessor at Śringeri, Bhärati-Krsna-Tīrtha. Another reading of this record, which is evidently badly edited, concludes that Immadi Bukka made his grant with the permission of, rather than in memory of, Kriyāśakti.96 A grant of Harihara II dated A.D. 1384 states that the king listened to the teachings of both Vidyāranya and Kriyāśakti.97 A grant issued in the year 1403 registers gifts of land both to Kriyāsakti-deva-rāya-vodeyar and to the guru of the Śringeri mațha.98 The Pāśupatas and Kālāmukhas were philosophical dualists and for this reason were regarded with disfavour by advaita theo- logians such as Samkarācārya and Sāyaņa-Mādhava, the author of the Sarvadarsanasamgraha. This latter priest has been identified as either Vidyāraņya himself or his nephew.99 If Kriyāśakti was a Pāśupata, it is highly unlikely that he was Vidyāraņya's disciple or that a temple of Vidyasamkara was set up in Kriyāśakti's memory. On the other hand, there is no need to assume that the two groups were overtly hostile to each other. Relations between the various Hindu sects in the early Vijayanagar empire were generally cordial. One Kriyāśakti was tolerant enough to grant

93Ed. and trans. Rice, EC, VII, Sk. 281. 94See ARMAD 1932, p. 105. 95See ARMAD 1941, p. 169. 96See Mysore Gazetteer, II, Part III, 1652. 97See ARMAD 1941, p. 169. 98See ibid., p. 170. 99Mysore Gazetteer, II, Part III, 1433-42.

OTHER KĀLĀMUKHA PRIESTHOODS 163

land to a temple of Varadarāja (Vișņu) in A.D. 1377.100 This cordia- lity was probably greatest between the Kālāmukha and Vīraśaiva schools since the latter appears to have gradually absorbed the former.101 Some of the royal disciples of the Kriyāśakti priests are in fact claimed by the Vīraśaivas.102 It is possible that the Kriyāśaktis were Viraśaivas, but we feel that the evidence favors their identi- fication as Pāśupatas or Kālāmukhas. The dates of the Kriyāsakti epigraphs extend from A.D. 1347 to 1431, indicating that there were at least two and probably three or four priests by this name. As we have noted, one may have died in the year 1389. Several variants of the name occur, including Kāśīvilāsa-Kriyāśakti (1368), Kriyāśaktyācārya (1378), Vāņīvilāsa- Kriyāśakti (1379), Kriyāśakti-deva (1398, 1399, 1431), Kriyāśakti- guru-munīśvara and Kriyāśakti-deva-rāya-vodeyar (1403), Kriyā- śakti-guru (1410), and Kriyāśakti-deśika (1410). In Mādhava- mantrin's Tātparyadīpikā the author identifies his guru as Kāśī- vilāsa-Kriyāśakti, and Gangādevī does likewise in her Kamparāya- carita. Like several Kālāmukha priests,103 Kriyāśakti is given the title rājaguru in a number of records. In some he appears as the guru of Madhava-mantrin, a minister of Prince Marapa.104 Other records praise Kriyāśakti as the kula-guru or rājaguru of Harihara II, of a governor called Vitthanna Odeyar (1403), of Devarāya I and his son Vijaya-bhūpati (1410), and possibly also of Devarāya II (1429). In the Vīra-Kamparāya-carita he is also called the kula-guru of Kampana II.105 Since the term kula-guru means family preceptor. it is likely that these priests fere held in high esteem by most of the early Vijayanagar rulers. Many inscriptions describe Kriyāśakti as a worshipper of the god Siva in the form Svayambhū-Triyambakadeva. Evidently this was the tutelary divinity of the priesthood. A grant made by Devarāya II in 1429 to some Brāhmaņas headed by Kriyāśakti-guru at Cöļisatțipalli in Kolar District states that this priest belonged

100See ibid .. pp. 1651-53. 101See ibid., p. 1654, and below, pp. 167-72. 102Ibid., p. 1654. 103See above, p. 123. 104Märapa was a brother of Harihara I and Bukka I and governor of the province Araga in the western part of the Vijayanagar empire. 105See Mysore Gazetteer, II, Part III, 1651-52, and ARMAD 1941, pp. 169-70.

164 THE KĀPĀLIKAS AND KĀLĀMUKHAS

to the Kasyapa-gotra and followed the Yajur Veda.106 Since the donated village was renamed Tryambaka-pura, there is little doubt that this Kriyasakti was a member of the same priesthood. Apart from their devotion to Tryambaka, however, little is known about the religious beliefs of these priests. The best source is an inscription of Madhava-mantrin dated A.D. 1368 which registers his gift of the village of Muchchandi in Shimoga District to eighty learned Kashmir Brahmanas.107 The lengthy eulogy of this minister asserts that he, 'through the astonishing favour of his master Kāśīvilāsa Kriyāśakti, a manifest incarnation of Girīśa, gained celebrity as a Śaiva [śāmbhava].' He also 'cleared and made plain the ruined path of the upanisads,' and 'on the advice of the Śiva guru Kāśīvilāsa Kriyāsakti, he worshipped in the manner of the Śaivāmnāya the god of gods embodied in his own favourite linga [ișta-linga], Tryambaka-nātha, by means of daily special ceremonies, and by a number of rites and practices.' The grant to the eighty Brähmanas was made to mark the completion of a 'great Śaiva vow' (mahac-chaiva-vrata) lasting one year which he had undertaken in accordance with the directions of the Śiva- samdhyā. Unfortunately the specific details of this great vow and the other rites and practices are never spelt out. The 'great Saiva vow' is conceivably the Mahävrata of the Kāpālikas although for reasons already stated we prefer not to make this identification.108 The donees in the present grant must have had some connection with Kriyäsakti since they are described as 'pre-eminent by their virtues and the country of their birth [Kashmir], travellers to the farthest point of the Cārāyanīya-aticaraņāmnāya, daily observers of all the rites appointed in the pure Śivāmnāya, ever devoted to the worship of the Astamurti.'109 From a record of A.D. 1347 we learn that Kriyāśakti's disciple Mādhava-mantrin aided Prince Mārapa in the compilation of a work called Śaivāgama-sārasamgraha.110 The minister was also the author of the Sūtasamhitā-tātparya-dīpikā on the Sūtasamhitā of the Skanda Purāņa.111

106Ed. and trans. ARMAD 1941, pp. 157-70. 107Ed. and trans. Rice, EC, VII, Sk. 281. 108See above, pp. 81-82. 109Ibid. Carānīya is a school of the Black Yajur Veda, the Veda followed by the Kriyāsakti of 1429. 110Ed. and trans. Rice, EC, VIII, Sb. 375. 111 Mysore Gazetteer, II, Part III, p. 1444.

OTHER KĀLĀMUKHA PRIESTHOODS 165

Andhra Pradesh and Madras The two chief centers of Kālāmukha activity in what is now Andhra Pradesh were Vijayawāda-Amarāvatī and Śrīparvata. These we have already discussed.112 At least a few other Kālāmukha sites must have existed in the region. An inscription of A.D. 1021 found at Melpādi in Chittoor District registers a grant by some shepherds of the town for ghee for a lamp in the temple of Ariñ- jīśvara.113 The head of the matha of the temple was a priest called Lakulīśvara-paņdita, who was probably a Kālāmukha. A Mahāv- ratin Lakulīśvara-pandita, who was possibly the same teacher, is mentioned in an inscription of A.D. 1068-69 found at Jambai in South Arcot District, Madras.114 Although the Kalamukhas were much less influential in Madras State than in Mysore, Kālāmukha temples existed in Chingleput, North Arcot, Thanjavür (Tanjore), and Tiruchchirappali (Tri- chinopoly) Districts. Tamil inscriptions of A.D. 1127, 1205 and 1231 found in the Tiruvālīśvara temple at Tiruvānaikkōyil in Chingleput District mention the Kālāmukha priests Sailarāśi- paņdita and Ñānarāsi-paņdita. They controlled the kāņi (land- revenue) of the temple.115 A Tamil record of A.D. 926 from Vēdal in North Arcot District refers to a Kalamukha Dasapuriyan of the Harīta-gotra and the Āpastamba-sūtra.116 In A.D. 1123 Gomadattu Aruļāļa Bhațțan, a Kālāmukha, sold some land to a temple at Kōyil Tēvarāyanpēțțai in Thanjavūr District.117 An important Sanskrit inscription from Kodumbāļūr in the southern part of Tiruchchirappalli District shows that the Kālā- mukhas had penetrated into the heart of Madras State by at least the middle of the tenth century.118 The Kodumbāļur chieftain Vikrama-kēsarī, a contemporary of Sundara Cōļa Parāntaka II (957-73), erected there three temples (vimāna-traya) named after himself and his two queens. After enshrining the god Maheśvara he donated the Big Matha (brhan-matha), together with the eleven

112See above, pp. 136-37 and 141-43. 113Ed. and trans. E. Hultzsch and H.K. Sastri, SII, III, no. 18. 114See K.A.N. Sastri, The Coļas (1st edition), I, 603. Sastri summarizes the evidence on the Madras Kālāmukhas in ibid., II, 493-94. 115See ibid., II, 623, 702 and 739. 116See ibid., I, 420, and Rangacharya, Inscriptions ... , II, 1162. 117See K.A.N. Sastri, The Coļas (1st edition), II, 610. 118Ed. and trans. K.A.N. Sastri, 'The Kodumbāļūr Inscription of Vikrama- Kēsarī,' JORM, VII, 1-10.

166 THE KĀPĀLIKAS AND KĀLĀMUKHAS

villages attached to it, to the chief ascetic (yati) of the Kālāmukhā- dāna, Mallikārjuna. Vikrama-Kēsarī also seems to have made provision for the feeding of 50 Asita-vaktra ascetics resident there and for offerings, perfume, incense, flowers, lamps and betel for the service of the god of the three temples. Asita-vaktra, Black-face, is a synonym for Kālāmukha. This tends to show that Kālamukha, rather than Kalamukha, is the correct Sanskrit form of the name. Mallikārjuna's own name may reflect devotion to the god of Śrīparvata. According to the inscription Mallikārjuna was a member of the Ätreya-gotra, a resident of Mathura, a master of the Vedas, and a pupil of Vidyārāśi and Taporāśi. Mathurā might be either of two famous cities-Uttara-mathurā (modern Mathura in U.P.) or Dakşiņa-mathurā (modern Madurai in Madras). The latter city is more probable since it is less than 100 miles from Kodumbāļūr, but the former is also possible since many Kālā- mukhas were emigrating from the North at about this time. Inscriptions found at Paļļimadam in Rāmanāthapuram (Ramnad) District and Tiruvorriyūr in Chingelput District refer to Mahāvratin ascetics who must have been Kālāmukhas. At Paļļimadam some sheep were given for a lamp in the matha of Mahavratigal attached to the Sundarapāņdya-īśvara temple.119 The grant was issued during the reign of Vira Pāndya and has been tentatively dated at about the middle of the eleventh century. At Tiruvorriyur an important matha of Mahavratins was founded or brought to prominence by Vaļabha, a general of Coļa Rājāditya, in about the middle of the tenth century.120 When Rājāditya died in A.D. 948 during the battle of Takkolam, Vaļabha was not at his side. In grief and shame the general went to bathe in the Ganges and resolved to become an ascetic. He returned to the South and entered a cave named after the guru Nirañjana at Tiruvorriyūr. There he obtained enlightenment and devoted himself to the per- formance of the Mahävrata for the sake of the protection of the mațha. He assumed the spiritual name or title, Caturānana-pandita, and, in about A.D. 959, made a gift of some gold to the assembly (sabhā) of Narasimha-mangala for a special service to the god Siva on the day of Dhanisthā, the star of his own birth.

119Ed. SII, XIV, no. 88. 120The story of the founding of the matha is contained in the Sanskrit and Tamil inscription of A.D. 959 edited and translated by V. Raghavan (EI, XXVII, 292-303). Raghavan's introduction gives a complete history of the mațha.

OTHER KĀLĀMUKHA PRIESTHOODS 167

Mahāvratins are mentioned in inscriptions from Tiruvorriyūr dated as early as A.D. 942.121 The priest Caturānana is first referred to in a grant of 957.122 The matha continued under a succession of teachers by this name until at least 1172.123 V. Raghavan notes (EI, XXVII, 300) that the Tiruvorriyūrp-Purānam tells of a Tondai- mān of Kāñci who erected a Siva temple at Tiruvorriyūr and established 500 Siva lingas. He also brought from the banks of the Ganges 500 Brähmaņa Mahāvratins and dedicated several images of Kālī and Bhairava and one of Siva in the form of a teacher of the Mahävratins. Some of these images can still be identified and seem to date from later Pallava times.

Kālāmukhas and Vırasaivas A considerable amount of circumstantial evidence points to the existence of a close historical link between the Kālāmukhas and the Viraśaivas. A definitive analysis of the problem would require extensive research both in the field and in the library. In particular, the voluminous hagiology of the Viraśaivas, mostly written in Kannada and Telugu, would have to be digested and painstakingly compared with the available epigraphic data. We will be content to draw attention to some of the more important clues which have turned up in the course of our investigation of the Kālāmukhas. The early history of the Viraśaivas is buried in a maze of legends.124 The principal early leader of the sect was Basava (Sanskrit vrsabha = bull), a minister of the Kalacuri king Bijjala (c. 1145-67). Vīraśaiva tradition claims that the sect antedates Basava, who was merely a major reformer. Modern authorities disagree about this point, but it seems probable that to most intents and purposes Basava was the founder. Not only is there no epigraphic evidence of the existence of Virasaivas before Basava, but the epigraphic allusions to Viraśaiva activity in the two or three centuries after Bijjala are few and far between. This is not to say, of course, that the sect had no antecedents. The evidence suggests that it was a reformist schism from the Kälämukha church

121See K.A.N. Sastri, The Colas (1st edition), I, 433, and Rangacharya, Inscriptions I, 445. 122See K.A.N. Sastri, The Colas (1st edition), I, 433-444. 123See ed. H.K. Sastri, SII, V, no. 1358. 124The most readable account of the life of Basava in English is still that in R.G. Bhandarkar's Early History of the Dekkan (pp. 101-104). See also Yazdani (ed.). The early History of the Deccan, I, 461-65, and Mysore Gazetteer, II, Part II, 873-93.

168 THE KĀPĀLIKAS AND KĀLĀMUKHAS

with Basava cast in the role of Luther.125 The chief Virasaiva sources for their own early history are two Kannada works-the Basava Purāna, written in about A.D. 1370,126 and the Canna-Basava Purāna, written in about 1585.127 A quite different account of the life of Basava is contained in a Jain work, the Bijjalarāyacarita. The Basava Purāna avers that Basava was the son of a Brahman named Mādirāja and his wife Mādalāmbikā of Bāgevādi (in Bijāpur District). Basava was married to the daughter of Baladeva, Bijjala's chief minister, and was appointed to Baladeva's pcsition after the latter's death. Basava's sister Nāgalāmbikā had a son named Canna-Basava. After Basava's appointment, he and his nephew began propounding the new Virasaiva doctrine and won a great number of converts. In the process they rapidly depleted Bijjala's treasury with munificent gifts to the jangamas, the Viraśaiva priests. This alienated the king who sought to punish Basava, but before he could do so the latter fled. The king set out to capture him, but Basava gathered together a large number of his followers and defeated the king in battle. The king then reinstated Basava to his old position but their relations were never again the same. Basava eventually commission- ed one of his followers to murder the king. After the murder Basava hurried to the shrine of Sangameśvara at the confluence of the Malaprabha and Krishna rivers and was absorbed into the godhead. The major portion of the Jain version is similar, but there are several important differences. After the regicide, for instance, the murdered king's son is said to have chased Basava to Ulavi on the Malabar coast, where the former minister ignominiously committed suicide by throwing himself into a well. Basava's nephew Canna-Basava was later reconciled with the new king and became sole leader of the Vīraśaivas. Both these sources are relatively late and there is no solid epigraphic confirmation of the story. For this reason J.F. Fleet

125S.C. Nandimath, in his A Handbook of Virasaivism (p. 9), notes that the Kālā- mukha matha at Hūli is now an important Virasaiva mațha and tentatively concludes that the transformation of Kalamukha mathas into Virasaiva mathas may have occurred elsewhere as well : 'Slowly and imperceptibly they were absorbed into Vīraśaivism.' The Mysore Gazetteer (II, Part II, 885) offers the opinion that the Saivite revival under Basava and the other early Virasaivas 'seems to have followed as the natural result of the work of these Saiva teachers of the Pāsupata [sic for Kālāmukha] school at Balagami [=Belagāve]. 126Trans. G. Würth, JBBRAS, VIII (1865-66), 65-97. 127Trans. G. Würth, JBBRAS, VIII, 98-221.

OTHER KĀLĀMUKHA PRIESTHOODS 169

thought it best to ignore it (EI, V, 242-45). K.A.N. Sastri feels that Fleet went too far in this rejection and cites a genealogy contained in a subsequently edited Arjunawada inscriptions of the Yādava Kannara (A.D. 1260) which mentions 'Basava or Sangana-Basava as the younger son of Mādirāja described as Tardavādi-madhyagrāma-Bāgavādi-puravarādhīsvara.'128 These two persons, Sastri believes, must be the famous Viraśaiva and his father. Although this identification is perfectly plausible, there is still no epigraphic corroboration of the legendary biographies or even of Basava being a minister of Bijjala. An inscription of about A.D. 1200 found at the Somanātha temple at Ablūr in Dharwar District provides better evidence about a Śaivite priest named Ēkāntada Rāmayya,129 who is des- cribed as a Viraśaiva saint in the Canna-Basava Purāna.130 A great contest was held in the town between him and the Jains. He van- quished them by offering his own head to Siva, who restored it as good as new after seven days. The losers still refused to destroy their image of Jina and establish one of Siva in its place. Ēkāntada Rämayya destroyed their shrine and built a large temple of Vira- Somanātha in its place. The Jains appealed to Bijjala for retribution but declined his offer of a second contest for bigger stakes. Bijjala therefore dismissed their appeal and 'bestowed on Ēkāntada Rāmayya, in the public assembly, a jayapatra or certificate of success.' This record contains an important clue to the possible relations existing between the Viraśaivas and Kālāmukhas. In it Ēkāntada Rämayya is said to have delivered a sermon in the Brahmeśvara temple at Ablür. As we have seen,131 this temple was headed by Kālāmukha priests of the Mūvara-koņeya-samtati until at least A.D. 1144. There is no reason to assume that it was not still in their hands when Ēkāntada Rāmayya visited it a few years later. Evidently Ēkāntada Rāmayya and these Kālāmukha priests

128In Yazdani (ed.), The Early History ... , I, 463. 129Fleet, EI, V, no. E. 130Trans. G. Würth, JBBRAS, VIII, 198. The story of 'Yekanta Rämeiya' in this work differs considerably from the epigraphic account, but the essentials are similar enough to confirm that he is the same priest. According to the Purāņa, 'Yēkānta Rāmeiya, a great saint,' heard of the fame of Basava and went to Kalyāna to see him. This suggests that Ekāntada Rāmayya became a Vīraśaiva after his reputation was already established. This Ablūr inscription gives no specific indication that he belonged to this sect. 131See above, p. 118.

170 THE KĀPĀLIKAS AND KĀLĀMUKHAS

maintained cordial relations. It is even possible that at the time of delivering his sermon Ēkäntada himself was a member of the Kālāmukha sect.132 The Brahmeśvara temple is today known as the temple of Basaveśvara and is a Vīraśaiva shrine.133 Many other former Kālāmukha temples are now controlled by the Virasaivas. The exact number and percentage is not known134 but they include the Kedāreśvara temple in Belagāve, the Trikū- teśvara temple at Gadag (Dharwar District), and the Kālāmukha temples at Hüli (Belgaum District). Belagäve, the former seat of the Śakti-parişad, is now a center of Viraśaiva activity and is visited by Vīraśaiva pilgrims from the surrounding areas.135 Of the five most sacred Vīraśaiva mathas, one is at Srīparvata in Kurnool District and one at Kedaranāth in the Himalayas-both sites with important Kālāmukha associations. Even more signi- ficant, perhaps, is the fact that few Virasaivas are found in areas not formerly dominated by the Kālāmukhas. The similarities we can trace between Kālāmukha and Vīraśaiva cult and philosophy are regrettably few. This is not altogether surprising since very little is in fact known about the early Viraśaivas apart from the broad outlines of their history. Our information about the Kālämukhas is not much greater except insofar as we assume that they followed the doctrines and rituals of the Pāsupatas. The most characteristic feature of later Viraśaiva philosophy, the doctrine of sat-sthala, is not mentioned in the early vacanas. According to S.N. Dasgupta, the philosophical content of Basava's vacanas is negligible.136 One of the earliest Viraśaiva philosophical works, Revaņārya's Siddhānta-śikhāmani, written in about the thirteenth century, does present certain similarities to Kālāmukha- or at least Päsupata-doctrine. This work states that Siva-Brahman is without any form or differentiation, yet is endowed with will by which he creates and destroys the world.137 This corresponds to the view of the Kālamukha priest Bonteyamuni of Hombal that the Creator (kartr) is formless.138 The idea that the distribution of the fruits of karman is managed and controlled by God, rather

132See above, note 130. 133See Fleet, EI, V, 213. 134A systematic collection of this information would be of considerable interest. 135Information gathered from local informants at Belagave in March, 1966. 136 History of Indian Philosophy, V, 44. 137Ibid., pp. 48-49. 138See above, pp. 132 and 134.

OTHER KĀLĀMUKHA PRIESTHOODS 171

than being automatic and autonomous, is, according to Dasgupta, a doctrine which Revaņārya borrowed from the Pāsupatas.139 We should also note here that Sripati Pandita, a fourteenth century Vīraśaiva commentator on the Brahma-sūtra, quotes approvingly Haradatta, the author of the Pāsupata Gaņakārikā.140 The social doctrines preached by the early Virasaivas included contempt for the caste system and child marriage as well as approval of widow remarriage. We know virtually nothing about Kālāmukha social attitudes, but it seems probable that they were more orthodox.141 Two important similarities between the respective cults of the Kālāmukhas and Vīraśaivas are the organization into large mațhas and the emphasis placed on linga worship. The former needs no comment. There is no epigraphic evidence that the Kālāmukhas ever followed the Viraśaiva practice of each devotee wearing a small linga, but the Basava Purāna itself admits that this practice existed even before Basava. In any case, the Kālāmukhas valued linga worship very highly. In the Kedāreśvara record of c. A.D. 1164, for instance, the Belagāve priest Kedāraśakti asserts that 'Whoso sets up but one linga, obtains a myriad-fold all the merit described in the ägamas.'142 The priests of the Viraśaivas are called jangamas, a term they explain as 'lingas in movement.'143 In a number of inscriptions the same word, with the same interpretation, is applied to Kālāmukha priests. The A.D. 1192 record from Gadag describes the Kālāmukha Candrabhūșana-pandita-deva as the fourth, and moving (jangama) linga of the god Trikuteśvara.144 The donor of the A.D. 1189 grant from Hombaļ, the Kālāmukha priest Rudraśakti, is called a jangama,145 and the Kālāmukha ascetic Sarveśvara-deva, who headed the Jagadīśvara temple at Munavaļļi in A.D. 1252, is called a jangama-lingāvatāra, an incarnation of a moving linga.146 In the Vīraśaiva Purāņas the relations between Basava and the jangamas, especially the profligate twelve thousand, are not altogether

139 History ... , V, 49-50. 140Ibid., pp. 9-10. 141We have alluded to the Kālāmukha attitude to caste above, p. 149. 142Ed. and trans. Rice, EC, VII, Sk. 108. 143L. Renou and J. Filliozat, L'Inde classique, I, 638. 144See above, pp. 120-21. 145Ed. Desai, SII, XV, no. 73 (1. 50). 146See above, pp. 152-53.

172 THE KĀPĀLIKAS AND KĀLĀMUKHAS

clear.147 It appears, however, that the jangamas were organized even before the saint's appearance. It is by no means inconceivable, we feel, that these early jangamas were none other than the Kālāmukhas.

147See Basava Purāņa, trans. G. Würth, JBBRAS, VIII, pp. 68 and 71.

CHAPTER VI

LAKULĪŚA AND THE PĀŚUPATAS

We have noted more than once that Rāmānuja describes four sects as following the doctrine of Pasupati : the Kāpāla, the Kālā- mukha, the Pasupata, and the Śaiva. The Pāsupata sect is the oldest of the four and was the spiritual parent of the Kālāmukha sect, if not of the others. In the period of Kālāmukha dominance in Mysore, which is also the time in which Ramanuja preached, the epigraphs of the Pāśupatas and Kālāmukhas display many simi- larities. Both sects revere the legendary teacher Lakulīsa. The ascetics of both bear similar or identical names and undertake pilgrimages to Kedāranāth and Śrīparvata.1 The philosophical content of the Iśvara-kartr-vāda propounded by the Kālāmukha priest Bonteyamuni of Hombal2 can be little different from the Pāśupata doctrine of Isvara as Cause (kāraņa) of the Material Universe (karyā). We have quoted above the passage from ins- criptions at Nesargi and Sirasangi which seems to equate Kālā- mukha, Mahāvratin and Mahāpāśupata.3 Most sources, however, describe the Kālāmukhas and Pāśupatas as separate, though closely allied, Śaivite sects.4 A number of modern scholars have written about the history, ritual and philosophy of the Pasupatas. Since these topics lie somewhat at the periphery of our study, we will be content to summarise and review these scholars' arguments and to comment on possible reflections of Pāśupata beliefs and practices in Kālā- mukha epigraphs. The best discussions of the early history of the Pāsupatas are those by J.N. Banerjea.5 Other contributions in this field have

1See above, pp. 107-109. 2See above, p. 134. 3See above, pp. 151-52. 4See above, pp. 7-12. 5In K.A.N. Sastri (ed.), The Mauryas and Satavahanas, pp. 393-403; in his own Development of Hindu Iconography; and his article 'Lakuliśa-The Founder or the Systematiser of the Pasupata Order' in Indian History Congress, Proceedings of the Fourteenth Session, Jaipur, pp. 32-36.

174 THE KĀPĀLIKAS AND KĀLĀMUKHAS

been made by D.R. Bhandarkar, who first noted the purāņic data on Lakulīśa and enabled scholars to establish his approximate date,6 and by J.F. Fleet, R.G. Bhandarkar, M. Hara and others.7 Studies more concerned with later developments have been made by G. Bühler, K.K. Handiqui, H.D. Sankalia and others.8 The earliest references to the Pasupatas are probably those in the Mahābharata. The most important mentions five religious doctrines-Sāņkhya, Yoga, Pāñcarātra, the Vedas, and Pāśupata- and says that the last was propounded by Siva, who is also called Lord of Umã, Lord of Beings, Śrīkantha, and Son of Brahmā.9 Extrapolating from a very tentative suggestion of R.G. Bhandarkar, who first noted the passage,10 V.S. Pathak has attempted to prove that a historical person named Srikantha was the founder of the Pāśupata order.11 The passage in question, however, clearly refers to the god Siva and not to a divinised human being. Of the allusions

6'An Eklingji Stone Inscription and the Origin and History of the Lakulisa Sect, JBBRAS, XXII (1904-07), 151-65; 'Lakulīśa' in Archaeological Survey of India, Annual Report : 1906-7, pp. 179-92; 'Some Published Inscriptions Reconsidered,' IA, XXXII (1913), 57-64; and 'Mathura Pillar Inscription of Chandragupta II : G.E. 61,' EI, XXI (1931-32), 1-9. 7J.F. Fleet, 'Siva as Lakulīsa,' JRAS for 1907, pp. 419-26; R.G. Bhandarkar, Vaişņavism ...; M. Hara, 'Pāśupata kenkyū II-Pañcārthabhāşya ad Pāšupatasūtra I, i,' Hikata Hakushi koki kinen ronbunshū, pp. (65)-(80); R.N. Mehta, 'Karavan- The Seat of the Lakulisa Sect' in Indian History Congress, Proceedings of the Four- teenth Session, Jaipur, pp. 71-76, and his 'Avākhal : the Traditional Ulkāgrāma of Kārvaņa Māhātmya, JOIB, VII (1957), 169ff .; P.C. Divanji, 'Lakulīśa of Kārvān and his Pasupata Culture,' JGRS, XVII (1955), 267-74; A.P. Karmarkar,: The Vrātya or Dravidian Systems; and V.S. Pathak, History of Saiva Cults in Northern India. 8G. Bühler, 'Cintra Praśasti of the Reign of Särangadeva,' EI, I (1888), 271-87; K.K. Handiqui, pp. 337-54, 467-510; H.D. Sankalia, The Archaeology of Gujarat; T.V. Mahalingam, 'The Pasupatas in South India,' JIH, XXVII (1949), 43-53, and his 'A Family of Pasupata Grhasthas at Jambukeśvaram,' JORM, XXV (1957), 79-85; B.P. Majumdar, 'Lakulīśa Pāsupatas and their Temples in Medieval India,' JBRS, XXXIX (1953), 1-9; H.K. Narasimhaswami, 'Dommara-Nadyala Plates of Punyakumara; 10th Year,' EI, XXVII, 268-76, and his 'Bhairavakonda Inscription of Vikramaditya,' EI, XXXIII, 78-81; N. Venkataramanayya in G. Yazdani (ed.), The Early History of the Deccan, II, 704-13; and A.K. Vyas, 'Paldi Inscription of Guhila Arisimha; V.S. 1173,' EI, XXX, 8-12. Relevant material is also found in the works mentioned in the previous note. 9 Śāntiparvan, Part III, ed. V.S. Sukthankar and S.K. Belvalkar, xii. 337. 59 and 62: 'sāmkhyam yogam pāñcarātram vai// ... umāpatir bhūtapatih śrīkaņțho brahmaņaḥ sutaḥ/ uktavān idam avyagro jñānam pāśupatam śivaḥ//.' 10Vaișnavism ... , p. 116. 11 History of ... , pp. 4-8.

LAKULĪŚA AND THE PĀŚUPATAS 175

to Śrīkantha which Pathak cites in support of his argument, most seem to denote the god Śiva-Śrīkantha and only one makes any connection between Śrīkantha and Pasupata doctrine. This passage, from the Tantrāloka of Abhinavagupta (c. A.D. 1000), merely says that Śrikantha and Lakuleśvara are the two authorities on Śiva-sasana. This is hardly conclusive, or even very useful, evidence. Also following R.G. Bhandarkar, but more plausibly, J.N. Banerjea has suggested that the Siva-bhagavatas mentioned by Patañjali (c. 150 B.c.) in his comments on Pāņini v. 2. 76 were pre-Lakulīśa Pāśupatas.12 Banerjea sees Lakulīśa as the 'systema- tiser' of this earlier Pāsupata order. Unfortunately, this theory finds no support in either of the two extant Pasupata texts-the Pāupata-sūtra with the Pañcārtha-bhāşya of Kauņdinya13 and the Gaņakārikā with the Ratnațīkā attributed to Bhāsarvajña.14 In his gloss on Pāśupata-sūtra i. 1, Kauņdinya (pp. 3-4) says that the Lord assumed the body of a Brähmana and came to earth at Kāyāvatarana. Then he went to Ujjayini where he imparted the sütras to a disciple named Kuśika. This, as we shall see, is a clear allusion to the legend of Lakulīsa. The Ratnatīka instructs the. devotee to honor the tīrthakaras beginning with Lord Lakulīśa and ending with Rasīkara. Thus by as early as the Gupta period, the time to which Kaundinya is generally assigned, Lakulīśa was regarded as the founder of the order by the Pāsupatas themselves. Banerjea's theory must be viewed as pure speculation. It is certain that Saivite ascetics existed before Lakulisa, and some of these undoubtedly inculcated similar beliefs and practices. Religious orders are never founded in a vacuum. Wine and wineskins are never completely new. Nonetheless this does not justify calling these early Śaivite ascetics Pāśupatas. Lakulīśa was in all likelihood the founder of the Pasupata order. The sources for his personal history are fairly numerous and varied

12Indian History Congress, Proceedings of the Fourteenth ., pp. 32-36. See also his, The Development ... , pp. 448-52 and his discussion in K.A.N. Sastri (ed.), The Mauryas ... , pp. 396-400. 13Ed. R.A. Sastri. 14Ed. C.D. Dalal. The authorship is disputed. The colophon attributes the Ganakā- rikā and not the Ratnațīkā to Bhāsarvajña. We agree with most authorities, however, that it is the commentary and not the text that this famous tenth century Naiyāyika logician composed. The author of the Ganakārikā was perhaps named Haradattā- cārya. See R.A. Sastri's introduction to Pāśupata-sūtra, p. -

176 THE KĀPĀLIKAS AND KĀLĀMUKHAS

but not very complete or consistent. In addition to the allusion in Kauņdinya's bhāsya, legends of his birth and priesthood appear in the Vāyu15 and Linga Purānas,16 the Kāravaņa Māhātmya,17 and three early mediaeval inscriptions.18 The Kāravaņa Māhātmya, a comparatively late work, gives the most complete version. According to it Siva was born as the son of a Brāhmaņa couple named Viśvarāja and Sudarśanā in the village of Ulkāpurī. The divine infant performed several superhuman feats but died at only seven months. His mother put him into the water of a nearby tirtha, and from there he was taken by tortoises to the Jāleśvara- linga. He was brought back to life after this initiatory journey to the underworld and later went to Kāyavarohana where he took up his priestly mission. In the Vāyu and Linga Purāņas Śiva predicts, in puranic fashion, that he will become incarnate as the brahmacārin Lakulin by entering a corpse found in a cremation ground at Kāyārohaņa (Vāyu) or Kāyāvatāra (Linga). This was to occur in the twenty-eighth yuga when Krna was incarnate as Vāsudeva. The stone inscription of A.D. 971 from the Ēklingjī temple near Udaipur states that in the country of Bhrgukaccha, the region around modern Broach, the sage Bhrgu was once cursed by Visnu. The sage propitiated Siva for aid and the god became incarnate as an ascetic holding a club (lakula). This occurred at Kāyāvaro- hana.19 The Paldi inscription of A.D. 1116, also found near Udaipur, says that when Siva saw the tree of dharma being destroyed by the axes of the Kali-yuga, he descended to earth at Kāyāvarohaņa in Bhrgukaccha.20 The Cintra praśasti of Sārangadeva, a late thirteenth century record from Somnäth in Gujarat, relates how Śiva came to Lāța and dwelt at Kārohaņa as Lakulīśa 'in order to bestow favour on the universe' and also 'to favour the offspring of Ulūka, who long were deprived of sons in consequence of a curse of their father .. . '21

15Ānandāśram edition, xxiii. 219-224. 16Ed. J. Vidyasagara, i. 24. 124-34. 17Ed. C.D. Dalal in Gaņakārikā, pp. 37-57. It is well summarised by D.R. Bhandar- kar in Archaeological Survey of India, Annual Report : 1906-7, pp. 180-83. 18D.R. Bhandarkar, JBBRAS, XXII, 151-65; Vyas, EI, XXX, 8-12; and Bühler, EI, I, 271-87. 19D.R. Bhandarkar, JBBRAS, XXII, 166. Was this club used to do battle with Vişņu? Bhrgu is usually portrayed as a Vaișņava. 20Vyas, EI, XXX, 11. 21Trans. Bühler, EI, I, 274. Fleet (JRAS for 1907, p. 419) offers an alternative interpretation in which Śiva-Lakuliśa, 'in order to favour the boys of Ulūka, who

LAKULĪŚA AND THE PĀŚUPATAS 177

It is evident that not much solid historical information can be derived from these accounts. They are, however, unanimous on two points-that Lakulīsa was an incarnation of Siva, and that he settled at a place called Kāyāvarohaņa, Kāyāvatāra, Kārohaņa, or Kāyārohaņa locatedin the Lāta or Bhrgukaccha region. It also seems that Lakulisa belonged to a Brāhmana family and that he travelled at least as far as Ujjain to preach his doctrines. Kāyāvarohaņa is unanimously identified with the modern village of Kārvāņ about 19 miles north of Baroda.22 Ulkāpurī, Lakulīśa's birthplace in the Kāravaņa Māhātmya version, is modern Avākhal in the same region. The legend of the sons or offspring of Ulüka alluded to in the Cintra prasasti is not mentioned in other sources. Ulūka may be somehow connected with Ulkāpurī.23 The name Ulūka also appears in the purāņic lists of the avatāras of Śiva who preceded Lakuliśa and is associated with Kaņāda, the founder of the Vaisesika system of philosophy. Fleet's attempt (JRAS for 1907, pp. 425-26) to connect the inscriptional Ulūka with the Mahābharata story of Sakuni, the son of a king of Gandhara named Subala, and Sakuni's son Ulūka is not altogether convincing. The name Lakulīśa-with its variants Nakulīśa, Lakuleśa, Lakulin, and Lakulīśvara-is derived from the word lakula, 'laguda or lakuta meaning 'club.' This is clearly shown in the ex- pression 'whose hand was characterised by a club (lakulopalakşita- kara)' from the Eklingji inscription24 and in the epithet Lakuta-pāņi from the Kāravaņa Māhātmya.25 Lakulīśa is thus the Lord (īśa) who bears a club (lakulin). A plain club-not the khatvānga suggested by Bühler (EI, I, 274)-is in fact the identifying mark of Lakulīśa sculptures. D.R. Bhandarkar quotes a verse from a work called Viśvakarmāvatāra-vāstu-šāstra which specifies that sculptures of Nakulīśa should show him seated in the padmāsana, with his penis erect, and with a citron (mātulinga) in his right hand and a club (danda) in his left.26 Most known sculptures

were for a long time without sons in consequence of a curse laid upon (his) father, ... settled (adhyuvāsa) at Kārohaņa.' 22See Bühler, El, I, 274. The name variants Kāyāvarohana (descending of the body) and Kāyārohaņa (ascending of the body) have opposite meanings but it is clear that both represent the same place. See Hara, 'Pāśupata kenkyū II ... ' 23See Bühler, EI, I, 274. 24D.R. Bhandarkar, JBBRAS, XXII, 166. 25Gaņakārikā, p. 37. 26 Archaeological Survey of India, Annual Report : 1906-7, p. 186.

178 THE KĀPĀLIKAS AND KĀLĀMUKHAS

depict him in this pose.27 Recently Daniel H.H. Ingalls has pointed out the striking resem- blances between the religious practices of the Pasupatas and those of the Cynics.28 He notes that 'one cannot avoid the suspicion that the name Lakulīśa is derived both semantically and phonetically from the patron saint of Cynicism,' Hercules ('HpakAns), another man-god who wielded a famous club.29 The physical resemblance, though not the phonetic one, between Lakulīśa and Hercules was recognised much earlier by Fleet (JRAS for 1907, p. 424). He suggests that the Indian god ancient Greek writers called Hercules might be Siva and not Krsna as scholars usually assume. Whether or not this is true in every case, he adds, 'we can hardly doubt that the club of Śiva as Lakulīśa is the club of Hēraklēs.' He also points out that in about the first century A.D. the figure of Hercules on the coins of the Kusāņas was being replaced by Siva. On one of the coins of Huviska Śiva is shown holding a club. In spite of the similarities in the cult practices of the Pāśupatas and the Cynics and the resemblances between Lakulīsa and Her- cules, Ingalls (HTR, LV, 296) is forced to 'doubt that the evidence permits one to speak of a genetic relation.' Thus he rejects the conclusion of F. Sayre, who, without being aware of the Pāsupata parallels, sought to derive Cynicism from India.30 On the whole we are inclined to agree with Ingalls since, as he indicates (HTR, LV, 296), there is virtually no direct evidence of a foreign back- ground within either cult. Nonetheless there are one or two circum- stantial details not noted by either scholar which point to some sort of relationship. Sayre has well documented the Greek evidence for the migration of Indian religion and philosophy to Greece but ignored the evidence for migration in the other direction. His statement (p. 45) that 'the Greeks were receptive of ideas from other nations while the Indians were not' has been amply refuted by R.A. Jairazbhoy.31

27See ibid., pp. 184-89. For a recent discussion of some variant sculpture poses, see R.C. Agrawala, 'Two Standing Lakulīśa Sculptures from Rajasthan,' JOIB, XIV, (1965), 388-91. 28'Cynics and Pasupatas : The Seeking of Dishonor,' Harvard Theological Review, LV (1962), 281-98. 29Ibid., p. 296. See also pp. 292-93. 30 Diogenes of Sinope : A Study of Greek Cynicism, pp. 38-47. 31 Foreign Influence in Ancient India. For the impact of Greek culture on India see especially chapter five.

LAKULĪŚA AND THE PAŚUPATAS 179

One must agree with Ingalls (HTR, LV, 296) that if there were any borrowings, 'there exists a stronger possibility that the Pāsupatas were influenced by the Cynics.' Sayre (pp. 39-40) stresses that Sinope, the home of Diogenes, the effective founder of the Cynic cult, was an entrepôt on the ancient trade route between India and Greece. This, we might add, was also true of the Broach or Bhrgu- kaccha region where Lakuliśa taught. At the time of the author of the Periplus (first century A.D.), Barygaza (Broach) was the chief trading port in western India. It seems to have held this position from as early as Mauryan times.32 Gautama or Akșapāda, the traditional author of the Nyãya-sūtra, is also believed to have resided in this region, and one modern authority has claimed that his logic borrows extensively from Aristotle.33 As we have seen, both the Pāsupatas and the Kālāmukhas are known to have had close connections with the Nyāya and Vaiśeșika schools of philo- sophy. It is also perhaps significant that Kashmir, the region from which many Kälämukhas migrated to the South, was on another important trade route to the West and had been exposed to Greek culture under the Indo-Greeks. Much controversy has centered around the date of Lakulīśa. Fleet (EI, VI, 228) originally identified him with the Kālāmukha priest Lakulīśvara-pandita who presided over the Pañcalinga temple in Belagāve in A.D. 1035. Fleet later abandoned this opinion (JRAS for 1907, p. 420) in the light of the discoveries of D.R. Bhandarkar. In view of the references to Lakulīśa in the Vāyu Purāņa, a work usually assigned to the early Gupta period, Bhandarkar (JBBRAS, XXII, 157) placed him 'as early as the first century A.D. at the latest.' R.G. Bhandarkar dated the rise of the Pāsupata system mentioned in the Mahābhārata, and presum- ably Lakulīśa as well, 'about a century after the rise of the Pancarātra system, i.e. about the second century B.C.'34 In 1931 D.R. Bhandar- kar published the Mathura pillar inscription of Candragupta II (EI, XXI, 1-9). This records a donation by the Mäheśvara teacher Uditācārya of two lingas named after his teacher, Bhagavat Kapila, and teacher's teacher, Bhagavat Upamita. Uditācārya is described as tenth in descent from Bhagavat Kusika and fourth in descent

32See R. Thapar, Asoka and the Decline of the Mauryas, pp. 80-83, 229, and also K.A.N. Sastri (ed.), The Mauryas ... , pp. 307-308, 437-46. 33See S.C. Vidyabhusana, 'Influence of Aristotle on the Development of the Syllogism in Indian Logic,' JRAS for 1918, pp. 471, 486-88. 34 Vaișnavism ... , p. 117.

180 THE KĀPĀLIKAS AND KĀLĀMUKHAS

from Bhagavat Parāśara. Bhandarkar identified this Kuśika with Kuśika, the disciple of Lakulīśa. Since the inscription dates from A.D. 380, he assigned Lakulīsa to the first half of the second century A.D. This date has justifiably commanded acceptance by most scholars, but there are a few problems and uncertainties about it which many have unfairly ignored. First, neither the word Lakulīśa nor Pāśupata occurs in the record. It was issued as a request to future Māheśvaras to protect and honour the two lingas. Māheśvara is normally simply a generic term for those who worship Śiva. Sāyaņa-Mādhava, for instance, applies it to the adherents of both the Pāsupata and Śaiva doctrines.35 Śamkarācārya seems to use it for the Pāsupatas alone,36 but Vācaspati Miśra and Bhāskarācārya divide the Māheśvaras into four distinct sects.37 The only reliable means of identifying Uditācārya's sectarian allegiance are the inscription's concluding line of praise to Lord Danda, who bears the staff of Rudra (Rudra-danda), and the Lakulīśa-like standing figure engraved on the pillar. These render it reasonable to assume that this is a Lakulīśa-Pāśupata record, but there is still a problem about the identification of Bhagavat Kuśika. As V.S. Pathak has noted,38 there are at least two Kusikas among the spiritual descen- dants of Lakulīśa. Kuśika I was his chief disciple and is mentioned in a number of epigraphic and literary sources.39.Kuśika II is mentioned in Rājaśekhara's Şaddarśana-samuccaya (c. 1350),40 in the commentary of Gunaratna (c. 1375) on Haribhadra's work of the same name,41 and indirectly in the Pāsupata Ratnațīkā

35 Sarvadar śana-samgraha, ed. U.S. Sharma, pp. 297, 320. 36 Brahma-sūtra-bhāşya ii. 2. 37. 37Vācaspati Miśra, Bhāmatī on Brahma-sūtra ii. 2. 37. Bhāskarācārya, Brahma- sūtra-bhāșya ii. 2. 37. 38 History of Saiva Cults ... , p. 9. 39In addition to the reference in Kaundinya's bhāşya on Pāšupata-sūtra i. 1, Kuśika I is named in the Cintra praśasti (Bühler, EI, I, 273 and 281), the Eklingji inscription (D.R. Bhandarkar, JBBRAS, XXII, 152 and 167), the Paldi inscription (Vyas, EI, XXX, 9 and 11), the A.D. 987 Udeypur inscription of Naravāhana (cited by Pathak, pp. 9-10), the Vāyu Purāņa, xxiii. 223, the Linga Purāņa, i. 24. 131, and, as Kuņika, in the Kūrma Purāņa i. 53 (p. 443). He also appears as Kauśika and Sauşya-Kauśika in the works by Rājaśekhara and Guņaratna cited below. 40Extract edited by Dalal in Gaņakārikā, pp. 35-36. 41 Extract edited and translated by D.R. Bhandarkar in Archaeological Survey of India, Annual Report : 1906-7, pp. 190-92. This is the same as the extract edited by Dalal in Gaņakārikā, pp. 29-30 although the readings vary slightly.

LAKULĪSA AND THE PĀŚUPATAS 181

commentary on the Ganakārikā.42 Kusika II is tenth in a list of seventeen or eighteen tīrtheśas or tīrthakaras beginning with Lakulīśa. Although the list is not in strict chronological order- Lakulīśa's four disciples appear to succeed one another-Kuśika II must have lived some time after Kusika I. If the Mathura inscription refers to Kuśika II, then Lakulīśa's date mut be pushed back about another one hundred years. A Kusika III or IV is, of course, also by no means impossible. Lakulīśa had three other important disciples besides Kuśika. The names of all four disciples are given, with variations, in the Kūrma, Vāyu and Linga Purāņas,43 by Rājaśekhara and Gunaratna,44 and in the Cintra praśasti. This last record describes Lakulīsa's arrival at Kārohaņa in Lāța and then continues :

His four pupils-Kuśika, Gārgya, Kauruşa, and Maitreya- arrived (avateruh) at this place in order to (learn) the special conduct (caryä) of the Pāśupata vow. The fourfold lineage (jäti) of those ascetics then came into being (and) adorned (all) the land girded by the four oceans.45

The abbot (sthānādhipa) Kārttikarāśi, who became 'an ornament of the gotra of Gärgya,' is then introduced. He belongs to the early thirteenth century. By this time, it seems, the followers of the four disciples of Lakulīśa were organised into separate groups. From the evidence of the Mathura pillar inscription, this division probably goes back at least as far as the fourth century A.D. No records survive of any priesthoods which traced their descent

42Ed. Dalal, p. 19. The Ratnatīkā merely mentions the tirthakaras beginning with Lakuliśa and ending with Räsikara, who is the seventeenth firthesa in the lists of Rājaśekhara and Guņaratna. These two authors name eighteen tīrthesas as follows (Rājaśekhara's reading is given first) : (1) Nakulīśa or Nakulin, (2) Kauśika or Sausya-Kauśika (= Kuśika I), (3) Gārgya, (4) Maitrya, (5) Kaurușa or Akauruşa, (6) Iśāna, (7) Pāragārgya, (8) Kapilāņda, (9) Manuşyaka, (10) Aparakuśika or Kuśika (= Kuśika II), (11) Atri, (12) Pingalākșa or Pingala, (13) Puşpaka or Puşyaka, (14) Brhadācārya or Brhadārya, (15) Agasti, (16) Santāna, (17) Rāśīkara, and (18) Vidyäguru. Most of the names in Dalal's edition of the Gunaratna extract coincide with those he gives for Rājasekhara rather than with Bhandarkar's version of Guņaratna. 43 Kūrma Purāņa i. 53 (Kuņika, Garga, Mitraka, Ruru); Vāyu Purāņa xxiii. 223 (Kuśika, Gargya, Mitraka, Rușța); and Linga Purāņa i. 24. 131 (Kuśika, Garga, Mitra, Kauruşya). 44See note 42 above. 45Ed. Bühler, EI, I, 281 (my translation).

182 THE KĀPĀLIKAS AND KĀLĀMUKHAS

from either Maitreya or Kauruşa. R.G. Bhandarkar attempted to connect the third disciple, Kauruşa, with the Kāruka-siddhāntins named as one of the four Saivite sects by commentators on Brahma- sūtra ii. 2. 37.46 These Kāruka-siddhāntins appear as Kāruņika- siddhāntins in Vācaspati's (c. 850) Bhāmatī, as Kāțhaka-siddhāntins in Bhāskarācārya's (c. 850) Brahma-sūtra-bhāșya, and as Kālā- mukhas in Rāmānuja's Śrī-bhāșya and other commentaries on this sutra.47 We cannot accept Bhandarkar's theory. In the first place, the word Kaurusa is not very close phonetically to Kāruka, Kāțhaka or Kālāmukha. Secondly, there is no precedent or reason for connecting the names of any of the other three Saivite sects- Pāśupata, Śaiva and Kāpālika-with the names of Lakulīśa's other three disciples. The followers of Kuśika and Gärgya both seem to have been Pāsupatas. Thirdly, if the Kāruka-, Kāruņika- and Kāțhaka-siddhāntins later became known as Kālāmukhas, 48 as appears likely, and if they were all descended from Kaurusa, it is strange that no mention of this disciple is found in any Kālā- mukha epigraph. By the time of Harsa-vardhana (606-647), and probably as early as Gupta times, there were Pāśupata temples in most parts of India.49 The pilgrim Hsüan Tsang met or heard reports about 'ash-smeared (followers) of the outer way,' i.e. Pāśupata heretics,50 at Jālandhara in East Panjab, Ahicchatrā in U.P., Malakuța in

46R.G. Bhandarkar, Vaisnavism ... , p. 121. We have not been able to trace any of these commentators, but Karuka-siddhantin is listed in M. Monier-Williams' Sanskrit-English Dictionary. 47See above, p. 1. 48There is a tenuous relation between Kāthaka-siddhānta and the Pāsupata-sūtra. Many of the sutras in this work are based on the Taittirīya Aranyaka (see C. Chakra- varti, Pāśupatasūtra, IHQ, XIX, 271). Kāțhaka is a school of the Black Yajur Veda, the Veda to which the Taittirīya Āranyaka belongs. 49 Although several small studies and surveys of the later history of the Pāsupata sect have been published, a complete work is still needed. The purānic material especially should be examined in more detail. Several of these works show definite Pāsupata influence. A preliminary study of this influence on parts of the Kūrma Purāna has been made by R.C. Hazra, 'The Smrti-chapters of the Kūrma Purāna,' IHQ, XI (1935), 265-86. The Vāyu Purāna has two chapters (xi and xiv) on Pāsupata- yoga. Their contents bear only a partial resemblance to the doctrines of the Pāśupata- sūtra. 50The characters are 5/ ia Hodous and Soothill, Beal, Watters, and others translate this as Pasupata. There is no reason to quarrel with this interpretation, but it should be noted that the various Chinese phonetic equivalents of Pāsupata do not occur in Hsüan Tsang's text.

LAKULĪŚA AND THE PĀŚUPATAS 183

South India, Mālava, Khotan, Kapiśa (Nuristan) in East Afgha- nistan, Gandhara, Varanasi, and elsewhere.51 Two early seventh century inscriptions registering grants to Pāśupata ascetics have been found as far afield as South-East Asia.52 Imeortant early references to Pāśupatas also occur in Mahendravarman's South Indian drama, Mattavilāsa, and, indirectly, in Varāhamihira's Brhat-samhitā.53 Sanskrit writers from Bāna onwards mention them frequently, and from about the tenth century epigraphical references also become numerous. Post-Gupta sculptures of Lakulīśa have been found throughout India, although the center of gravity for both sculpture and epigraphy shifts to the South by about the end of the tenth century.54 No texts on the ritual regimen and religious philosophy inculcated by the Kālamukhas have so far been discovered. There are, however, several works composed by or about the Pasupatas. In the absence of contrary evidence we must assume that the Kālāmukhas main- tained the Paśupata regimen and theology more or less intact. The first Sanskrit sources on the Pāsupata system to be noticed by modern scholars were the brief passages in the Brahma-sūtra commentaries, the 'Pasupata vow' of the Atharvaśiras Upanişad,55 and the 'Nakulīśa-Pāśupata-darśana' chapter of Sāyaņa-Mādhava's Sarvadarśana-samgraha. Although Sāyaņa-Mādhava wrote his account in the fourteenth century, it is still the best short summary of the subject. It was translated into English by A.E. Gough in 1882, into French by S. Lévi in 1889, into German by P. Deussen in 1908, and partly paraphrased again in English by R.G. Bhandar- kar in 1913.56 All these translations suffered from an inadequate text and an inability to understand some of the technical termino- logy. In 1920 C.D. Dalal published the first actual Pasupata work, the Gaņakārikā with the Ratnatīkā commentary now attributed to Bhāsarvajña. In 1940 R.A. Sastri published the newly discovered Pāsupata-sūtra with Kauņdinya's Pañcārtha-bhāșya.57

51S. Beal (trans.), Chinese Accounts of India, pp. 118, 163, 209, 228, 291, 433, 453, 461, and 464-67. 52K. Bhattacharya, Journal asiatique, CCXLIII (1955), 479-81. 53See Banerjea, Development ... , pp. 230-31. 54For an attempted explanation of this migration see above, pp. 108-109. 55Trans. R.G. Bhandarkar, Vaisnavism ... , p. 112. The whole Upanişad is trans- lated by T.R.S. Ayyangar in Śaiva Upanișads, pp. 28-53. 56See M. Hara (trans.), 'Nakulīśa-Pāśupata-darśanam,' IIJ, II (1958), 8-9. 57A list of variant readings for some of the sūtras is given by C. Chakravarti, 'Pāupatasūtra,' IHQ, XIX, 270-71.

184 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Since 1940 several scholars have attempted to re-evaluate the Päśupata system in the light of the new evidence. The most important work is the critical translation of Sāyaņa-Madhava's 'Nakulīśa- Pāśupata-darśana' chapter by M. Hara. (IIJ, II, 8-32). D.H.H. Ingalls has translated most of text and commentary of the third chapter of the Pāśupata-sūtra.58 F.A. Schultz has published a valuable study of Pāśupata theology59 and useful general surveys have been written by S.N. Dasgupta,60 K.K. Handiqui (pp. 199-204, 234-44), and K.C. Pandey.61 The theology and ritual regimen or cult of the Pāupatas are rightly regarded by Schultz and Ingalls as basically separate. Since the oldest extant and possibly original work of the sect, the Paśupata-sūtra, is devoted almost exclusively to ritual, it is likely that philosophy was a secondary development. Already in the Pañcārtha-bhāșya, however, Pāśupata philosophy is presented in a systematic form. Since the cult seems older we will outline its basic features first. The ritual prescriptions of the sūtras do not have a very rigid arrangement or order. Kaundinya and later writers attempted to remedy this situation and also to incorporate both theology and ritual into a single theoretical scheme. Two rather pedantic systems of classification were evolved-the first best represented by Kauņdinya and the second by the Ganakārikā. Kauņdinya divides Pāśupata doctrine into five Principal Topics (pancārthas) : Effect (kārya), Cause (kāraņa), Union (yoga), Observance (vidhi) and End of Sorrow (duhkhānta).62 These five Topics are described as the central feature of Pāśupata (or Māheśvara) doctrine in the comments of Śamkarācārya, Vācaspati Miśra and Bhāskarācārya on Brahma-sūtra ii. 2. 37. Most of what may be called the Pasupata cult falls under the heading of Observance. The system of classification set out in the Ganakārikā is somewhat more complicated. In eight short mnemonic verses it divides Pāśupata doctrine into nine primary Groups (gaņas)-eight Pentads (pañcakas) and one Group of Three. The third Pentad

58HTR, LV, 285-91. M. Hara is preparing a translation of the entire work. 59 Die philosophisch theologischen Lehren des Pāupata-Systems nach dem Pañ- cārthabhāșya und der Ratnațīkā. See also M. Hara's review in IIJ, IV (1960), 165-70. 60In his A History of Indian Philosophy, V, 1-10, 130-49. 61 Bhāskarī, Vol. III. 62 Pañcārtha-bhāșya on Pāśupata-sūtra i. 1. (p. 6).

LAKULĪŚA AND THE PĀUPATAS 185

consists of the five Stages (avasthās) in the initiate's spiritual development. These are : (1) the Marked (vyakta), (2) the Un- marked (avyakta), (3) Victory (jaya), (4) Cutting (cheda), and (5) Cessation (nisthā).63 In the Marked Stage the Aspirant (sādhaka) adopts the marks of the sect and performs certain vows. He 'bathes' himself and lies down in ashes from a funeral pyre. He wears flowers taken from an image of Siva. He lives in a temple and performs there six Acts of Worship (upahāra): laughing, dancing, singing, uttering the auspicious sound huduk (or dumdum), offering homage (namaskāra), and pious incantation (japya). All this is to be done only in the company of other Pāsupatas. In the Unmarked Stage the Aspirant leaves the temple, abandons the identifying marks of his sect, and actively encourages censure from the populace by means of several peculiar practices, notably the six so-called Doors (dvāras) : kräthana (snoring or acting as if asleep when one is not), spandana (shaking one's limbs as if afflicted by 'wind-disease'),64 mandana (walking as if crippled), śrngāraņa (making amorous gestures in the presence of women), avitatkarana (acting as if devoid of judgement), and avitadbhāsaņa (uttering senseless or contradictory words). The third, Victory Stage is characterised by victory over the senses. In the fourth, Cutting Stage, the Aspirant presumably destroys all his remaining worldly ties.65 The final, Cessation Stage marks the absolute cessation of all exertion, mental or physical, religious or profane. Each of the five Stages is associated with a particular Place (deśa), Strength (bala), Impurity (mala), Purification (viśuddhi), Procedure (upāya), Attainment (lābha), and Aspect of Initiation (dīkşākārin). These form the remaining seven Pentads. The relationship of all these items is best seen in the table on next page. The last of the nine Groups is called Means of Livelihood (vrtti). It is threefold : Alms (bhaiksya), Left-over Food (utsrsta), and Food Acquired by Chance (yathalabdha). According to the

63Most of the verses of the Ganokārika are quoted by Sayana-Mādhava in the 'Nakulīśa-Pāśupata-darsana' chapter of the Sarvadarśana-samgraha. Our trans- lations of these technical terms closely follow those given by Hara in his translation of the latter work (IIJ, II, 12-32). Hara also gives elaborate cross references to the other Pāśupata works. 64våyu-abhibhūta. The wind humour is the cause of a great number of disorders according to Hindu medical works. See J. Filliozat, The Classical Doctrine of Indian Medicine, pp. 61-79, 196-228. 65The Ratnatika does not explain this term very clearly.

186 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Stage Marked Unmarked Victory Cutting Cessation

Place With the Among Men Cave Cemetery With Rudra Guru (jana) (guhā-deśa) (šmaśāna)

Strength Devotion to Tranquility Victory over Merit Constant the Guru of Mind Opposites (dharma) Caution (Guru- ( Mati- (dvandva- bhakti) (apramāda) prasāda) jaya)

Impurity False Demerit Cause of Deviation Creaturehood Knowledge (adharma) Attachment (mithyd-jflåna) (sakti-hetu) (cyuti) (pašutva)

Purification Removal of Removal of Removal of Removal of Removal of Ignorance Demerit Attachment Deviation * Creaturehood (ajfåna-hăni) Causes

Procedure Impregnation Prescribed Pious Constant Grace with Conduct Incantation Recollection (prasāda) Doctrine (caryă) and Meditation of Rudra (våsa)66 (japa and (sadā dhyāna) Rudra-smrti)

Attainment Knowledge Penance Constant Fixedness Magical (jRăna) (tapas) Association (in Rudra) Perfection with God (sthiti) (siddhi) (deva-nityatva )67

Aspect of Material Time Ritual Divine Image Guru Initiation (dravya) (kāla) (kriyā) (mūrti)

Ratnațikā these are the only sources of nourishment approved by the āgamas. The Ratnațīkā attempts to combine this system of classification with the five Principal Topics of Kaundinya by including these Topics under the heading of Knowledge, the first of the five Attain- ments. Since the Principal Topic of Observance (vidhi) has little to do with knowledge, however, it is mainly subsumed under the Procedure of the Unmarked Stage, Prescribed Conduct (caryā).68 The most important of the five Stages are the first two, the Marked and the Unmarked. The other three seem to denote mental states as much as courses of behaviour. The most notable feature of the Marked Stage is the 'six-limbed' Act of Worship (upahāra) : laughing, dancing, singing, and so forth. As we have seen, the

66The meaning and reading of this term is not certain. See Hara, IIJ, II, 15-16, and S.N. Dasgupta, A History ... , V, 148. 67See M. Hara, 'A Note on the Sanskrit Word Ni-tya-,' JAOS, LXXIX (1959), 90-96. 68See Gaņakārikā, pp. 9-15, 17-19.

LAKULĪSA AND THE PĀŚUPATAS 187

Kālāmukha priest Vāmaśakti from the Kedāreśvara temple in Belagave was known as 'the most skilled in the world in daily performing pleasant dances.'69 This might well be a reference to the Pasupata Act of Worship. It should also be recalled that several other Kālāmukha priests are described as experts in drama and music. The curious custom of courting dishonour by disreputable behaviour during the Unmarked Stage is the most distinctive feature of the Pūsapata cult. It is described in some detail in the third chapter of the Pāsupata-sūtra and in Kaundinya's commentary thereon.70 According to these two sources the chief rationale for this behaviour is the transfer of good and bad karman. The sūtras explain it thus : 'Because of the censure of others, he gives his (accu- mulated) demerit (pāpa or adharma) to them, and he takes the (accumulated) merit (sukrta or dharma) from them.'71 Without the censure of others these actions would result simply in the increase of the performer's own demerit.72 The idea of exchanging good and bad karman, as Ingalls notes (HTR, LV, 293) is common in classical Sanskrit literature. The idea of intentionally courting dishonour for this purpose, however, is very unusual and difficult to explain. Ingalls (HTR, LV, 295-98) seeks the origin of some of the ascetics' peculiar behaviour in the beast-vows mentioned by the Jaiminiya-Brāhmaņa and later Sanskrit literature and in other techniques of spiritual possession practised by shamans in primitive societies throughout the world. Without wishing to discard Ingalls' hypothesis entirely, we believe that most of the psychological and historical foundation for these practices can be found, without going so far afield, in the dominant asceticism complex of Indian religion itself. Courting the censure of one's fellow humans is, after all, an efficient means of cutting oneself off from them, of achieving isolation and worldly detachment. Under various names this state of detachment is an essential ingredient of Jainism, Buddhism and Upanişadic Hinduism. As we have seen, Removal of Attach- ment Causes (sangakara-hāni) is one of the five Purifications in

69Sce above, p. 128. 70Most of this chapter and its commentary have been translated by Ingalls, HTR, LV, 285-91. 71 Pāśupata-sūtra iii. 7, 8 and 9. 72Kauņdinya's Pañcārtha-bhāșya on sūtra iii. 7.

188 THE KÄPĀLIKAS AND KĀLĀMUKHAS

Pāśupata doctrine. Kauņdinya himself makes it clear that the cultivation of detachment as much as the transfer of merit is the motive behind the Pāsupata's actions. Under sūtra iii. 3, 'Dis- honored (avamatah),' for instance, he quotes a verse which declares : 'For he who is despised lies happy, freed of all attach- ment.'73 His comments on sūtra iii. 11 are equally explicit. The sūtra declares: 'He should go about like an outcaste (preta).'74 Kauņdinya comments :

He should appear as though mad, like a pauper, his body covered with filth, letting his beard, nails and hair grow long, without any bodily care. Hereby he becomes cut off from the respectable castes and conditions of men, and the power of passionless detachment is produced.75

This type of idea has no place in the world of the shaman. His babblings, animal noises and so forth do cut him off from his fellow men, but both he and they regard this behaviour as a sign of his superior spiritual power. However much the shaman controls his trance, he believes that he is in communication with the spiritual world. The Päsupata's mad behaviour, on the other hand, is completely feigned and wins only contempt from ordinary men. It is possible that the ultimate source of some of his practices may be found in shamanism, but their psychological basis has changed completely. For this reason we prefer to consider the Pasupata's courting of dishonour mainly as an extension, albeit a highly original one, of the search for worldly detachment through ascetic penance.76

73Trans. Ingalls, HTR, LV, 286. 74Trans. ibid., p. 289. The usual meaning of preta is, of course 'dead person' or 'ghost.' This may well have been the meaning intended by the sūtra although Kauņdinya seems to interpret it as 'outcaste.' 75Trans. ibid., except for the passage put into italics. Ingalls translation at this point seems significantly misleading. The whole of the Sanskrit sentence reads: ato varņāśrama-vyucchedo vairāgyotsāhaś ca jāyate.' Ingalls renders the italicised passage as 'and gives rise to disgust.' In the present context vairāgya seems more likely to denote the positive quality of 'freedom from all worldly desires.' This is the meaning used by Kaundinya elsewhere in his commentary. Under sūtra i. 42, for instance, we find the compound dharma-jñāna-vairāgyaiśvaryādharmājñānā- vairāgyānaiśvaryānām. 'Freedom from all worldly desires' would naturally arise in the worshipper, not in those who see him. 76The acceptance, if not the courting, of dishonour is prescribed for Jain ascetics

LAKULĪSA AND THE PĀŚUPATAS 189

One other point about the cult of the Pasupatas should also be noted. This is the great emphasis placed by Kaundinya, and to a lesser extent by the Ratnatīka, on the ten yamas and niyamas or 'major and minor restraints.' Kaundinya defines the five yamas as non-injury (ahimsā), celibacy (brahmacarya), truthfulness (satya), non-trade (asamvyavahāra), and non-theft (asteya). He defines the niyamas as non-anger (akrodha), attentiveness to the teacher (guru-śuśrūşā), purity (śauca), abstemious diet (āhāra- lāghava), and constant caution (apramāda).77 Other Sanskrit works define these yamas and niyamas somewhat differently.78 They form the first two of five 'limbs' of Yoga named in Yoga- sūtra ii. 2979 and invariably head the similar lists of yogic virtues repeatedly attributed to Kālāmukha ascetics in epigraphy. The five yamas also appear to be associated with the epithet Mahāvratin as it is applied to Kālāmukha ascetics. According to Yoga-sūtra ii. 31, when the five yamas are maintained under all circumstances- without regard for caste, place, time, or occasion-they are called the Mahāvrata.80 The theology of the Pasupatas is a large and rather complicated subject which we do not feel qualified to discuss in detail.81 Its basic outlines are summed up in the five Principal Topics (minus the third, Observance, which denotes the cult). The first Topic is Effect (kārya). This is divided into three categories : (1) Cognition (vidyā) including various types of conscious and unconscious mental activity; (2) World and Body Parts (kalä) including the physical elements and human organs, senses and mental faculties; and (3) the Individual Soul (paśu). All Effect is said to be dependent (asvatantra). in the following passage from the Ācārānga-sūtra (trans. H. Jacobi, iv. 16. 2-3): 'A mendicant, living thus, self-controlled towards the eternal (world of living beings), the matchless sage, who collects his alms, is insulted with words by the people assailing him, like an elephant in battle with arrows. Despised by such-like people, the wise man, with undisturbed mind, sustains their words and blows, as a rock is not shaken by the wind.' 77 Pañcārtha-bhāșya on Pāśupata-sūtra i. 9 (pp. 16-33). 78See P.V. Kane, HDS, V, Part II, 1418-24. 79'yama-niyamāsana-prāņāyāma-pratyāhāra-dhāraņā-dhyāna-samādhayoșțāv an- gāni.' See also above, p. 111. 80 ahimsa-satyāsteya-brahmacaryāparigrahā yamāņ/ jāti-deśa-kāla-samayānavac- chinnāņ sārva-bhaumāḥ mahāvratam.' Yoga-sūtra ii. 30-31. See also above, p. 81. 81 The reader is referred especially to the work of F.A. Schultz, to Hara's translation of the 'Nakulīśa-Pāśupata-darśana' chapter of Sayana-Madhava's Sarvadarśana- samgraha, and to S.N. Dasgupta's, A History ... , Vol. V.

190 THE KĀPĀLIKAS AND KĀLĀMUKHAS

The second Topic is Cause (kāraņa). This is defined quite simply as God or Iśvara. The Pasupata faith is thoroughly theistic and consequently God functions as the linchpin of its metaphysical system. He is described as the creator, destroyer and supporter of the universe.82 He has two major aspects-one which is immanent and manifold (sakala) and one which is transcendent and formless (niskala) Both are characterised by unlimited Power of Knowledge (jñāna-śakti) and Power of Action (kriyā-śakti).83 Speech is incapable of expressing his formless aspect.84 In his manifold aspect, however, he is called by various names in accordance with his several attributes and functions such as patitva, sattvà, ādyatva, ajātatva, and so forth.85 In this aspect he is also said to pervade all Effect (defined as the twenty-five categories or tattvas of Sāmkhya).86 This doctrine of God as Cause must be very similar to the Iśvara- kartr-vāda taught by the eleventh century Kalāmukha priest Bonteyamuni of Hombal. Although little is known about his doctrine apart from its name, one of the stories describing the miracles he performed during his travels provides the additional information that the Kartr of his doctrine also had a formless aspect. At a great debate his opponents 'questioned him how the Kartr he defended could be formless.' In reply 'he stood invisible (became formless) for a while and made them speechless ... '87 One of the most distinctive features of Pāśupata doctrine as propounded by Kaundinya and the Ratnatīka is the belief in God's absolute independence (svatantratā).88 In practical terms this independence means that God acts without regard for human karman (karmādinirapeksa). God's will is thereby placed over and above even the moral order (dharma). Sāyaņa-Mādhava contrasts this view with that of the Maheśvaras who follow the Śaiva-darśana. They reject this Pāśupata doctrine 'because it is blemished by the faults of cruelty and injustice' and hold that 'the Supreme Lord, the Cause, (acts) in conformity with (human) karman, etc.'89 In other words, the Saiva-darsana God cannot

82 Ratnațikā in Gaņakārikā, p. 11. 83Kauņdinya on Pāśupata-sūtra ii. 27 and v. 27. 84Ibid., v. 27. See also Ratnațīkā in Gaņakārikā, p. 11. 85 Ratnațtkā in Gaņakārikā, p. 11. 86Kauņdinya on Pāśupata-sūtra ii. 5. 87Ed. Desai, SII, XV, no. 73. Translated for us by H.S. Biligiri. See also above. 88Kauņdinya on Pāśupata-sūtra v. 47 and Gaņakārikā, p. 15. 89'karmādisāpeksaḥ parameśvaraḥ kāraņam iti.' Sarvadarśana-samgraha, p. 320.

LAKULĪŚA AND THE PĀŚUPATAS 191

act arbitrarily but must reward good deeds and punish evil ones. H. Jacobi points out that the Nyāya logician Uddyotakara (c. 620), who calls himself a Pāsupatācārya, adopts the Saiva-darśana point of view on this point.90 This shows that at least one important doctrinal split had occurred in the Pāsupata sect by the seventh century. This is by no means surprising since the radical view of Kauņdinya and the Ratnațīkā strikes at the heart of nearly all Indian ethical systems, the theory of karman. Sāyaņa-Mādhava allows the Päsupatas a rebuttal to the charge that their doctrine of God as an independent Cause would lead to a situation in which 'human deeds (karma) would produce no result and all effects would be produced at the same time,'91 but their reply, at least as Sāyaņa-Mādhava presents it, is not altogether clear or convincing. It does appear, however, that they were forced to temper this doctrine somewhat, although not to such a degree that it is possible to agree with Jacobi (p. 53) that the difference between Uddyotakara and Lakulīśa (sic) is only apparent. Salvation in Paśupata doctrine is the state called End of Sorrow (duhkhānta), the last of the five Principal Topics. As is to be expected in such an uncompromisingly theistic system, it is achieved only by the grace of God. Sūtra v. 40 declares : 'He who has constant caution attains the end of sorrows through the grace of God.'92 Preliminary to this final liberation, however, is Yoga, the fourth Topic, which Kaundinya repeatedly defines as 'the union of the atman and Iśvara.'93 The soul does not become absorbed or dissolved in Iśvara or Brahman as in monistic Vedānta, but remains inseparably tied to God in the state the sūtras call Rudra-sāyujya.94 The designation of Salvation as End of Sorrow has a rather negative ring. Bhāskarācārya claims that the Pāsupatas, Vaiśeșikas, Naiyāyikas, and Kāpālikas all hold End of Sorrow and mokșa to be identical. In this condition, he adds, the ätmans are without attributes and resemble only stones.95 Much the same claim is made by Yāmunācārya.96 This does not seem to be any more true for the Pasupatas than for the Kāpalikas. It is certainly not the

90 Die Entwicklung der Gottesidee bei den Indern, p. 53. 91Trans. Hara, IIJ, II, 31. 92'apramādī gacched duņkhānām antam īśa-prasādāt.' 93See, for instance, his commentary on sūtras i. 1 (p. 6), i. 20, and v. 2. 94 Pāśupata-sūtra v. 33. 95 Brahma-sūtra-bhāșya ii. 2. 37. 96See Handiqui, p. 235.

192 THE KĀPĀLIKAS AND KĀLĀMUKHAS

view of Kaundinya or the author of the Ratnatīkā. The latter work distinguishes between two types of End of Sorrow-the Impersonal (anātmaka) and the Personal (sātmaka). Impersonal End of Sorrow does resemble the state referred to by Bhāskara and Yämuna. It is characterised only by the absolute extirpation of all sorrows. Personal End of Sorrow, however, is a state of 'Perfection (siddhi) characterized by the Power of Lordship (aiśvarya) of Maheśvara.'97 This Perfection consists of Power of Knowledge or Perception (jñāna- or drk-śakti) and Power of Action (kriyā-śakti). These two are also divided into a'number of specific superhuman abilities. The Power of Knowledge is fivefold and comprises extraordinary powers of seeing, hearing, thinking, discrimination, and omniscience. Power of Action is threefold and comprises the ability to act with the swiftness of the mind (manojavitva), the ability to assume forms at will (kāma- rūpitva) and the ability to act without physical organs (vikarana- dharmitva). In addition, the ätman who has attained Personal End of Sorrow gains ten other Perfection characteristics including such qualities as fearlessness, agelessness, deathlessness, and lordship (patitva).98 He possesses, in short, nearly all the attributes of Iśvara himself.

97 Ratnațikā in Gaņakārikā, pp. 9-10. 98Ibid. See also Sāyaņa-Mādhava's 'Nakulīśa-Pāśupata-darśana,' trans. Hara, IIJ, II, 19-21.

BIBLIOGRAPHY

Aiyangar, S. Krishnaswami. Manimekhalai in its Historical Setting. London : Luzac and Co., 1927. Agrawala, R.C. 'Two Standing Lakulisa Sculptures from Rajastan,' JOIB, XIV (1965), 388-91. Agrawala, Vasudeva S. Vāmana Purāņa-A Study. Varanasi : Prithivi Prakashan, 1964. Ānandagiri. Śamkara-vijaya. Ed. J. Tarkapanchānana. ('Bibliotheca Indica.") Calcutta : Baptist Mission Press, 1968. Ānandarāyamakhin. Jīvānanda. Ed. M. Duraiswami Aiyangar. ('Adyar Library Series,' No. 59). Adyar : 1947. Vidyāpariņayana. Ed. Śivadatta and K.P. Parab. 2nd ed. ('Kāvyamālā,' No. 39) Bombay : Nirņaya Sāgar Press, 1930. Annigeri, A.M. (ed.) Karnatak Inscriptions, Vol. IV. Dharwar : Kannada Research Institute, 1961. Āpastambīya Dharmasūtra. With Haradatta's Ujjvala commentary. Ed. Mahādeva Śāstrī and K. Rangāchārya. Mysore : Government Press, 1898. Āpastambīya Dharmasūtra. Trans. G. Bühler. Vol. II of SBE. Oxford : The Clarendon Press, 1879. Atharvaśiras Upanisad. Ed. G.S. Murti and trans. T.R.S. Ayyangar in their Śaiva Upanișads. Madras : Adyar Library, 1953. Atiya, A.S. Crusade, Commerce and Culture. Bloomington, Ind .: University Press, 1962. Bādarāyaņa. Brahma-sūtra. See Bhāskarācārya, Rāmānuja, Šamkarācārya, and Vācaspati Miśra. Balfour, Henry. 'The Life History of An Aghori Fakir,' Journal of the Anthropological Institute [London], XXVI (1897), 340-57. Bāņa. Harșa-carita. Ed. P.V .- Kane. Two parts in one volume. 2nd ed. Delhi : Motilal Banarsidass, 1965. . Harsa-carita. Trans. E.B. Cowell and F.W. Thomas. Photo reprint of 1897 edition. Delhi : Motilal Banarsidass, 1961. . Kādambarī. 'Pūrvabhāga.' Ed. P.V. Kane. 2 parts. Bombay : by the editor, 1920-21. -. Kādambarī. 'Uttarabhāga.' Ed. P.V. Kane. Bombay : Nirņaya Sāgar Press, 1913. Kadambarī. Trans. C.M. Ridding. London: Royal Asiatic Society, 1896. Banerjea, Jitendra Nath. Development of Hindu Iconography. Calcutta : University of Calcutta, 1956. 'Lakuliśa-the Founder or the Systematiser of the Pāsupata Order,' Indian History Congress, Proceedings of the Fourteenth Session, Jaipur (1951), pp. 32-6. Bārhaspati-sūtra. Ed. and trans. F.W. Thomas. Lahore: Moti Lal Banarsi Dass, 1921. Barnett, Lionel David (ed. and trans.). 'Inscriptions at Yewur,' EI, XII, 268-98 and 329-40.

194 THE KĀPĀLIKAS AND KĀLĀMUKHAS

-. (ed. and trans.). 'Inscriptions of Huli,' EI, XVIII (1925-26), 170-218. . (ed. and trans.). 'Inscriptions of Sudi,' EI, XV (1919-20), 73-112. -. (ed. and trans.). 'Three Inscriptions of Lakshmeshwar,' EI, XVI (1921-2), 31-52. -. (ed. and trans.). 'Two Inscriptions from Kurgod,' EI, XIV, 265-84. . (ed. and trans.). 'Two Inscriptions from Ron, of Saka 944 and 1102,' EI, XIX, 222-36. Barrow, H.W. 'On Aghoris and Aghorapanthis,' JAnSB, III (1893), 197-251. Basava Purāņa. Trans. G. Würth in JBBRAS, VIII (1864-6), 65-97. Basham, A.L. History and Doctrines of the Ajivikas: A Vanished Indian Religion. London : Luzac & Co., 1951. -. The Wonder That Was India. London : Sidgwick and Jackson, 1954. Baudhāyana Dharmasästra. Trans. G. Bühler. Vol. XIV of SBE. Oxford: The Clarendon Press, 1882. Beal, Samuel (trans.). Chinese Accounts of India. 4 vols. New edition. Calcutta : Susil Gupta (India) Ltd., 1957-8. (First published in 1883.) Bhandarkar, D.R. (ed.) 'An Eklingji Stone Inscription and the Origin and History of the Lakuliśa Sect,' JBBRAS, XXII (1904-7), 151-65. -. 'Lakulīśa,' Archaeological Survey of India, Annual Report: 1906-7, pp. 179-92. . (ed. and trans.). 'Mathura Pillar Inscription of Chandragupta II : G.E. 61,' EI, XXI (1931-2), 1-9. -. (ed.). 'Some Published Inscriptions Reconsidered,' IA, XLII (1913), 57-64. Bhandarkar, R.G. Early History of the Dekkan : Down to the Mahomedan Conquest. Reproduced from the Bombay Gazetteer, Vol. I, Part II (1896). Calcutta : Susil Gupta (India) Private Ltd., 1957. . (ed. and trans.). 'A Revised Transcript and Translation of a Chālukya Copper-plate Grant,' JBBRAS, XIV (1878-80), 16-28. . Vaisnavism, Saivism and Minor Religious Systems. Photo reprint of 1913 edition. Varanasi : Indological Book House, [1965]. Bharati, Agehananda. The Ochre Robe. London : G. Allen and Unwin, 1961. -. The Tantric Tradition. London : Rider and Co., 1965. Bhāsarvajña. See Gaņakārikā. Bhāskarācārya. Brahma-sūtra-bhāşya. Ed. Vindhyeśvarī Prasāda Dvivedin. (Chowk- hambā Sanskrit Series,' Nos. 70, 185 and 209.) Benares : Vidya Vilas Press, 1915. Bhattacharya, Kamaleswar. 'La Secte des Pacupata dans l'ancien Cambodge,' Journal asiatique [Paris], CCXLIII (1955), 497-90. Bhavabhūti. Mālatī-Mādhava. With Jagaddhara's commentary. Ed. R.G. Bhandar- kar. Bombay : Government Central Book Depot, 1905. Malati-Madhava. Ed. and trans. C.R. Devadhar and N.G. Suru. Poona : By the editors, 1935. al-Birūni. Tarikh al-Hind. Trans. Edward Sachau as Alberuni's India. Two volumes in one. New edition. London : Kegan Paul, Trench, Trübner & Co. Ltd., 1910. (First published in 1888.) Bloomfield, Maurice, 'On False Ascetics and Nuns in Hindu Fiction,' JAOS, XLIV (1924), 202-42. Böhtlingk, O., and Roth, R. Sanskrit Wörterbuch. 7 vols. St. Petersburg : 1855-75. Brahmānda Purāņa. Bombay : Venkațeśvara Press, 1913. Briggs, George Weston. Gorakhnāth and Kānphatā Yogīs. Calcutta : Y.M.C.A. Publishing House, 1938.

BIBLIOGRAPHY 195

Brown, Norman O. Hermes the Thief. Madison : University of Wisconsin Press, 1947. Bühler, G. (ed. and trans.). 'The Cintra Prasasti of the Reign of Sarangadeva,' EI, I (1888), 271-87. Caņdapāla. See Trivikrama-bhațța. Canna-Basava Purāna. Trans. G. Würth in JBBRAS, VIII (1864-6), 98-221. Carstairs, George Morrison. The Twice Born. London : Hogarth Press, 1957. Chakravarti, Chintaharan, 'The Soma or Sauma Sect of the Saivas,' IHQ, VIII (1932), 221-3. Tantras : Studies on their Religion and Literature. Calcutta : Punthi Pustak, 1963. Chandra, Pramod. 'The Kaula-Kāpālika Cults at Khajurāho,' Lalit Kalā, Nos. 1-2 (1955 -- 56), pp. 98-107. Chattopadhyaya, Sudhakar. The Evolution of Theistic Sects in Ancient India. Calcutta: Progressive Publishers, 1962. Chaudhuri, Nirad C. The Continent of Circe. London : Chatto and Windus, 1965, Choudhary, Radhakrishna, 'Heretical Sects in the Purāņas,' ABORI, XXXVII (1956), 234-57. Cousens, Henry. The Chälukyan Architecture of the Kanarese Districts. ('Archaco- logical Survey of India,' New Imperial Series, Vol. XLII.) Calcutta : Government of India, Central Publication Branch, 1926. Crooke, William. 'Aghori,' ERE, I, 210-13. Cunningham, A. Report of a Tour in the Punjab in 1878-79. ('Archaeological Survey of India Reports,' Old Series, Vol. XIV.) Calcutta': Office of Superintendant of Government Printing, 1882. Daņdin. Daśakumāra-carita. Ed. and trans. V. Satakopan, V. Anantacharya, and N. Bhaktavatsalam. Madras : V. Ramaswamy Sastrulu and Sons, 1963. Dasgupta, Shashi Bhusan. An Introduction to Täntric Buddhism. 2d ed. Calcutta : University of Calcutta, 1958. Obscure Religious Cults. 2d ed. Calcutta : Firma K.L. Mukhopadhay, 1962. Dasgupta, Surendranatha. A History of Indian Philosophy. 5 vols. Cambridge : Cambridge University Press, 1922-55. Desai, P.B. (ed.) South Indian Inscriptions, Vol. XV. Madras : Government Press, 1964. 'Devaņņa Bhațța. Smrticandrikā. Ed. L. Srinivasacharya. 6 vols. Mysore : Govern- . ment Press, 1914-21. Devendra Gaņī. Commentary on Uttarādhyayana. Trans. J.J. Meyer in his Hindu Tales. London : Luzac & Co., 1909. Dhanapatisūri. See Mādhavācārya. Divanji, P.C. 'Lakulīśa of Kārvān and his Pāsupata Culture,' Journal of the Gujarat Research Society, XVII (1955), 267-74. 'The Mahesvara Cult and its Offishoots,' Journal of the Asiatic Society of Bombay, XXX (1955), Part II, 6-22. Eliade, Mircea. Cosmos and History: The Myth of the Eternal Return. Translated by W.R. Trask. New York : Harper Torchbooks, 1959. (First published in 1949 as Le Mythe de l'eternel retour : archetypes et repetition.) -. Yoga : Immortality and Freedom. Translated by W.R. Trask. London : Routledge & Kegan Paul, 1958. (First published in 1954 as Le Yoga : Immortalité et Liberté.) Filliozat, Jean. The Classical Doctrine of Indian Medicine. Translated by D.R.

196 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Chanana. Delhi : Munshiram Manoharlal, 1964. (First published in 1949 as La Doctrine classique de la Médecine indienne.) Fleet, John Faithful (ed. and trans.). Corpus Inscriptionum Indicarum, Vol. III. Calcutta : 1888. . (ed. and trans.). 'Inscriptions at Ablur,' EI, V (1898-9), 213-65. . (ed. and trans.). 'Notes on Inscriptions at Gaddak in the Dambaļ Tālukā of the Dharwad District,' IA, II (1873), 296-303. -. (ed. and trans.). 'Sanskrit and Old-Canarese Inscriptions,' IA, IX (1880), 123-5. -. (ed. and trans.). 'Sanskrit and Old-Canarese Inscriptions,' IA, X (1881), 126-32. -. (ed. and trans.). 'Sanskrit and Old-Canarese Inscriptions Relating to Yādava Kings of Dēvagiri,' JBBRAS, XII (1876), 1-50. 'Śiva as Lakulīśa,' JRAS for 1907, pp. 419-26. -. (ed. and trans.). 'Sravana Belgola Epitaph of Marasimha II,' EI; V (1898-99), 151-80. . (ed. and trans.). 'A Series of Sanskrit and Old-Canarese Inscriptions Relating to the Ratta Chieftains of Saundatti and Belgaum,' JBBRAS, X (1871-74), 167-298. Gai, G.S. (ed.), South Indian Inscriptions, Vol. XX, Mysore : Government Press, 1965. Gait, Edward Albert. 'Human Sacrifice (Indian),' ERE. VI, 849-53. Gaņakārikā. Often ascribed to Haradattācārya. With Ratnațīkā commentary often ascribed to Bhāsarvajña. Ed. C.D. Dalal. ('Gaekwad's Oriental Series,' No. 15.) Baroda : Central Library, 1920. Gautama Dharmaśāstra. Trans. G. Bühler. Vol. II of SBE. Oxford : The Clarendon Press, 1879. Ghurye, G.S. Indian Sadhus. 2d. ed. Bombay: Popular Prakashan, 1964. Gokulanātha. Amrtodaya. Ed. Śivadatta and K.P. Parab. 2d ed. revised. ('Kāvya- mālā,' No. 59.) Bombay : Nirņaya Sāgar Press, 1935. Gorakşa-siddhānta-samgraha. Ed. Gopi Nath Kaviraj. ('Princess of Wales Saraswati Bhavana Texts,' No. 18.) Benares : Vidya Vilas Press, 1925. Guņaratna. See Haribhadra. Hāla. Gāthāsaptašatī. With Gangādhara's commentary. ('Kāvyamālā,' No. 21.) Bombay : Nirņaya Sāgar Press, 1889. Haldar, R.R. (ed.) 'Inscription of the Time of Hammir of Ranthambhor, Dated (V.S.) 1345,' EI, XIX, 45-52. Handiqui, Krishna Kanta. Yaśastilaka and Indian Culture. Sholapur: Jaina Samskrti Samrakshaka Sangha, 1949. Hara, Minoru. 'Pāśupata Kenkyū I,' Journal of Indian and Buddhist Studies [Tokyo], XII (1964), 57-73. -. 'Pāśupata kenkyū II-Pañcārthabhāşya ad Pāsupatasūtra I, i,' Hikata Hakushi koki kinen ronbunshū [Fukuoka: 1964], pp. (65)-(80). -. 'A Note on the Sanskrit Word ni-tya-,' JAOS, LXXIX (1959), 90-6. Review of Die philosophisch-theologischen Lehren des Pāsupata-Systems nach dem Pancārthabhāșya und der Ratnațīkā by F.A. Schultz, IIJ, IV (1960), 165-70. Haradattācārya. See Gaņakārikā. Haribhadra. Şaddarśana-samuccaya. With Guņaratna's Tarkarahasya-dīpikā com- mentary. Ed. Luigi Suali. ('Bibliotheca Indica.') Calcutta : 1905-14.

BIBLIOGRAPHY 197

Samarāiccakahā. Ed. H. Jacobi. ('Bibliotheca Indica.') Calcutta : 1926. Hauer, J.W. Der Vratya. Stuttgart : W. Kohlhammer, 1927. Hazra, Rajendra Chandra. 'The Kālikā-Purāņa,' ABORI, XXII (1941), 1-23. . 'The Smrti-chapters of the Kūrma-Purāna,' IHQ, XI (1935), 265-86. Heesterman, J.C. 'Vrātya and Sacrifice,' IIJ, VI (1962-3), 1-37. Hemacandra. Trişaşțiśalākapurușa-caritra. Trans. Helen M. Johnson. 6 vols. ('Gaekwad's Oriental Series.' Nos. 51, 77, 108, 125, 139 and 140) Baroda : Central Library, 1931-62. Hultzsch, E. (ed. and trans.) 'Sravana Belgola Epitaph of Mallishena,' EI, III (1894-95), 184-207. Hultzsch, E., and Sastri, H. Krishna (ed. and trans.). South Indian Inscriptions, Vol. III. Madras : Government Press, 1929. Indraji, Bhagvānlāl, and Bühler, G. (ed. and trans.) 'Inscriptions from Nepal,' IA, IX (1880), 163-94. Ingalls, Daniel H.H. 'Cynics and Pasupatas: The Seeking of Dishonor,' HTR, LV (1962), 281-98. Jacobi, Hermann. Die Enteicklung der Gottesidee bei den Indern und deren Beweise fur das Dasein Gottes. Bonn : K. Schroeder, 1923. Jairazbhoy, R.A. Foreign Influence in Ancient India. New York : Asia Publishing House, 1963. Jambhaladatta. Vetālapañcavimšati. Ed. and trans. M.B. Emeneau. ('American Oriental Series,' Vol. IV.) New Haven : American Oriental Society, 1934. Jātakā. Ed. V. Fausbøll. 7 vols. Photo reprint of 1st edition of 1877-97. London : Luzac and Co., Ltd., 1962-4. Joshi. Purushottam Balkrishna. 'On the 'Rite of Human Sacrifice in Ancient. Mediaeval and Modern India and Other Countries.' JAnSB III (1893), 275-300. Kalhaņa. Rājataranginī. Ed. M.A. Stein. Bombay : Education Society Press, 1892. -. Rājataranginī. Trans. M.A. Stein. 2 vols. Photo reprint of 1900 edition. Delhi : Motilal Banarsidass, 1961. Kālikā Purāņa. Chapter entitled 'Rudhirādhyāya.' Trans. W.C. Blaquiere in Asiatick Researches, V (1797), 371-91. Kane, Pandurang Vaman. History of Dharmaśāstra. 5 vols. Poona : Bhandarkar Oriental Research Institute, 1930-62. Kāņhapāda. See Shahidullah. Kāravaņa Māhāimya. Edited by C.D. Dalal in his edition of the Gaņakārikā. See above. Karmarkar, A.P. The Vrātya or Dravidian Systems. Vol. I of a projected The Religions of India. Lonavla (India): Mira Publishing House, 1950. Kathākośa. Ed. Jagdish Lal Shastri. Lahore : Shanti Lal Jain, 1942. Kathākosa. Trans. C.H. Tawney. London : Royal Asiatic Society, 1895. Kavikarņapūra. Caitanyacandrodaya. Ed. Kedāranātha and W.L. Śāstrī. ('Kāvya- mālā,' No. 87.) Bombay : Nirņaya Sāgar Press, 1906. Keith, A. Berriedale. A History of Sanskrit Literature. Photo reprint of 1920 edition. London : Oxford University Press, 1961. -. The Sanskrit Drama. Photo reprint of 1924 edition. London : Oxford University Press, 1964. Kielhorn, F. (ed.) 'Gadag Inscription of the Yadava Bhillama; Śaka-Samvat 1113,' EI, III (1894-5), 217-220.

198 THE KĀPĀLIKAS AND KĀLĀMUKHAS

-. (ed.). 'Twenty-one Copper-plates of the Kings of Kanauj; [Vikrama-] Samvat 1171 to 1233,' EI, IV, 97-129. Koestler, Arthur. The Lotus and the Robot. London : Hutchinson, 1960. Kosambi, Damodar Dharmanand. Myth and Reality. Bombay : Popular Prakashan, 1962. Krishnamacharlu, C.R. List of Inscriptions Copied by the Office of the Superintendent for Epigraphy, Madras. Delhi : Government Press, 1941. -. (ed.). South Indian Inscriptions, Vol. XI, Part I. Madras : Government Press, 1940. Krishnarao, Bhavaraj V. (ed.) 'Tandikonda Grant of Ammaraja II,' EI, XXIII, 161-70. Krşņamiśra. Prabodhacandrodaya. With Candrikāvyākhyā and Prakāśațīkā com- mentaries. Ed. V.L. Pansīkar. 6th ed. Bombay : Nirnaya Sagar Press, 1965. Kşemendra. Dašāvatāra-carita. Ed. Durgāprasād and K.P. Parab. Bombay: Nirņaya Sāgar Press, 1891. Deśopadeśa and Narmamālā. Ed. Madhusūdan Kaul Shastri. Poona: Āryabhūshan Press, 1923. Kşemīśvara. Caņdakauśika. Ed. and trans. Sibani Das Gupta. Calcutta: The Asiatic Society, 1962. Kudalkar, J.S. (ed. and trans.) 'A Note on Tilakwāda Copper-plate Inscriptions of the Time of King Bhoja Paramara of Malwa,' All India Oriental Conference, Proceedings and Transactions of 1st session, Poona, 1919, Vol. II, pp. 319-26. Kulārņava-tantra. Edited, with a rough paraphrase, by Tārānātha Vidyāratna. Madras : Ganesh and Co. (Madras) Private Ltd., 1965. Kürma Purāņa. Ed. Nilamani Mukhopadhyaya. ('Bibliotheca Indica') Calcutta: Giriśa-Vidyāratna Press, 1890. Lal, Rai Bahadur Hira, 'The Golaki Matha,' Journal of the Bihar and Orissa Research Society, XIII (1927), 137-44. Lalitavistara. Ed. P.L. Vaidya. ('Buddhist Sanskrit Texts,' No. 1.) Darbhanga : Mithila Institute, 1958. Linga Purāņa. Ed. Jīvānanda Vidyāsāgara. Calcutta : Nūtana Vālmīki Press, 1885. Lüders, H. (ed.) 'Gadag Inscription of Vira-Ballala II; Saka-Samvat 1114,' EI, VI (1900), 89-97. Mādhavācārya. Samkara-digvijaya. With Dhanapatisūri's Dindima commentary. (Ānandāśram Sanskrit Series,' No. 22.) Poona : Ānandāśram Press, 1915. Mahābhārata (critical edition). Vols. III-IV .: Aranyakaparvan. Ed. V.S. Sukthankar. Poona : Bhandarkar Oriental Research Institute, 1942. Mahābhārata (critical edition). Vols. XIII-XVI : Sāntiparvan. Ed. V.S. Sukthankar and S.K. Belvalkar. Poona : Bhandarkar Oriental Research Institute, 1949-54. Mahābhārata (critical edition). Vol. XI: Salyaparvan. Ed. R.C. Dandekar. Poona : Bhandarkar Oriental Research Institute, 1961. Mahalingam, T.V. 'A Family of Pāsupata Grhasthas at Jambukēśvaram,' JORM, XV (1957), 79-85. . 'The Pasupatas in South India,' JIH, XXVII (1949), 43-53. Mahanirvāna-tantra. Ed. and trans. John Woodroffe as The Great Liberation (Mahānirvāņa Tantra). 3d ed. Madras: Ganesh & Co. (Madras) Ltd., 1953. Mahendravarman. Mattavilāsa. Ed. T. Gaņapati Sāstrī. ('Trivandrum Sanskrit Series,' No. 55.) Trivandrum : Government Press, 1917. Mattavilāsa. Trans. L.D. Barnett in BSOS, V (1930), 697-710.

BIBLIOGRAPHY 199

Maitrāyaņīya Upanīsad. Ed. and trans. J.A.B. van Buitenen. The Hague : Mouton and Co., 1962. Majumdar, B.P. 'Lakulīśa Pāsupatas and their Temples in Medieval India,' JBRS, XXXIX (1953), 1-9. Majumdar, Ramesh Chandra (ed.). The Age of Imperial Kanauj. Vol. IV of The History and Culture of the Indian People. Bombay : Bharatiya Vidya Bhavan, 1955. . (ed.). The Age of Imperial Unity. Vol. II of same series. Bombay : Bharatiya Vidya Bhavan, 1951. -. (ed.). The Classical Age. Vol. III of same series. Bombay : Bharatiya Vidya Bhavan, 1954. . (ed.). The Struggle for Empire. Vol. V of same series. Bombay : Bharatiya Vidya Bhavan, 1957. Manu-smrti. Trans. G. Bühler. Vol. XXV of SBE. Oxford: Oxford University Press, 1886. Matsya Purāņa. ('Ānandāśram Sanskrit Series,' No. 54.) Poona: Ānandāśram Press, 1907. Mehta, R.N. 'Karavan-The Seat of the Lakulisa Sect,' Indian History Congress, Proceedings of the Fourteenth Session, Jaipur (1951), pp. 71-76. Mitra, Rajendralāla. 'On Human Sacrifices in Ancient India,' JRASB, XLV (1876), 76-118. Monier-Williams, Monier. A Sanskrit-English Dictionary. Photo reprint of 1899 Ist edition. Delhi : Motilal Banarsidass, 1963. Mysore Gazetteer. Newedition edited by C. Hayavadana Rao but mostly based on B.L. Rice's original edition. 5 vols. Bangalore: Government Press. 1927-30. Nandimath, S.C. A Handbook of Virasaivism. Dharwar : L.E. Association, 1942. Narasimhachar, R. (ed. and trans.) Epigraphia Carnatica, Vol. II (revised edition). Bangalore : Mysore Government Press, 1923. Narasimhaswami, H.K. (ed.) 'Bhairavakonda Inscription of Vikramaditya,' EI, XXXIII (1959-60), 78-81. -. (ed.). 'Dommara-Nandyala Plates of Punyakumara; 10th Year,' EI, XXVII (1947-8), 268-76. Pañcatantra. Ed. Nārāyaņa Rāma Āchārya. Bombay: Nirņaya Sāgar Press, 1959. Panchamukhi, R.S. (ed.) 'Karnatak Inscriptions, Vol. I. Dharwar : Kannada Research Office, 1941. Pandey, Kanti Chandra. Bhāskarī. Vol. III. ('Princess of Wales Sarasvati Bhavana Texts,' No. 84.) Lucknow : Superintendent, Printing and Stationery, U.P., 1954. Panigrahi, Krishna Chandra. Archaeological Remains at Bhubaneswar. Bombay: Orient Longmans, 1961. Pāsupata-sūtra. With Kauņdinya's Pañcārthabhāșya commentary. Ed. R. Anantha- krishna Sastri. ('Trivandrum Sanskrit Series,' No. 143.) Trivandrum : University of Travancore, 1940. Pāśupata-sūtra. Variant readings of several sūtras edited by C. Chakravarti in IHQ, XIX, 270-1. Patanjali. Yogasūtra. Ed. and trans. J. Ballantyne. 4th reprint. Calcutta : Susil Gupta (India) Private Ltd., 1963. Pathak, V.S. History of Saiva Cults in Northern India from Inscriptions. Varanasi : Dr Ram Naresh Varma, 1960.

200 THE KĀPĀLIKAS AND KĀLĀMUKHAS

.Ancient Historians of India-A Study in Historical Biographies. Bombay : Asia Publishing House, 1966. Pillai, J.M. Somasundaran. Two Thousand Years of Tamil Literature. Madras: By the author, 1959. Pires, Edward A. The Maukharis. B.G. Paul & Co., 1934. Raghavan, V. (ed. and trans.) 'Tiruvorriyur Inscription of Chaturanana Pandita : 20th Year of Krishna III,' EI, XXVII (1947-8), 292-303. Rahamāna, Abdula. Samdeśa-rāsaka. Ed. and trans. Śri Jina Vijaya Muni. Bombay : Bharatīya Vidyā Bhavan, 1944. Rājaśekhara. Karpūramañjarī. Ed. Sten Konow and trans. Charles Rockwell Lan- man. ('Harvard Oriental Series,' Vol. IV.) Cambridge, Mass. : Harvard Uni- versity, 1901. Rāmacandra. Kaumudī-mitrānanda. Ed. Muni Punyavijaya. Bhavanagar: Jaina Atmananda Granthamāla, 1917. Nalavilasa. Ed. G.K. Shrigondekar and Lalchandra B. Gandhi. ('Gaekwad's Oriental Series,' No. 29). Baroda : Central Library, 1926. Rāmānuja. Šrībhāşya. Ed. and trans. R.D. Karmarkar. 2 vols. Poona : University of Poona, 1959-62. Rangacharya, V. Inscriptions of the Madras Presidency. 3 vols. Madras : Government Press, 1919. Rangaswamy, M.A. Dorai. The Religion and Philosophy of Tēvāram. 4 vols. in 2 books. Madras : University of Madras, 1958-59. Rao, N. Lakshminarayan (ed.). South Indian Inscriptions, Vol. XI, Part II. Madras : Government Press, 1953. Rao, R. Rama. 'Origin and Development of Siva-Worship with Special Reference to Vīrasaivism,' QJMS, IV (1924), 190-210 and 282-301. Rao, T.A. Gopinatha (ed.). 'The Huzur Treasury Plates Belonging to the Vișnu Temple at Tiruvalla,' TAS, II (1910), 131-207. Elements of Hindu Iconography. 2 vols. Madras : The Law Printing House, 1914-16. Renou, Louis, and Filliozat, Jean. L'Inde classique. 2 vols. Paris : Payot, 1947-53. Rice, B. Lewis (ed. and trans.). Epigraphia Carnatica. 12 vols. Bangalore : Mysore Government Press, 1886-1904. Sakhare, M.R. History and Philosophy of Lingāyat Religion. Belgaum : Mahavir Press, 1942. Śaktisangama-tantra. Ed. Benoytosh Bhattacharya. 3 vols. (Gaekwad's Oriental Series,' Nos. 61, 91 and 104.) Baroda : Central Library, 1932-47. (An intended final volume was never published.) Saletore, Bhasker Anand. Mediaeval Jainism : With Special Reference to the Vijaya- nagara Empire. Bombay : Karnatak Publishing House, [1938]. Śamkarācārya. Brahma-sūtra-bhāsya. Published together with Vācaspati Miśra's Bhāmatī, Āmalānanda-Sarasvatī's Kalpataru, and Appayadīkșita's Parimala. 2d ed. Ed. Bhārgav Sāstrī. Bombay : Nirņaya Sāgar Press, 1938. (1st edition edited by Anantkrişņa Śāstrī.) Brahma-sūtra-bhāysa. Trans. G. Thibaut. Vols. XXXIV and XXXVIII of SBE. Oxford : Oxford University Press, 1904. Sankalia, Hasmukh D. The Archaeology of Gujarat. Bombay : Natwarlal and Co., 1941. Śankhadhara. Lațaka-melaka. Ed. Durgāprasād. 3d ed. ('Kāvyamālā,' No. 20.)

BIBLIOGRAPHY 201

Bombay : Nirņaya Sāgār Press, 1923. Sansom, G.B. Japan : A Short Cultural History. Revised edition. New York : Apple- ton-Century-Crofts, Inc., 1943. Sarkar, Jadunath. A History of Dasnami Naga Sanyasis. Allahabad : Śri Panchayata Akhara Mahanirvani, 1958. Sastri, K.A. Nilakanta. The Colas. 2 vols. Madras : University of Madras, 1935-7. . The Colas. 2d ed. revised. Madras : University of Madras, 1955. . Development of Religion in South India. Bombay : Orient Longmans, 1963. . A History of South India. London : Oxford University Press, 1958. -.. (ed. and trans.). 'The Kodumbāļūr Inscription of Vikrama-Kēsarī,' JORM, VII (1933), 1-10. . (ed.). The Mauryas and Satavahanas. Vol. II of A Comprehensive History of India. Bombay : Orient Longmans, 1957. Sastry, R. Shama, and Rao, N. Lakshminarayan (ed.). South Indian Inscriptions, Vol. IX, Parts I and II. Madras : Government Press, 1939-41. Şat-cakra-nirūpaņa. With Kālācarana commentary. Ed. and trans. J. Woodroffe as The Serpent Power. 7th ed. Madras : Ganesh & Co. (Madras) Private Ltd., 1964. Sāyana-Mādhava. Sarvadarsana-samgraha. Edited with a Hindi translation by Uma Shankar Sharma. Varanasi : Chowkhamba Vidyabhawan, 1964. . Chapter of Sarvadarsana-samgraha entitled 'Nakulīsa-Pāsupata-darśana.' Trans. Minoru Hara in IIJ, II (1958), 8-32. Sayre, Farrand. Diogenes of Sinope : A Study of Greek Cynicism. Baltimore : [Printed by J.H. Furst Company], 1938. Schultz, F.A. Die philosophisch-theologischen Lehren des Pāsupata-Systems nach dem Pañcārthabhāșya und der Ratnațīkā. Walldorf-Hessen : 1958. Shahidullah, M. (ed. and trans.) Les Chants Mystiques de Kānha et de Saraha; les Dohā-koșa et les Caryā. Paris : Adrien-Maisonneuve, 1928. Shāstri, Dakshinaranjan. 'The Lokāyatikas and the Kāpālikas,' IHQ, VII (1931), 125-37. Śiva Purāna Vāyavīyasamhitā. Ed. Mallikārjunaśāstrī. 2 vols. Sholapore: Datta- prasäda Press, 1905-6. Skanda Purāņa. 7 vols. Bombay : Venkateśvara Press, 1909-11. Skanda Purāņa Sūtasamhitā. With Mādhavācārya's commentary. Ed. V.S. Paņa- śīkara. 3 vols. ('Ānandāśrama-samskrita-granthāvaliḥ,' No. 24.) Poona: 'M.C. Apte, 1893. Somadeva. Kathāsaritsāgara. Ed. Durgāprasād and K.P. Parab. Bombay : Nirņaya Sagar Press, 1889. Kathāsaritsāgara. Trans. C.H. Tawney as The Ocean of Story. 10 vols. London : Chas. J. Sawyer, 1924-8. Sreenivasachar, P. (ed. and trans.) A Corpus of Inscriptions in the Telingana Districts of H.E.H. the Nizam's Dominions. Vol. XIII, Part II, of HAS. Hyderabad : 1940. Śrīharşa. Naisadhacarita. Translated, with extensive notes and extracts from several commentaries, by Krishna Kanta Handiqui. 2d ed. revised. ('Deccan College Monograph Series,' No. 14.) Poona : 1956. Subandhu. Vasavadatta. Ed. and trans. Louis H. Gray. Reprint of 1913 edition. New York : Ams Press Inc., 1965. Subbiah, A. Venkata. 'A Twelfth Century University in Mysore,' QJMS, VII (1917), 157-96.

202 THE KĀPĀLIKAS AND KĀLĀMUKHAS

Thapar, Romila. Asoka and the Decline of the Mauryas. London : Oxford University Press, 1961. Trivikrama-bhatța. Nalacampū. With Caņdapāla's Vişamapadaprakāśa. Ed. Durgā- prasād and Sivadatta. 3d ed. Bombay : Nirņaya Sāgar Press, 1931. Tucci, Giuseppe. 'Animadversiones Indicae,' JRASB, n.s. XXVI (1930), 125-60. Umāpati. Śivaprakāsam. Trans. Henry R. Hoisington as 'Siva-Pirakāsam, -Light of Siva' in JAOS, IV (1854), 125-244. Upadhyay, Vasudeva. Socio-Religious Condition of North India (700-1200 A.D.). ('Chowkhamba Sanskrit Studies,' Vol. XXXIX.) Varanasi: Chowkhamba Sanskrit Series Office, 1964. Vācaspati Miśra. See Šamkarācārya. La Vallée Poussin, Louis de. 'Tantrism (Buddhist),' ERE, XII, 193-97. Vāmana Purāņa, Bombay: Venkațeśvara Press, 1908. Varāhamihira. Brhajjātaka, Ed. and trans. V. Subrahmanya Sastri as Brihat Jataka. Mysore : Government Branch Press, 1929. -. Brhajjataka. With Utpala's commentary. Bombay : 1863. -. Brhatsamhitā. Ed. H. Kern. ('Bibliotheca Indica.') Calcutta : 1865. -. Brhatsamhitā. Trans. H. Kern in JRAS, n.s. IV(1870), 430-79; n.s. V(1871), 45-90 and 231-88; n.s. VI (1873), 36-91; and n.s. VII (1875), 81-134. (Trans- lation only up to chapter 134). . Brhatsamhitā. 2 vols. Ed. and trans. V. Subrahmanya Sastri. Bangalore : V.B. Soobbiah and Sons, 1947. Vāsișțha Dharmasūtra. Trans. G. Bühler. Vol. XIV of SBE. Oxford : The Clarendon Press, 1882. Vāyu Purāņa. ('Ānandāśram Sanskrit Series,' No. 45.) Poona : Ānandāśram Press, 1905. Venkayya, V. (ed.) 'Triplicane Inscription of Dantivarman,' EI, VIII, 290-96. Vidyābhūșana, Satis Chandra. 'Influence of Aristotle on the Development of the Syllogism in Indian Logic,' JRAS for 1918, pp. 469-88. Vinayacandra. Mallinātha-carita. Ed. Hargovinddas and Bechardas. Benares : Harshchand Bhurabhai, 1911. Vișņu Purāņa. Trans. H.H. Wilson. 3d ed. Calcutta : Punthi Pustak, 1961. (First published in 1840.) Vişnu-smrti. With Nandapandita's Keśavavaijayanti commentary. 2 vols. Adyar: Library and Research Centre, 1964. Vișnu-smrti. Trans. J. Jolly. Vol. VII of SBE. Oxford : The Clarendon Press, 1880. Vyas, Akshaya Keerty (ed.). 'Paldi Inscription of Guhila Arisimha, V.S. 1173,' EI, XXX, 8ff. Vyāsācala .. Śamkara-vijaya. Ed. T. Chandrasekharan. Madras : Government Press, 1954. Watters, Thomas. On Yuan Chwang's Travels in India. 2 vols. London : Royal Asiatic Society, 1904. Yājñavalkya-smrti. With Vijñāneśvara's Mitākșara commentary. Ed. N.R. Āchārya. 5th ed. Bombay : Nirņaya Sagar Press, 1949. Yaśaņpāla. Moharājaparājaya. Ed. Muni Chaturavijaya. ('Gaekwad's Oriental Series,' No. 9.) Baroda : Central Library, 1918. Yazdani, G. (ed.) The Early History of the Deccan. 2 vols. London : Oxford University Press, 1960. Zimmer, Heinrich. The King and the Corpse. New York : Pantheon Books, 1948.

INDEX

The following abbreviations are used; Āpastambīya Dharmasūtra, 75-76. 165 Kp: probable Kālāmukha priest and Appar, 23, 49 Kt : probable Kālāmukha temple. ārādhya, 125. 151 Ariñjīśvara, Kt, 165 Abhayacandra, 148 Aristotle, 179 Abhinava-Someśvara, Kt, 155 Amodara, 11-12 Ablūr, 45, 98, 100, 106, 110-11, 115- āsana, 111n, 156 16, 118, 120, 126, 130, 157, 169 Asanga, 82 Acaleśvara, Kt, 114 ashes, 2, 4-6, 15, 21, 23, 40, 43, 47, 64, Act of Worship (upahāra), 185-87 68-70, 152, 185 Ādūr, 140, 157 Asitänga, 84-85 āgama, 11, 109-110. 125-27, 142-43, Asitavaktra, 10, 166 151, 171, 186 Assam. 53, 86 Agastya, Kp, 130-31 aştānga-yoga. 127-28 Agastyeśvara, Kp, 154 Aśvapāda, 66, 94 Aghora, Kp, 150 Atharva Veda, 28, 103, 152 Aghoraghanta, 50, 56-57, 85, 95 Atharvaśiras Upanişad, 183 Aghoraśakti, Kp, 159 ātman (Self), 2. 34, 58, 85, 92, 95, Aghorī, 53 191-92 Ahavamalla Deva, 123 Avadhūta, Kp. 131 Aitareya Brāhmaņa, 86 āvali, or āmnāya. 98-99 Ajayapāla (Caulukya). 49n, 52 Ājīvika, 14-15, 45, 105 Bāļacandra, Kp. 140 Ajñānarāśi, 62 Bāla-Sivācārya, Kp, 154 Akalańka, 26, 112, 148 Bālasūrya or Trailocana, Kp. 155 ākhādā (religious regiment), 46 Balegrāma, 27 Akulavīratantra. 82 Baleśvara, Kt, 144 akriyā, 36 Ballāla I, 158 alchemy, 18, 58 Balleśvara, Kt, 152 Amarakośa, 73, 82 Bamma-jīya, Kp, 161 Amaravațeśvara, Kt, 142-43 Bammarāśi, Kp, 160 Amarāvatī, 133, 143, 165 Bammēśvara, Kt, 161 Ambale, 29 Ammarāja II (E. Cālukya), 141-43 Bāņabhațța, 16-23, 51, 183 Banerjea, J.N., 173, 175 Amrtarāśi, Kp, (a) 130; (b) 135; (c) Bārhaspati-sūtra, 68, 88 154; (d) 159 Basava. 167-71 Ānandagiri, xi-xii, 9, 31-32, 39, 42-46, Basava Purāņa, 168. 171 83,89 Basham, A.L., xiv, 105 Ānandarāyamakhin, 49, 82, 88 Bațuka or Bațuka-nātha. 37. 39, 42, 47 Anantaśakti, Kp, 153 Baudhāyana Dharmašāstra, 175-76 Andhāsura, Kt, 150-51 Beal, S., 16 Andhra, 17, 39, 51-52, 97 Apastamba, 11 Bēgūr-agrahāra, 140 Belagāve or Baļļigāva or Balipura. etc ..

204 INDEX

xii. 81. 97. 99-100. 109-11. 113-16, al-Bīrūnī. 109 120-29. 131. 133, 138-40, 146, 148-49. Bodholbaņa-nityānanda. See Krakaca 153. 156. 170-71. 179 Bonteyaguru. Kp, 131 Beļleya-santāna. 98-99. 130-31, 135- Bonteyamuni, Kp, xiv, 45. 131-35, 36 143. 149, 170, 173, 190 Bengal. 52. 71. 86 Boppeśvara, Kt, 150 bhaga. 2-4 Brahmā (god). 11, 20, 60, 77-81. 85. Bhairava (god). 4. 16. 20. 34-35. 37. 90.124, 174 41-43. 46-47. 61, 76, 84-89. 91, 156, Brahmadatta. (a) 64; (b) 134 167 brahmahan and brahmahatyā. 13-14. Bhairava (sect). 8-9. 11 70-71, 74-81 Bhairavācārya. 20-22, 63n Brāhmaņa (varņa). 6, 28-30, 38, 41- Bhairavāgama. 40-41, 43. 88-89 42, 47-48. 50. 55. 60, 63-64. 67-69. Bhairavananda. 49 74-81. 104, 114. 124. 131, 145, 149. Bhairava-tantra, 40 161-64, 167, 175 -- 77 bhakti. 83-95. 107 Brāhmaņas (texts). 79. 81. 86 Bhandarkar, D.R .. 174, 177. 179-80 Brahmāņda Purāņa, 24 Bhandarkar, R.G., xii-xiii. 3, 5. 148. Brahma-sutra and commentaries. 1-8. 174-75. 179, 182-83 171. 182-84 Bharadvaja. 11, 18 Brahmeśvara (now Basaveśvara), Kt. Bharata or Bharata-muni, 121. 126 99.106. 111. 113, 118. 126. 130, 169-70 Bharati. A .. 3, 89-90 Buddhism, xi. 3-4. 7. 13-15, 25-26. 45, Bhasarvajña. 134. 175, 183 47, 50. 54-55. 59. 61. 68-69, 82, 91. Bhāskarācārya. 1, 7. 180, 182. 184, 94.104-105. 112. 115-16, 124. 148. 156 191-92 Bühler, G., 174. 177 Bhavabhūti. xi-xii. 2. 17. 19. 22-23. 44.49-51. 56-57. 95 Caitrāpura, Kt. 130 Bhaujanga or Bhuvana or Trilocana. Kp. Cālukyas (Western of Kalyāņa), 51, 155-56 117-19, 122-23, 130. 133, 136-37. Bhayilēśvara. Kt. 135 143-44, 147-48, 154-55, 157-58 Bhikșu (sect). 14. 59. 61 Cāmeśvara, Kt, 138 Bhillama. 119 Cāmuņdā, 30, 57, 85 Bhīșaņa. 84 Caņda, 84 Bhoja (Paramāra), 28. 93 Candāla. 55, 58-59, 65. 68 Bhrgu. 176 Caņdamārī, 76 Bhrgukaccha. 176-77, 179 Caņdapāla, 73 bhrūnahan. 75-76 Caņdikā. 16-18, 51. 87 Bhubaneshwar. 30. 52 Candra, 85 Bhujanga-devācārya, Kp. 140. 155-57 Cāņdupaņdita. 82 Bhujangāvali. 98-99, 108. 138-40. 156 Canna-Basava Purāņa. 168-69 Bhuvana-karttāra. Kp. 158 Candrabhūşaņa. Kp, (a) 101, 103. 119. Bhuvanarāśi, Kp, 142 120-23. 171; (b) 153; (c) 159 Bijāpur. 108. 140. 146. 155-57 Candrasvāmin, 63 -- 64, 94 Bijjala. 123-26. 129-30. 157. 167-69 Cārvāka. 47. 50. 59, 82 Bijjalarāyacarita. 168 caste (jāti and varņa), 6-7, 47-48, 90- Bilhaņa. 114, 133 92, 135, 149, 171, 188 Biligiri. H.S .. xiv. 132n Caturānana. Kp, (a) 83. 166-67; (b) Billa-matha. Kt. 135-36 140 Bīreya-jīya. Kp. 150 Cause (kāraņa). 134, 173. 184, 190

INDEX 205

celibacy (naişthika and brahmacāra). Digambara. 15. 25. 50. 59-60, 62. 82. 58. 76. 81. 92, 121, 127, 139, 145, 148. 148 150, 157-58, 176, 189 Dinakara, 28 Cikkadēva, Kp, 145. 147 Diogenes, 179 Cikkakavi-jīya, Kp, 159 Divyaśakti, Kp, 154-55 Cōļas. 39. 165-66 Dombi, 7,69-70, 77 crossroads. 16-17 Doors (dvāra). 139, 185 Cuttack. 133 dualism (dvaita). 1-2. 35-36. 84. 134. Cynic. 178-79 162 Dugga-Tribhuvanamalla (Kākatīya). Dadhīca or Dadhyañc. 33, 91-92 136 dakșiņācāra, 48. 90 Dakşiņa-Someśvara, Kt. 153 Effect or Material Universe (karyā). Dalal, C.D., 183 134, 173, 184. 189-90 Daņda (god), 180 ekadaņda and Ekadaņdin. 14-15, 104- Daņdin. 22-23 105 Dantidurga. 26 Ekāmbiranātha, 23. 50 Daśanāmi Nāgā. 46 Ēkāntada Rāmayya, 45-46, 169-70 Dasgupta, S.N., 170-71, 184 Ēklingjī, 176-77 Deccan. 18, 20, 52 Eliade, M., 16 Dēkavve, 159 End of Sorrow (duḥkhānta). 134, 184. detachment (vairāgya. etc.), 38, 187-88 191-92 Deussen. P .. 183 Devadatta. 64, 94 Fleet, J.F., xii, xiv, 28, 99, 103, 112. Devangiri, 130 117-18, 168-69, 174, 177-79 Devaņna Bhatta. 68 Devapāla, 65 Gadag, 100, 103, 116, 119-23. 131. Devarāśi, Kp, 130 157,170-71 Devarāya I and II. 163 Gaņakārikā, 134, 171, 175, 181, 183-84 Devaśakti, Kp, (a) 137; (b) 150; (c) 153 Gandhi. M.K., 39n Devasinga-jīya, Kp, 157 Ganga or Ganges, 63-64, 167 Devaśiva, Kp, 153 Gangādevī. 163 Devasomā, 54-55 Gańgaraśi. Kp, (a) 137;(b) 159;(c) 161 Devavrata. Kp, 126 Gargya or Garga. 109, 181-82 Devendra Gaņī, 64 Gauladeva. Kp. 146-47 Devendraśakti, Kp, 159 Gauri, 50, 92. 121 Dhanada, 11 Gautama-muni, Kp, 101, 116-19. 123 -- Dhanapatisūri, 31-32, 42-43, 46 Dhānyakațaka, 133, 143 Gavareśvara, Kt. (a) 153; (b) 157 dhāraņā. 111, 115, 124, 156 Ghoraghona or Meşamukha. 62 dharma. 60, 92. 126, 142, 149. 176. Ghurye. G.S., 5. 46. 105 186-87.190 Gīrvāņaśakti, Kp. 158 Dharma (god). 58. 61 Gogga, 98, 136 Dharma-kattāra, Kp, 158 Gōjēśvara. Kt, 159 Dharmarāśi, Kp. (a) 146-47; (b) 158; Gokulanātha. 49. 73. 82 (c) 159 Gomadatta Aruļāļa Bhațțan. Kp. 165 Dharmaśiva, Kp. 124 Gorakhnāth or Gorakșa-nātha. 35. 37 Dharmeśvara. Kt. 160 44 dhyāna, 111, 115, 124, 156, 186 Gorakşa-siddhānta-samgraha. 31-32.

206 INDEX

35-38, 42-43 173, 190 Goravas, 115 Honnaya. Kp. 151 gotra. 11, 149, 160, 164-66, 181 Hospitallers, 45 Gough, A.E., 183 Govardhanēśvara, Kt., 138 Hoysalas, 119. 123. 127, 135. 138-39. 158,160 Govinda III, 160 Hsüan Tsang. 15-16. 182 Govinda-candra (Gāhadavāla), 30, 52 Hūli or Pūli, 150-51, 170 Govindarasa (daņdanāyaka), 111, 113 Human flesh (oblations and sale of). -14 16-17. 21. 56-57, 60. 62, 76. 85-87 grace (prasāda), 157, 186, 191 Huvina-hadagalli, 144 Grāmeśvara, Kt, (a) 130; (b) 152 Greece, 178-79 Immadi Bukka. 162 guggulu. 16-17, 27-28 Indo-Greeks. 179 guild, 118, 151 Indra. 33. 85, 106. 115, 121 Guņaratna, 68, 88, 134, 180-81 Ingaleśwar, 146 Gupta, 108, 175, 179, 182 Ingalls, D.H.H., xiii, 178-79. 184, 187 Gurubhakta, Kp, 138 Iśa. 32. 88. 159 Iśānarāśi. Kp, 145, 147 Habbeśvara, Kt, 151 Iśānaśakti, Kp, (a) 138; (b) 157 Hāla, 13 Iśvara. 42n, 60, 106, 173, 190-92 Halagondi, 144 Iśvaradāsa, Kp. 142, 160 Hammīra (Cāhamāna), 29 Iśvara-kartrvāda or Issarakāraņavāda hamsa and paramahamsa ascetics, 104 or Kartrvāda. etc .. 131-32. 134-35, 105 170, 173, 190 Haļebidu or Dorasamudra, 138, 158 Haļe-Nidnēgila, 100, 129, 157 Itțige-samtati. 98-99, 138

Handiqui, K.K., xii, 19, 105, 174, 184 Jacobi. H .. 191 Hanungal, 130 Jagaddhara, 73 Hara (god), 11, 60, 106, 124, 135 Jagadekamalla, (I) 119; (II) 157 Hara, M., 174, 184 Jagadēkamallēśvara. Kt. 124 Haradatta, (a) 75; (b) 171 Jagadīśvara, Kt, 152, 171 Haraśakti, Kp, 135 Jagateśvara. Kp, 160 Hari, 60, 124 Jagateśvara. Kt. 160 Haribhadra, 65, 68. 134, 180 Jaimini, 115 Harihara II, 162-63 Jaiminīya Brāhmaņa. 74, 187 Harihareśvara, Kt, 135-36 Jain and Jainism, xi, 14-16. 24-27, 39 Hariścandra, 37, 57-59, 86. 94 45, 47, 49-50. 52. 54. 59-61. 64-65. Harivarman. 82 68, 82. 104-105. 112, 116, 143, 148. Harşavardhana, 16-17, 20, 22, 51, 182 168-69 Hatha Yoga. 94-95 Jairazbhoy, R.A., 178 Hațhayogapradīpikā, 91 Jaitugi I, 146 Hauer. J.W .. 74 Jālapāda, 64 Hemacandra. 64 Jaleśvara, 176 Hercules. 178 Jambhaladatta. 65. 94 Himalayas. 78. 99-100. 104-105. 109. jangama. 9. 121, 153. 168. 171-72 170 japa. 111, 115, 124. 186 Himaśitala. 26 Jayasimha. (I) 32-33; (II) 117-48 Hiriya-mațha. Kt. (a) 149; (b) 150 Jhañjheśvara. Kt. 156 Hombal. xiv. 130-31.149. 157. 170-71. Jina. 45.65. 115. 124. 169

INDEX 207

Jinasena. 26 Kāņhapāda, 4, 7. 52, 69-71. 77 Jişņugupta. 30 Kankāla, 8, 10-11 Jñānarāśi. Kp, 145-47 Kannara. 169 Jñānaśakti. Kp, 107-108; (a) 101, 107. Kānphațā. 2, 31, 35-38, 44 45, 53 118; (b) 101, 127, 129; (c) 118, 140; Kāpāla or Kāpālika or Kapālin. etc .. (d) 138; (e) 150-51; (f) 153; (g) 155 xi-xiii, 1-11, 13-71. 73-95. 107. 110. jñānaśakti (Power of Knowledge), 161. 164, 173. 182, 191 190. 192 Kāpāla or Kāpālika vow (See also Jñāneśvara, Kp. 144, 147 Mahāvrata). 14, 22. 54. 63, 73. 77 Jogēśvara, Kt, 140 Kapāla-kuņdalā. 2. 50. 56-57, 95 Kapālamocana, 29-30. 77-79 Kabbina-Sirivūr, 130 Kapāleśvara or Kapālīśvara or Kapālin. Kailāsa. 21, 32, 88 etc. (god and temples). 20, 23, 27-32. Kālabhairava, Kp. (a) 155-56; (b) 158 40-43. 62. 73-74. 80-81, 84-85 Kalacuris, 123. 126-27, 129-30, 138, Kapālika (village), 29 145-46, 157. 167-68 Kāpālika-tantra. 75 Kalacuris of Tripuri. 52. 62 Kapālimangalam. 29 Kāladamana, 9-11 Kāpālinī (god). 30 Kāļajñāni-vakkhāņi-jīya. Kp. 122 Kāpila. 25. 68. 156 Kālakācārya. 14-15 Karahada, 132-33 Kālāmukha or Kālamukha. etc .. xi- Karālā. 50. 56-57 xiii. 1. 3-12, 19-20, 44 45, 51. 67-68. Kāravaņa Māhātmya. 176-77 73, 81-82, 97-172, 173, 179. 182-83. karman. 41. 84, 170. 187, 190-91 187, 189-90 Karņāța or Kamņāțaka. 27. 32. 39-40. Kālāmukha-dīksita. Kp, 160 43-44. 52. 97, 108. 114. 161, 173 Kalanana. 10 Kārtavīrya IV (Rațța), 151 Kālapriya. 44 Karttāra. Kp. 158 Kāļaśakti. Kp. 160 Kāruka-siddhāntin. 1. 182 Kālāsya. 10-12 Kāruņika-siddhāntin. 1. 7, 182 Kāļeśvara, Kt, 140 kaşāva and Kāșāyin. 13-14. 24 Kalhaņa. 51.66-67. 94 Kashmir. 52, 66-67. 94, 105-106. 108- Kālī, 16, 47, 65, 81n. 167 109, 114, 157. 162, 164. 179 Kalideva. Kt. (a) 144; (b) 152; (c) 152 Kāśmīra. Kp. 105. 108; (a) 108. 140. Kalideveśvara-Svayambhu, Kt. 147- 155. 157; (b) 139-40 48 Kaśmīra-deva. Kp. 155 Kālikā Purāņa. 86 Kațaka. 132-33, 143 Kalinga, 23. 141 Kāthaka-siddhāntin, 1. 7. 182 Kali-yuga. 24, 67. 156, 176 Kathā-kośa. 65 Kallēdēva, Kt, 159 Kathmandu. 30 Kalsi. 139 Kaula. 49. 61n. 62 Kalyāņa. 130. 133. 137 Kaulāvalinirņaya. 89 Kalyāņaśakti. Kp. (a) 138; (b) 150; Kaumāra-vyākaraņa. 103. 130-31 (c) 159; (d) 159: (e) 160 Kaundinya. 81. 134. 175-76. 183-84. Kāmadeva (? Kadamba), 148 186-92 Kammateśvara. Kt. 160 Kauruşa. 181-82 Kampana II. 163 Kavikamņapūra. 49 Kaņāda. 115. 177 Kāyāvataraņa. Kāyāvarohaņa. Kāyāro- Kāñcī. 4. 23. 26. 39. 50. 52. 54. 167 haņa· Kāyāvatāra. Kārohaņa. or Kār- Kane. P.V .. 94 vān. 175-76. 181

208 INDEX

Kedāraśakti. Kp, (a) 100-102, 106. Kula-kundalinī Śakti. 69 112. 114, 116, 126, 171; (b) 126; (c) 131 Kulārņava-tantra. 6-7. 89 Kedāreśvara or Dakşiņa-Kedāreśvara. Kumāra-Bontēśvara. Kt. 131 Kt. 97-98. 100. 102-104. 106-107. Kumāra-deva. Kp. 154 110-29. 147. 156. 170-71. 187 Kumārapāla (Caulukya). 49n. 52 Kedārnāth or Kedāra Mountain. 99- kuņdala. 2. 13. 21, 69-70 100.104-105. 109. 170, 173 Kuppațūr. 100. 123. 129-30 Keith. A.B .. 148 Kūrma Purāna. 8. 11. 24. 181 Kern. H .. 15 Kuşāņas. 178 Keśirāja or Keśava. etc .. (daņdanāyaka) Kuśika (Mathurā priest), 179-80 124-27 Kuśika I and II (successors of Lakulīśa). Kēta-jīya. Kp. 159 175.180-82 Khajuraho, 139 Kusuveśvara. Kt. 102 Khaleśvara. Kp. 144. 147 Kuyibāl. 140. 157 khatvānga. 2. 5. 14. 56. 63. 66. 75. 77. 177 laguda or lakula or lakuța. etc .. 5. 108. Kīrttivarman (Candella). 52 176-77 Kiruvada. 130 Lakaśipu or Paśupati I, Kp. 141-43 Kōdiya-mațha or Kedāra-mațha. etc .. Lakkunda. 29 Kt. 81. 103. 107. 111-12, 119. 123- Laksmaņa temple. 139 27. 131. 135, 153 Lakshmeśvar or Puligere. 130-31. 133. Kodumbāļūr. 4. 10. 165-66 157 Koestler. A .. 94 Lākula or Lāguda or Lākura or Nākula Kogaļi and Morigeri. 143-44. 147 or Vākula. etc .. 5. 8-11. 102. 104- Kollāpura or Kolhāpur. 132-33 105. 107. 111. 113. 137. 154 Kollāpura-mahālakșmī. 133 Lākulāgama or Lākuliśvarāgama, 110. Koppina-vakkhāņi-deva, Kp. (a) 122; 116.125. 137. 153-54. 156-57. 159-60 (b) 122, 138 Lakuleśvara. Kp. 144. 147 Kotīśvara, Kt. 100. 123. 129-30 Lakulīśa or Lakulīśvara or Nakulīśa. Koyil Tēvarāyanpețțai. 4. 165 etc. (Pasupata founder). xiii. 2-3. Krakaca or Bodholabaņa-nityānanda. 81. 102, 106-109. 114. 126, 129. 32. 39-47. 53, 83-85. 88-89, 91. 94 134-35. 137. 142-44. 146. 160, 173. Krātheśvara, 11. 12 175-81.191 Krishņā River. 32, 133, 141, 143, 168 Lakulīśa-deva. Kp. 151 Kriyāśakti, Kp. 107; (a) 101. 122-23; Lakulīśvara, Kp. (a) 108. 157; (b) 148- (b) 131; (c) 136; (d) 149; (e) 150;(f) 152 49, 179; (c) 155; (d) 165 kriyāśakti (Power of Action), 161, 190. Lalitavistara. 14 192 Lambodara or Kosthaka or Bhasmaka. Kriyāśaktis of Vijayanagar. Kp. 161- 62 64 Lenkeśvara, Kt. 143 Krodha. 84 Lévi. S .. 183 Krsna (god). 38. 176. 178 linga. 51. 103-104. 109. 121, 126. 131. Krsņa II. 27 135. 143. 150. 153, 156. 164, 167. Krsnamiśra. xi. 49-50. 59. 82-83. 94 171. 176. 179-80 Kşapaņaka. 47.59-61. 104-105 Linga Purāņa. 9. 176, 181 Kşemendra. 67 liquor. 3. 5-6. 24. 38. 41, 42n. 43. Kşemīśvara. 49-50. 52. 57 47-48. 54-55. 60-61. 67. 87-90 Kşīrasvāmin. 73. 82 logic (tarka-vidyā. etc.). 106. 110-12. Kubera. 18.63 115. 127-28. 132. 148. 150. 152. 154

INDEX 209

Lōkābharaņa, Kp. 140 Mallikārjuna. Kt. (a) 51. 99. 132, 136- Lokāyata or Lokāyatika. 9. 47. 68. 37; (b) 100. 129; (c) 138; (d) 139; 112.148 (e) 140; (f) 153 Mallişena-Maladhārideva. 25 -- 27 Macesvara. Kt. 161 Managoli. 146 Madanamañjari, 63, 94 Maņeśvara, 28 Madhava-bhatta. 148 Māņikeśvara, Kt, 151 Mādhavācārya or Vidyāraņya. xi -- xii. Manimekhalai. 23-24 25, 31-32, 34-35. 38-39. 42-44. 46. 51. mantra 16. 18. 21. 41-42. 57. 63-65. 91. 162 68.92-94. 128. 156 Madhava-mantrin, 162-64 Mantragupta. 22-23 Mādirāja, 168-69 Marapa. 163-64 Mahabhairava, 60-61. 85 Mārasimha III (W. Ganga). 24-25 Mahābhārata. 51, 55. 79. 91. 174. 177. Marked ( vyakta). 185-86 179 matha. xi. 103, 105, 117-19, 123-26. Mahābodhi Jātaka, 134 130. 137, 143-44. 154. 158, 160. 162. Mahādeva (god), 149 165 -- 67, 170-71 Mahādeva-jīya. Kp, 159 Mathura, 109, 166, 179, 181 Mahākāla, 17-18, 21-22, 44 Mātrkā or Mātr. 16-17.20 Mahākāla-hrdaya, 21-22, 63n Matsya Purāņa. 51. 77 Mahanirvana-tantra, 4, 7,89 Mattikote. 98. 138. 140 Mahāpāsupata (god). 11-12 maunānuşthāna. 111, 115. 124 Mahāpāśupata (sect). 9, 12. 18-19. Mauryas. 179 107, 151, 173 meat eating. 3-4. 38. 55. 62. 67n. 80. Mahavrata. xiii. 19. 58, 73-82, 87-88, 88-89 164, 166. 189 Meru. 59. 92. 136 Mahāvrata-dhara or Mahāvratika or Mēvuņdi. 83 Mahavratin, etc .. 8-12. 17. 19, 27- Mihireśvara. 28 28, 64. 68. 73. 76. 82-83, 87, 107. 129. militaristic ascetics. 24-25. 41-46 151. 165-67, 173. 189 Mīmāmsā. 104. 112. 115-16. 148. 154. Mahavratin (god). 11-12 56 Mahendravarman. 19, 23. 27. 49-50. monism (advaita) 35-36. 50. 59. 148. 53. 88. 110, 183 162, 191 Maheśvara. Kt. 165 mudrā, 2-3, 67n. 69.89 Maheśvara (god). 14-15. 90. 192 Mūlasthāna, Kt. (a) 130; (b) 150 Māheśvara (sect). 1. 21. 134. 179-80. Mūlasthana Siva. Kt. 160 184.190 Mūlasthana Vosavanteśvara. Kt. 130 Mahesvaradatta. 65 Mulugunda. 130. 154 Mahīpāla I (Pratīhāra). 52 Munavalli. 6. 110. 123. 152. 171 Mahmūd of Ghaznī. 108-109 Murugan. 53 Maiļaladēvī. 137 Muttagi, 108. 140. 157 Maitrāyaņīya Upanișad. 13 Muttinakeyya-Indreśvara, Kt, 131 Maitreya. 181-82 Mūvara-koņeya-samtati. 98-130. 169 Māleśvara, Kt. 138 Mylapore. 23. 29 Maleyāļa, Kp. 144-47 Maleyāļa-Jñānarāśi, Kp, 146-47 Nāgarāśi. Kp. (a) 151; (b) 160 Malleśvara. Kt. (a) 152; (b) 154 Nagareśvara or Nakhareśvara. Kt. (a) Mallideva, 29 100. 111. 114-15: (b) 102, 111. 113 Mallikārjuna. Kp. (a & b) 131; (c) 166 Nāgavardhana (early Cālukya). 27

210 INDEX

Nairātmyā. 69-70 Parśvanātha-caritra. 65 Nānarāsi-paņdita, Kp, 165 Parvatāvali. 51. 98-138 Nandi-Hill. 142. 160 Pārvatī, 77. 83. 91. 99. 128 Nar Nārāyaņ. 88 Pāśupata, xi-xiii. 1. 3-5. 7-12, 15. Nārādīya Purāņa. 18-19, 23-25, 30. 54-55. 66-68. 81. Narakāpālin, 62. 80 83. 94. 107-10. 126. 133-35. 139. 146. Nārāyaņa. 15. 78 149, 161-63, 170-71. 173-92 Nāstika. 62. 68. 88 Pāśupata-sūtra. 81. 107. 134, 139. 175. Näth Siddhas. 35. 37 183-84, 187-88 Natha. 37-38 Paśupati (god), 2. 21. 30. 107, 110. 134 Nēsargi. 107. 151-52. 173 Paśupati. Kp. (a & b) 142 Nirañjana. Kp. (a) 153; (b) 166 Patañjali (grammarian). 175 Nirāśraya. 12. 19 Patañjali (Yogasūtra). 81-82. 93-94, Nirgrantha. 14. 24 104-05. 111. 189 Nirmand. 28 Pathak, V.S .. 174-75, 180 Nirvāņa. Kp. (a) 131; (b) 154 Periplus, 179 niyama. 82. 111. 115. 124-25. 127. 156. Pinākapāņi. Kp. 135 160,189 Pires, E.A., 28 Nolambeśvara. Kt. (a) 143-44; (b) Prabhūtarāśi, Kp, (a & b & c) 142-43 154-55 Prajñā. 4, 70 Nrsimha, 34-35. 38 prāņāyāma. 68, 95. 111, 113, 156 nyāsa. 95 pratyāhāra. 111, 113, 156 Nyaya and Naiyāyika. 104. 110. 112- Pravarasena, 66-67 13. 115-17. 122. 134-35. 137. 148. Principal Topic (pañcārtha) 184, 186. 156, 179. 191 189. 191 Nyāyaśāstra. 113 Prōla I (Kākatīya). 136 Nyāva-sūtra. 179 prostitution (devadāsī, sūļeyar. veśyā. etc.). 115, 121-2. 127, 139 Observance ( vidhi ) 134.184. 186, 189 Pūliya. Kp. 122 Pūliyadeva, Kp. 151 Padmapada or Padmapāda. 34-35. 42 Purāņas, 7-12, 24. 29, 57. 73, 77, 79. Padmāvatī, 44. 50. 56 80, 82, 84, 104, 154, 171 Pālas. 71 Pure Land (Jodo) sect, 45 Pallavas. 24-26, 49. 158. 167 Pușpabhūti, 20-22 Pañcalinga. Kt. 109. 124. 148-49 pañca-makāra. 3. 67n. 89-90. 92 Rācamalla I, 153-54 pañca-mațha. Kt, 124, 149 Raghavan. V .. 167 Pāñcarātra. 8, 174. 179 Raghūttama, 82 Pañcatantra. 139 Rahmān, Abdul, 66 Pandey, K.C .. 184 Rājāditya, 166 Paņdyas. 39.45. 166 rājaguru, 122-23, 125-27. 129, 143, Panigrahi, K.C .. 30 152,163 Pāņini. 175 rājanīti. 121-22 Pāņinīya. 103. 126 Rājaśekhara. 9. 49. 52. 135, 180-81 Parameśvara. 104. 113 Rāma Rāghava, 79 Parameśvarī, 24 Rāmacandra. 49-50, 52, 62 paramātmāgama. 106.110 Rāmacandra-rāya (Seuna). 100 Paraśurāma. 38 Rāmaidēva or Rāmaiya-deva. Kt. 130 parişad or parșe. 98-9. 141. 146 Rāmanātha. Kt. (a) 130: (b) 159

INDEX 211

Rāmānuja. xi-xiii, 1-8. 44, 92. 105. 62. 67-68. 70. 83. 86. 88-92. 134. 107. 109-110. 122. 135, 139. 173, 182 156-57. 184. 191-92 Rāmaśakti. Kp. 159 samādhi. 34, 93. 111. 115. 124. 156 Rāmeśvara. Kp. 51. 99. 132, 136-37 Samasta-bhuvanāśraya. Kt. 141-42 Rāmeśvara, Kt. (a) 130: (b) 144 Sāma Veda. 103, 131, 152 rasa (sentiment). 53-54, 56-57 Sāmavedi, Kp. 130-31 Rāsțrakūțas, 26-27. 160 Sambandar or Tiruñānasambandar or Rāśikara. 175 Ñānasambandar. 23. 29. 45 Ratnacandra. 65 Samhāra or Samhāra-Bhairava. 42. 84 Ratnațīkā (ascribed to Bhāsarvajña), -85 175, 180. 183, 186, 189-92 Śamkara (god). 28. 74. 152 Rāyamurari-Sovi-Dēva or Someśvara. Śamkara Miśra. 19 126,146 Śamkarācārya. xi. 1. 27. 31-48. 51. Revaņārya. 170-71 80. 85. 88. 94. 130. 134. 162. 180. 184 Rg Veda, 103. 152 Sāmkhya. 104. 112-13. 115-16. 148. Rice. B.L., xii. xiv. 26, 100. 126 156, 174. 190 Riśīśenga. Kp. 151 Sāmkhyāgama. 113 Roņ, 138, 157 Rucikarațīkā. 49 samnyāsa and samnyāsin. 6. 35-36. 42n, 46 Rudra, 58. 85, 142. 180. 186. 191 samtati or santāna. 98-99. 136. 142. Rudrabharaņa. Kp. 101-102 144. 154. 156 rudrākșa. 6. 21, 152 Samudrasena. 28 Rudraśakti. Kp, (a) 123, 129; (b) 131; samyama. 93 (c) 131. 134. 171; (d) 136; (e) 138; sandha-terminology. 3. 70 (f) 147; (g) 149; (h) 151-52; (i) 160 Sankalia. H.D., 174 Ruru. 84 Sankama II. 138. 145 Sankama-Deva. 127 Śabara. 16-17. 68 Śankhadhara, 49-50. 61 Śābara-tantra. 37. 42 Sāntara, 119 Śabdānuśāsana. 103 Śāntarāśi, Kp. 144, 147 sacred thread. 2. 22 Sārasvata, 114. 149 sacrifice. 17. 32. 41-43, 57. 60. 65. 69. Sarasvatī, (god). 82. 106-107. 124. 157 76. 80.85-87 Sarasvatī River. 79. 92 Sahajiya Buddhism. 4.69.71 Śarvavarman. 28 Śailarāśi. Kp. 165 Sarveśvara. Kp. (a) 123, 130; (b) 123. Śaiva or Śaiva-siddhānta. xiv. 1. 5-12. 152-53. 171; (c) 124, 149; (d) 158 18. 23, 25. 103, 105. 107. 109-11. 135. Sarveśvara, Kt. 138 164, 173. 180, 182, 190-91 Sastri, K.A.N .. 169 Śaivāgama. 6. 110, 150 Sastri, R.A., 183 Śākațāyana. .103 Śatapatha Brāhmaņa. 86 Śākta. 86. 133 Sātēnahaļļi. 135, 157 Śakti. 56. 61. 70. 89-90. 92 Satyasoma. 53-55. 62 Śakti-parişad. xii. 97-141. 143. 146. Saugata. 47, 82 155-58.170 Sauma and Saumya. 5. 8. 10-11. 82 Śaktisangama-tantra. 9.90 Saudāminī. 22, 50. 57 Śakuni. 177 Sāyaņa-Mādhava, 113. 129. 134. 162. sallekhanā. 24-25 180, 183-84. 190-91 salvation (duhkhānta. mokşa. mukti. nir- Sayre. F .. 178 vāņa. etc.). 2. 4. 41, 43. 47-48. 60- Schultz. F.A .. 184

212 INDEX

sect. xi Somasiddhānta (fictional character). Śekkiļār. 68 53, 59-61. 82. 87-88, 95 sexual activity. 13. 41. 43. 47-48. 54- Someśvara. (I) 51. 137-38. 140, 143. 55, 61. 67. 74, 83, 88-92. 139. 185 158; (II) 137-38, 155, 157; (III) 117-18 shaman. xiii, 187-88 Someśvara, Kp, (a) 100-103, 110-17, Shravan Belgola, 24-25 120, 124-26, 149. 156; (b) 143-44. 147; siddha or Siddha. 16. 23. 35. 37-38. (d) 152; (e) 155 59. 66. 71, 92 Someśvara. Kt. 150 Siddhanātha. Kt. 130 Someśvara Deva (king). 123, 146 Siddha-siddhānta-paddhati. 10 Somnāth. 109, 176 Siddhēśakti, Kp, 131 Śrighaņteśvara. 28 Siddheśvara, Kt. (a) 150; (b) 151 Śrīharșa. xiii, 82 siddhi. 22, 33. 38, 51, 61. 63, 65, 77. 80. Śrīkaņțha. Kp. (a) 101-102, 106-107. 83. 92-95, 156, 186, 192 110-11. 114, 116, 126; (b) 101, 129; Simhana-Deva or Singhana-Deva. 129 (c) 149 Simha-parișad. 97, 141-46, 153, 155 Śrīkaņțha (god), 106. 174-75 Sindas. 129, 138. 153 Srinivas, M.N., 16 Sinope. 179 Śrīparvata or Śrīśaila, 18-19, 23. 32. Sirasangi. 107. 152, 173 50-52, 57, 66-67, 99, 109, 121. 132-33. Śiva Purāņa. 8 136-38, 143, 153, 155. 161, 165-66, Śiva-bhāgavata, 175 170,173 Śiva-dharma-purāņa. 154 Śrīpati Paņdita. 171 Śivāgama. 64 Śrīrangam, 4, 107 Śivalińga-deva, Kt. 157 Śrī-Svayambhū-Siddeśvara. Kt, 155 Śivapura, (a) 136; (b) 138 śrngāraņa, 139, 185 Śivarāśi. Kp. 135 Stage (avasthā). 185-86 Śivaśakti. Kp. (a) 147; (b) 159; (c) 160 Subandhu. 51, 126 Siva-śāsana, 10, 175 Subbiah, A.V., xii, 105. 108 Skanda. 53, 126 Sudhanvan. 39-41 Skanda Purāņa, 8-11. 24 Sūdi, 100-111, 114-115, 120. 157 skull and skull bowl, 2. 4-6. 13-15. Śūdra, 6, 11-12, 47 17-18. 20, 24-25, 27. 30, 32-33, 38. Sundarapāņdya-īśvara. Kt, 166 40-43. 47. 54-55. 60-61. 63. 66. 70. Suprabhedāgama, 10 75.77-81.88 Sureśvara, Kp. 51, 137 Sobbēśvara. Kt, 161 Sūrili. 135 Soma (Kadamba), 150 Sūrya. 85. 131, 135 Soma or Someśvara (god). 83. 90 svādhyāya. 111, 115, 124 Soma, Soma-siddhānta and Soma-sid- Svarņņādri-mahodaya. 30 dhāntin. xiii. 8-11. 30 49, 73, 82-83, Svayambhū (god), 24. 121, 142 90 Svayambhū. Kt, (a) 130, 154; (b) 153- Somadeva. Kp. 151 Somadeva ( Kathāsaritsāgara). 17. 62, 54; (c) 158 Svayambhū or Svayambhu-Somanātha. 64.73 Kt. 145 Somadeva ( Yaśastilaka). xii, 17, 68. 76 Svayambhu Purāna. 9 Somakeśvara. 11-12 Svayambhū-Bhairavadeva. Kt. 137 Somakhadduka. 30. 83 Svayambhu-Kedāreśvara, Kt. 146 Somanātha. Kt. (a) 126; (b) 146 Svayambhū-Nagareśvara. Kt. 137 Somanātha (Ablūr temple). 169 Somarāśi. Kp. 160 Svayambhū-Someśvara, Kt. 157 Svayambhū-Triyambakadeva, Kt.163-64

INDEX 213

Taila III. 123 Unmarked (avyakta). 185-87 Tāņdikoņda, 141 Unmatta (god). 84 Tantra and Tantrism. xiii. 3-7, 9. 11. Unmatta-Bhairava. 32. 46-48. 53-54. 16-23, 28. 30. 35-40, 48, 50-53. 62. 88.91 67. 70, 73, 76-77. 82. 86, 89-90. Upanișads, 60. 164 93-94. 110, 128. 138 Upāya (god). 70 Tārā. 89 Uśana-samhitā 68 Tatpurușa, Kp. (a) 137; (b) 150 Utpala. 14 Tāvaregere or Tāvareyakere. 102-103, 111.113 Vācaspati Miśra, 1. 7. 112. 180, 182. Tejonidhi. Kp, 158 184 Templars, 45 Vādibhasimha. 148 Tilakwādā, 28 Vādirāja, 26, 148 Tiļuvaļļi, 130 Vāgīśa Bhațța, 83 Tirumakūdal-Narsipur Taluk, 24-25 Tiruvālīśvara, Kt, 165 Vaidyarāj, V.S., 92 Vaiśeşika. 104, 110. 113. 115-16, 134- Tiruvalla, 29 35, 154, 156, 177, 179, 191 Tiruvānakkoyil, 4, 165 Vaișņava, xi, 6, 39. 46. 50. 59. 105. 125 Tiruvorriyur. 83, 166-67 Vaitāl, 30 Tiruvorriyūr Purāņam Vajrayāņa, 4, 91 Trailokya-karttāra. Kp, 158 Vakkhāņideva, Kp .. 151 Trailokyaśakti, Kp. 159 Vaļabha, 166 Tribhuvanaśakti, Kp. (a) 137; (b) 159 Vālmīki-muni, Kp, 100-102 tridanda and Tridandin, 14, 104-105 Vālmīki-rāśi, 109 Trikūțeśvara or Svayambhū, Kt. (a) 100, 116. 119. 121-23, 170-71; (b) 151 Vāma (sect), 8, 11 vāmācāra. 22. 48-49 Trilocana. Kp. (a) 138-39; (b) 140. Vāmadeva. Kp. 154 155-56; (c) 154 Vāmana Purāņa. 9-12. 18. 20 Tripurāntaka, Kt. 103, 109, 138-40. Vāmaśakti. Kp. (a) 101, 116-18; (b) 147 101, 110, 116. 119. 123-29, 187; Tripurāntaka (Pāśupata priest), 109 (c) 101. 129; (d) 131; (e) 151; (f) 151; Trivikrama-bhatta. 65, 73 (g) 153; (h) 155; (i) 159 Tucci, G., 82 Vāmeśvara. Kp. 142 Varadarāja, 19 Udayana. 19, 82 Varāhamihira, 14-15. 183 Uddyotakara. 19.1 Varanasi, 52, 58. 64. 77, 109, 134, 161. Uditācārya. 179-80 183 Ugra (god). 32, 87-88 Vardhamāna. 82 Ugra-Bhairava. 32-39. 43-44, 53. 80, Vareśvara. Kp. (a) 138-40; (b) 155; 87-88,94 (c) 157 Ujjayinī or Ujjain, 16-17. 39, 43 -- 44. varņāśramadharma. 6-7. 24, 149 51-52. 63, 175, 177 Varuņa, 86 Ulkāpuri or Avākhal, 176-77 Vāsişțha Dharmasūtra. 76 Ullrich, H., xiv, 151n Vaśiştha Purāņa. 9 Ulūka, 176-77 Vātsyāyana, 82, 121-22 Umā, 41. 43, 83, 86, 90, 174 Vāyu Purāņa. 24. 176, 179. 181 Umāmaheśvara, Kt. 141 Vēdal, 4, 11, 165 Umāpati, 68 Vedānta. 33, 46, 55. 60. 94, 127, 150. Union (yoga), 134, 184, 191 152, 191

214 INDEX

Vedas, 2, 11, 88, 103, 105, 114, 131. Viraśaiva or Lingāyat. 43n, 45. 51. 110. 142-43, 149, 154, 156, 161, 174 121. 125-26. 135. 149, 163. 167-72 Veļūr, 29 Virūpākșa, Kt. 144 Venkațācaleśa, 39 Vişņu. 12, 29. 34. 38, 61, 66, 78. 85. 90. Vetāla, 17, 21, 58-59, 63. 65 125. 163, 176 Vidarbha (place and dynasty). 39-40. Vişņu-smrti. 74. 76. 81 43, 52, 56 Vişnuvardhana. 138 Vidyābharaņa, Kp, 101, 116-20, 123 Viśvakarmāvatāra-vastu-śāstra. 177 Vidyādhara, 21, 64 Viśvāmitra Kauśika. 58-59. 152 Vidyādhararāja (Bāņa), 160 Vrātya, 74 Vidyāraņya-kālajñāna, 162 Vrddha or Vrddha-śrāvaka. 14-15 Vidyeśāna, Kp. 151 Vyāsa. 93 Vijaya-bhūpati, 163 Vijayāditya II (E. Cālukya), 141-42 Yādavas, 119, 129, 146, 169 Vijayameśvara, Kt, 152 Yājñavalk ya-smrti. 13. 75, 93 Vijayanagar dynasty, 162-64 Yajur Veda. 50, 103. 152, 164 Vijayanagar (place). 161-64 Yakşa, 63. 131 Vijayawāda, 141, 143, 165 yama. 81, 111, 115, 124-25. 127. 156, Vijñāneśvara, 75 189 Vikramāditya VI, 114, 122, 130, 133. Yama. 85, 148 136-38, 144-45, 147, 158 Yāmunācārya, xi-xii. 1-2. 4-8. 11. 44. Vikramäditya (Sinda), 138 105, 191-92 Vikrama-kēsarī of Kodumbāļūr. 165- Yaśahpāla. 28.49-50.62. 80. 88 66 Yēwūr. 144 Vilāsavatī, 16 Yoga. 36, 38, 60. 70, 87, 91-95, 108. Vimalacandra, 25-26 110. 117, 127-28, 132-33, 135, 150-51, Vinayacandra. 65 154.156. 174. 189 Vindhya, 16, 52, 66 Yogaśāstras. 81. 104, 111 Vīra Pāņdya, 166 Yogeśvara or Yogīśvara. Kp, (a) 155 -- Vira-Ballāla II, 119-21, 123. 127, 135. 57; (b) 157 160

APPENDIX A

NEW DATA ON THE KĀPĀLIKAS

The early Indian "ascetics" known as the Kāpālikas can be regarded as archetypal in at least two senses.1 In the first place, in the eyes of their opponents, whose writings have so far been our principal source of information on Kapālika beliefs and practices, the Kāpālikas have served as the archetype, or stereotype, of an immoral and heretical ascetic. In Sanskrit literature, Kāpālika characters mock Vedic doctrines while ' they either prepare a human sacrifice or drink and carouse with comely female disciples. In the second place, however, some of these same sources suggest that the Käpalikas in fact ritually modelled their lives on a divine archetype, on the god Siva-Kapālin who must endure a lengthy penance to atone for the sin of having cut off one of the five heads of Brahma. This symbolic re-enactment of Siva's Great Vow (mahāvrata) earned the Kāpālikas their title of Mahāvratins. This doubly archetypal character of the Kāpālikas, or at least of their traditional portrait in Sanskrit literature, has made them apt objects of modern academic analysis from a variety of structuralist, symbolicist and archetypicist points of view. Examples of such analysis can be found in the present book, in Wendy O'Flaherty's discussion of the Siva-Kapalin myth in her The Origins of Evil in Hindu Mythology, and in two articles on Bhairava by Elizabeth-Chalier Visuvalingam.2

1. This appendix is a revised version of an essay published in Alf Hiltebeitel (ed.), Criminal Gods and Demon Devotees: Essays on the Guardians of Popular Hinduism (Albany: State University of New York Press, 1989). 2. See chapter III above; W. O'Flaherty, The Origins of Evil in Hindu Mythology (Berkeley: University of California Press, 1976); E.C. Visuvalingam, "Bhairav: Kotwal of Varānasi," in V.P. Verma (ed.), Vārānasī Through the Ages (Varanasi: Bharatiya Itihas Sankalan Samiti, 1986), pp. 241-60; and her "Bhairava's Royal Brahmanicide: The Problem of the Mahabrahmana," in Alf Hiltebeitel, Criminal Gods .... See also H. Stietencron, "Bhairava," Zeitschrift der Deutschen Morgenländischen Gesellschaft, Supplementa I, Teil 3 (1969), pp. 863-71; S. Kramrisch, The Presence of Śiva (Princeton: Princeton University Press, 1981); S. Gupta, D.J. Hoens and T. Goudriaan, Hindu Tantrism (Leiden: E.J. Brill 1979); and J. Parry, "Sacrificial death and the necrophagous ascetic", in M. Bloch and J. Parry (eds.), Death and the Regeneration of Life (Cambridge: Cambridge University Press, 1982), pp. 74-110.

216 APPENDIX A

Recently Alexis Sanderson has claimed to have identified a corpus of unpublished Saivite tantric texts which describe Kapalika worship and observances. The most important of these texts is the Jayadrathayāmala, manuscripts of which are preserved in the national archives of Ncpal and have been photographed by the Nepal-German manuscript preservations project. Sanderson has yet to publish a full descriptive analysis of this material, but some preliminary discussion of it is available in his recently and soon to be published writings.3 From these it remains unclear to what extent the material can be directly attributed to the Kāpālikas as opposed to Kāpālika-like trantric cults such as the Kaulas.4 Most so far published source material relevant to the Kāpālikas found in Indian literature and epigraphy is discussed in the earlier chapters of this book. The most important literary omissions are from Tamil sources. First is the story of Sirnttonda, contained in the Tamil Periya Puranam, a text of the twelfth century.5 This legend relates how a "Bhairava" ascetic came to the house of the devotee Siruttonda and demanded to be fed the flesh of Siruttonda's son. The father dutifully killed his son and had his wife, the boy's mother, cook the flesh. When the mcal was

3. Alexis Sanderson, "Purity and Power among the Brahmans of Kashmir", in M. Carrithers et al. (eds.), The Category of the Person (Cambridge: Cambridge University Press, 1985), pp. 190-216; "Mandala and Agamic Identity in the Trika of Kashmir," in Mantras et diagrammes rituels dans l'hindouisme (Paris: Editions du Centre National de la Recherche Scientifique, 1986), pp. 169-207; review of two āgama texts edited by N.R. Bhatta in Bulletin of the School of Oriental and African Studies, 48 (1985), pp. 564-68; entries on "Krama Saivism", "Saivism in Kashmir" and "Trika . Saivism", in M. Eliade (ed.), Encyclopedia of Religion (New York: Macmillan, 1987); and "Saivism and the Tantric Traditions", in Stuart Sutherland (ed.), The World's Religions (London: Coom Helm Press, forthcoming). Sanderson kindly supplied me with this information and copies of most of the material. Two other possible Kāpālika manuscripts-Kāpālikamatavyavastha and Somasiddhänta-are listed in G. Kaviraj's catalogue, Tāntrik sāhitya (Lucknow: Hindi Samiti, 1972). A yāmala text that has been publishcd, the Rudrayāmalam (ed. Bhagīrathaprasādatripāthi [Varanasi: Sampurnanand Sanskrit Vishvavidyalaya, 1980]), contains much material mentioning terrific forms of Siva-Bhairava and Durga-Kāli, skulls, bones, Khatvangas and the like, but nothing I have been able to detect that specifically has to do with the Kāpālikas. 4. On the Kaulas, see the recent attempt to organize and decipher their doctrines and literature by Mark S.G. Dyczkowski, The Canon of the Saivagama and the Kubjika Tantras of the Western Kaula Tradition (Albany: State University of New York Press, 1988). 5. See M.A. Dorai Rangaswamy, The Religion and Philosophy of Tevaram, 4 vols. in 2 (Madras: University of Madras, 1958-59), p. 1018; and G. Vanmikanathan (trans.), Periya Puranam (Madras: Sri Ramakrishna Math, 1985), pp. 353-64.

APPENDIX A 217

served the ascetic revealed himself to be the god Siva and returned the son to life. Although the text does not specifically call the ascetic a Kapālika, his character is obviously congruent with the Kāpālika archctypc. Second is the story of Tirilokacolan, a king of the Cola country, and a group of Brahman Kāpalikas from Nepal. This story is found in the late medicval Perūrppurānam and has bcen translated by David Shulman.6 In this tale the Kāpālikas successfully persuade the king to adopt a life of hedonistic self-indulgence with the following argument:7 "Your Highness, our home is in Nepal, but we have come here after hearing of your fame and the richness of your country. This is our knowledge: the true source of joy in the world is sexual union with beautiful women. Some say that the world is created by a god, but thcy are fools; has any one ever seen such a god? There is no immortal deity; only the body is lasting and real. The world is created and increases through sexual union, which is the finest thing there is; and it is enhanced by drinking wine. So discard all this talk of salvation, O great king!" The king accepts their arguments and abandons "the Vedic way" with disastrous results for his kingdom. The god Brahma eventually sends the sage Närada to bring the king back to his senses. The Kāpālikas then flee the land. Given the rather late date of this text, it seems unlikely that the Kāpālikas' specch is more than a stereotyped argument against heresy. The fact that they are portrayed as materialists as well as hedonists also suggests that the author of the passage did not really know much about the actual historical Kapalikas. On the other hand, the fact that they are identificd as Brahmans and as having come from Nepal is noteworthy. At the Icast, this reflects an awareness that Nepal was by this time the chief stronghold of extreme forms of tantric religion and that this rather unorthodcx religious current-somewhat like Jewish-Christian gnostic cults-was essentially a phenomenon controlled by elite social groups rather than a popular movement of the lower classcs. Another interesting and previously unnoticed reference to the Kāpālikas is found in the well-known Nāth-sampradāya text, Svātmārāma's

6. Perürppuranam 30. 1-33. David Dean Shulman, The King and the Clown in South Indian Myth and Poetry (Princeton: Princeton University Press, 1985). 7. Translated by Shulman, The King ..... , pp. 305-06.

218 APPENDIX A

Hatha-yoga-pradipīkā.8 Svātmārāma claims to describe the Amarolī mudra "according to the khanda doctrine of the Kapālikas." The Amaroli mudrā is a variant of the better known Vajrolī mudrā, a sexual yogic exercise in which the adepts reabsorb the seminal fluid released during intercourse. A few other previously unnoticed references to the Kāpālikas in published Sanskrit sources have been identified and discussed by Minoru Hara and A.C. Barthakuria.9 Unfortunately, none of them adds much of significance to our understanding of Kāpālika beliefs and practices. Recently, Nagendra Nāth Upādhyāy of Banaras Hindu University has published a monograph in Hindi on Bauddha Kāpālik sādhanā aur sāhitya in which he discusses the beliefs and practices of what he claims to be a tradition of Buddhist Kapalikas.10 He has in fact produced a useful study of Buddhist tantricism, some early schools of which do seem to have been strongly influenced by the Kāpālikas. It also seems clear, however, that the original Kāpālikas were Saivites and not Buddhists. Apart from Krsnācāryapāda's well-known and probably metaphorical references to himself as a Kāpālin in his caryāpadas,11 references.which come from a tradition as much associated with the Näths as with the Buddhists in any case, Upādhyay has not offered, in my opinion, sufficient evidence to justify the identification of the Tantric Buddhists of his study as Käpālikas. On the other hand, Alexis Sanderson, who has been working with unpublished manuscripts of the Buddhist yoginitantra tradition, claims that this tradition should be considered to be "a variant of the Śaiva Kāpālika."12 Mark Dyczkowski has suggested to me that the Kāpālikas were simply followers of a Kāpālika "vow", recruited from various tantric groups, rather than members of a specific "sect." Although plausible, this view fails to explain the many references to the Kāpālikas as one of several

8. Svātmārāma, Hathayogapradipikā (Madras: The Adyar Library, 1972), verse 3.96. 9. M. Hara, Review of The Kāpālikas and Kālāmukhas, by D. Lorenzen, in Indo-Iranian Journal, 17 (1975), pp. 253-61; A.C. Barthakuria, The Kapalikas: A Critical Study of the Religion, Philosophy and Literature of a Tantric Sect (Calcutta: Sanskrit Pustak, 1984). 10. N.N. Upādhyāy, Bauddha Kāpālik sādhanā aur sāhitya (Allahabad: Smriti Prakāsan, 1983). 11. See above, pp. 69-71; and P. Kvaerne, An Anthology of Buddhist Tantric Songs (2nd ed .; Bankok: White Orchid Press, 1986), pp. 37, 113-22, 150-55. 12. Letter dated 16 January, 1987.

APPENDIX A 219

Śaivite sects and to their special doctrine of Somasiddhānta. Nonetheless, one must still try to better locate them in historical time and space by means of solid epigraphical evidence. In the first edition of this book I was able to identify only two inscriptions that clearly record donations to persons who are likely to have been Kāpālika ascetics (above, pp. 27-28). Both are from western India. A copper plate grant found in Igatpuri, Nasik District, registers a donation of a village to finance the "Guggula-pūjā" at a temple of the god Kapalesvara. The donees were the Mahavratin ascetics who lived at the temple. The donor was the early Cälukya king Nägavardhana who lived about the middle of the seventh century A.D. A second copper plate grant, found near Tilakwādā in Baroda District, registers a grant made in A.D. to "the muni named Dinakara, a Mahāvratadhara who was like the Kapālin, Samkara, in bodily form." The donor was a vassal or official of the Paramära king Bhoja. One other earlier identified reference to a contemporary Kāpālika (here called a Somasiddhāntin) ascetic is found in an A.D. 1171-72 inscription from the Tiruvorriyūr matha in present day Chingleput District of Tamilnadu, although it is not really clear whether the actual donee of record, one Caturānana-pandita, was a Kāpālika or not (above, p. 83). In recent years two other epigraphs have bcen edited and published which can be identified as grants from and to Kāpālika ascctics, although the editors of the inscriptions have not made this identification. The more important inscription was found at Kolanupaka, ancient Kollipāke, in Nalgonda District of Andhra Pradesh.13 It is written in highly Sanskritized Kannada prose and is dated in Saka year 973 (A.D. 1050) during the reign of the Cālukya (of Kalyāna) king Trailokyamalla (Someśvara I). It records the donation of some land to Candamayya, a servant of Gangamarāja, by a Brahman (vipra) named Somi-bhattāraka. This Somi-bhattäraka is said to be the chief (sthänadhipati) of the temple of Śamkareśvara (=Śiva) in Kollipāke. These inscriptions give us the first word portrait of a Kāpālika ascetic from a source sympathetic to these ascetics. After a few words whose joint sense is not completely clear, the inscription refers to Somi- bhattäraka as "devoted to the meaning of the Somasiddhanta issued from the lotus mouth (? of Siva)." Somi-bhattāraka is said to be "sprinkled with ashes; adorned with the six insignia (mudra); and holding a

13. P.V. Parabrahma Sastry, Select Epigraphs of Andhra Pradesh (Hyderabad: Government of Andhra Pradesh, 1965), pp. 7-10.

220 APPENDIX A

khatvānga club, a skull (kapāla), damaruga and mrdanga drums, and a kāhala ( ?= trumpet)." Next he is called a "bee (buzzing round) the lotus fcet of the glorious (god) Srinatha." Finally he is said. to be a "Mahāvratin, an ocean of gencrosity and a treasury of truth and asceticism."14 The most interesting aspects of this description of Somi-bhattāraka arc the mention of his devotion to the doctrine of Somasiddhanta and the mention of the six insignia, the khatvanga club and the skull. Somasiddhänta can be better discussed after revicwing the contents of the second new Kāpālika inscription. The mention of the six insignia, the skull and the khatvanga club provides an important and surprising confirmation of the statements about the Kāpālikas made by Yāmunācārya and his disciple Rāmānuja, both contemporarics of Somi-bhattāraka. In his Agamaprāmānya, Yāmunācārya claims that the Kāpālikas assert that:15 The fruit of liberation (apavarga) is attained through knowledge of the six insignia (mudrika-satka) and through wearing them, not through the understanding of brahman. He who knows the essence of the six insignia, who is proficient in the highest mudra, and who meditates of the Self as seated in the vulva (bhagasana-stha) attains nirvāna. (The Kāpālikas) define the six insignia as the karnika (earring), the rucaka (necklace), the kundala (carring), the sikhāmani (crest-jewel), ashes (bhasma), and the sacred thread (yajñopavita). The skull (kapäla) and the khatvanga club are declared to be the sccondary insignia (upamudra). If the body is marked with these (various insignia) one is not born again here (on carth). In Somi-bhattaraka's inscription, the reading of the term "six insignia" (san-mudrā) is not completely clear, but a comparison with Yāmu- näcarya's text confirms the reading, particularly since the skull and khatvanga appear in both texts immediately afterwards. The damaru and other drums and musical instruments are frequently associated with Śaivite ascetics, but they do not seem to have been special insignia of

14. mukha-kamala-vinirggata-Somasiddhāntā bhiprāya-parāyanam [*bha] smoddhūlita-sa [*ņ-mudrā] [ *= drā] lamkrita-khatvamga-kapāļa-dāmaruga- mrdamga-kā halam sri-Srīnātha-pāda-pamkaja-bhramara-mahāvrati-dānāmbu- rāsi-satya-tapo-nidhi -..... The starred additions and correction are my own as is

15. Yāmunācārya, Āgama-prāmānya, ed. J.A.B. van Buitenen (Madras: Ramanuja the translation.

Rescarch Socicty, 1971), p. 43, lines 7-16. The translation is mine. Rāmānuja repeats the same verses minus he first and last (see above, pp. 1-4).

APPENDIX A 221 the Kapalikas. Ashes are of course worn by ncarly all Saivite ascetics. Detailed discussions of the six insignia, the two secondary insignia and the term "Mahavratin" are found in the carlier chapters of this book and necd not be repcated here (above, pp. 2-7, 73-82). I would only note here that I do not agree with Minoru Hara's suggestion that "the adjective mahat in the maha-vrata may have had a pcjorative connotation" as in the case of the term mahabrahmana.16 It is clear from this inscription and from many other sources that the Kapalikas themselves called themselves Mahavratins. It is unlikely that they would have done so if the term was pejorative. In the second new inscription, it is the donce who appears to belong to the Käpālikas, although the identification is somewhat more problematic in this case. The inscription was found in Bangalorc District.17 It is written in Sanskrit and Kannada and has been datcd in about the sixth century A.D. The donor was the "Prthivi-Komkani- mahādhirāja" named Durvinīta. In his fourth regnal ycar Durvinīta gavc the village of Penna-ūr as a brahmadeya to "Kāpāli-śarman, who belonged to the Kutsa gotra, the Taittiriya-carana, and the Pravacana- kalpa; who was a resident of Upakotta; whose rites were characterized by sacrifices with Soma; and who was the son of Agni-sarman."'18 The name Kāpāli-sarman makes it fairly likely, though by no means certain, that he was a Kapalika. We have noted that Somi-bhattāraka was called a vipra, i.e. a Brahman. Kāpāli-śarman is not merely a Brahman, he is a learned Brahman evidently well-versed in Vedic literature and rituals. His association with Vedic Soma sacrifices is particularly noteworthy since it suggests some possible relation between such sacri- fices and the Somasiddhānta of Somi-bhattāraka and other Kāpālika

16. See Hara, n. 257n. 17. K.V. Ramesh, Inscriptions of the Western Gangas (Delhi: ICHR and Agam Prakashan, 1984), pp. 70-74 (no. 19). 18. Kutsa-gotrāya Taittirīya-caranāya Prāvacana-kalpāya Upakotta-nivsine Somenesta-visista-karmmanah Agni-sarmmaņaḥ putrāya Kāpāli-śarmmaņe ...... My translation differs slightly from that of Ramesh who renders the phrase Somenesta-viśista-karmmanah as "whose religious rites were rendered special by his performance of those rites with soma." Ramesh notes, however, that an earlier edition of the inscription in the Mysore Archaeological Reports reads somenesti- rather than somenesta -. Even granting that Remesh's reading is orthographically correct, it makes better sense if the text is amended to the reading in the earlier edition.

222 APPENDIX A

ascetics. Although Somasiddhänta is discussed at some length in chapter III (above, pp. 82-83), I would like to briefly review the basic argument. In the lists of four (or more) Saivite sects frequently found in Sanskrit sources, the term Soma, Sauma or Saumya sometimes appears in place of Kāpālika. Ksīrasvāmin's commentary on the Amarakosa equates Kāpālins, Mahavratins, Somasiddhāntins, and Tāntrikas. In Krsnamiśra's Prabodhacandrodaya and in Anandarāya's Vidyāparinayana, the Kāpālika characters are named Somasiddhānta. Several commentaries on Krsnamiśra's Prabodhacandrodaya derive the term soma from sa-Umā or Umaya sahitah, i.e. "he who is with Umā," or the god Siva as the lover of Uma-Parvatī. This didactic etymology obviously agrees well with the sexual symbolism characteristic of tantric Hinduism, but Kāpāli- sarman's association with Soma sacrifices or rites suggests the possibility of a more orthodox, Vedic background to the term. Vinodacandra Śrīvāstav has criticized me for identifying Somasiddhänta as the doctrine of the Kāpālikas alone, suggesting that Somasiddhānta and Nyāyasiddhänta may have been equivalent terms.19 This equation would make it possible to connect the philosophical affiliation of the Käpalikas with that of the Kalāmukhas, wno are known to have been mostly Naiyāyikas, but Šrīvāstav supports his hypothesis with insufficient evidence. In all the so far identified occurrences, the term Somasiddhānta refers to the doctrine of the Kāpālikas or, in one or two references, other tantric ascetics. Until further evidence is discovered, there is little reason to connect Somasiddhänta with Nyāya or any other known philosophical school. The image of the Kapalika ascetic in Indian literature is an interesting and valid topic of study independent of its possible correspondence or non-correspondence with historical reality. Nonetheless, the attempt to reconstruct this historical reality is a task of equal or greater value. A. Sanderson's proposed identification of a corpus of Kāpālika texts

19. V. Śrivāstav, "Kāpālik evam Kālāmukh: ek aitihāsik samīksā," Journal of the Bihar Puravid Parishad, 3 (1979), pp. 155-78. Srivästav fails to cite the most interesting piece of evidence in favor of his hypothesis. G. Tucci has suggested that the term na ya xiu mo found in Chinese translations of Harivarman's lost Sanskrit work Tattvasiddhisāstra is equivalent to nyāya-sauma or nyāya-saumya (see above, p. 82). Śrivästav further speculates (p. 174) that Soma may be the name of the human author of the Somasiddhanta. This is a plausible suggestion but is not supported by any concrete evidence.

APPENDIX A 223

preserved in Nepal has the potential to revolutionize our understanding of actual Kāpālika beliefs and practices. Even so, a better understanding of the historical evolution, distribution and strength of the Kapālikas also requires the identification and analysis of whatever epigraphical material exists relevant to these ascetics. The inscriptions of Somi-bhattāraka and Kāpāli-śarman add important historical data to what is already known about the Kāpālikas.

APPENDIX B

THE KĀLĀMUKHA BACKGROUND TO VĪRAŚAIVISM

The first edition of this book was published in 1972 but in fact mostly written in 1967.1 Since then, a large amount of previously unavailable source material relating to the Kālämukhas, chiefly the texts of inscriptions, has been published. In this paper I want to review the more important new material and make a tentative reevaluation of the still largely unsolved historical problem of the transition from Kālāmukha and Virasaiva dominance in the Karnataka region starting from the appearance of Basava and his Virasaiva contemporaries about the middle of the twelfth century. The Kalamukhas formed the principal branch of the Pāsupata "sect" of Saivism present in the Karnataka region from about the ninth to the thirteenth centuries. Particularly under the royal dynasty of the Western Cālukyas of Kalyana, and to a somewhat lesser extent under the successor dynasties of the Kalacuris and the Seunas (Sevunas) or Yadavas of Devagiri, the Kalamukhas were the dominant religious force in this region. Many Kālāmukha preceptors claimed the title of rājaguru to the kings of these dynasties and their feudatories. The rapid and rather mysterious decline of the Kälämukhas during the thirteenth century probably had as much to do with the decline of these dynasties as with the activities of the Virasaivas and other rival religious movements. Unfortunately, the available historical record is not sufficient to raise the discussion on this point much beyond the level of deductive speculations. The numerous inscriptions which refer to Kālāmukha temples, monasteries and preceptors often reveal much about their religious beliefs and accomplishments. These inscriptions paint a quite different picture of the Kälamukhas than that contained in the attacks on them made by their theological and religious opponents such as Yāmunācārya and Rāmānuja. These attacks, together with somewhat peculiar religious practices prescribed in Pāsupata texts, have suggested to many modern scholars that the Kālamukhas were a heterodox, and non-Brahmanical,

1. David N. Lorenzen, The Kāpālikas and Kālāmukhas: Two Lost Saivite Sects (Berkeley: University of California Press; and New Delhi: Thomson Press, 1972).

APPENDIX B 225

sect replete with tantric influences like the Kāpālikas. In a recent public lecture, for instance, a leading historian of South India referred to the Kālāmukha and Pāśupata preceptors as "so-called Brahmans" and as priests who "claimed Brahman status." In fact, except in one or two cases, the inscriptions do not identify the caste of the Kālāmukha preceptors, but most inscriptions do claim that these preceptors were panditas well-versed in traditional Sanskrit learning. In what sense, then, can we consider the Kālāmukhas to be heterodox? It must be admitted that there is a sense in which any "sect" (sampradāya or pantha) within Hindu tradition may be regarded as somewhat unorthodox since the sect's very existence is defined by its maintaining certain spccific beliefs and practices which distinguish it from the varied, but very looscly specified or defined beliefs and practices of the majority of Hindus who do not profess allegiance to any particular sect. This is true even of the otherwise highly orthodox and "non-sectarian" organization of the smärta Brahmans, who generally accept a specific theological doctrine, usually Advaita Vedanta, and practise a specific form of worship, the pāñcayatana-pūja. The smārta cult is distinct from most other sects, however, insofar as it is basically domestic rather than centered in temples and monasteries. Apart from this weak sense of heterodoxy inherent in any scct, however, I can see no reason not to accept descriptions of Kālāmukha preceptors as orthodox pandits who fully merited their often used epithets pandita and pandita-deva, whatever their caste origins may have been. The elaborate praises of their religious and scholarly accomplishments in the inscriptions indicate that most of the leading Kālāmukha preceptors were followers of the orthodox theological schools of Nyāya and Vaiśesika and that they were also experienced adepts in the mental and physical disciplines of Yoga. Many are said to be learned in the theological doctrines of their opponents as well, striking evidence both of their religious tolerance and of their skill in religious debate. Often the Kālāmukha preceptors are described as "life-long brahmacārīs," suggesting that that sect was comprised principally of celebate sannyāsins and their lay followers. There is evidence, however, that at least some of the preceptors were married. The inscriptions show that the Kālāmukhas were divided into at least two different parisads ("councils"): the Sakti-parisad and the Simha- parisad. The Sakti-parisad was further divided into at least two separate

226 APPENDIX B

āvalis or āmnāyas ("traditions"): the Parvatāvali and the Bhujangāvali, both of whose monastic centers were mostly located in northwestern Karnataka and adjoining districts. These ävalis were further divided into a number of santatis or santānas ("lineages"). The Ittige-santati at Mattikote, Shimoga District belonged to the Bhujangavali while the Müvara-Köneya-santati at Belagāve, Shimoga District, and the Belleya- santana at Puligeri and Hombal, Dharwar District, both belonged to the Parvatāvali. The more recently published inscriptions reveal the names of two more santatis of the Parvatāvali of the Sakti-parisad: the Śiva- lingi-santāna with a center at Hippalgãon (Pippalagrāma) in Osmanabad District, Maharashtra, and the Kittagäve-santati at Tilivalli and other sites mostly in Dharwar District. The chief centers of the Simha-parisad seem to have been at Vijayawada (Bezwada), about 100 kilometers from the mouth of the Krishna River in Andhra Pradesh, and at Kogali in Bellary District of Karnataka. A large number of inscriptions refer to Kālāmukha preceptors without mentioning the organization divisions to which they belonged. Probably most were members of one or other of these two parisads. A number of the newly published Kālāmukha inscriptions come from districts where their presence was previously unknown. Such is the case of inscriptions from several temples in Nanded District of southern Maharashtra edited by S. Ritti and G.C. Shelke.2 A series of inscriptions relating to a Someśvara or Somanātha temple at Karadkhed in Deglur Taluk, mostly dated in the reigns of the Western Cālukya kings Vikramāditya VI (1076-1126 A.D.) and Someśvara III (1126-1138 A.D.), mention a number of preceptors whose names indicate that they were probably Kālāmukhas: Śrī-Candraśekhara-pandita-deva (1079, 1102 A.D.), the same preceptor plus Iśānaśakti and Jñāneśa-yogīśvara (an undated twelfth century inscription), and the latter preceptor as Jñāneśvara-pandita (1130 A.D.).3 An inscription from Tadkhel in Deglur Taluk of the same district, dated in 1047 A.D. during the reign of Someśvara of the Western Cālukyas, mentions a donee named Vägeśvara-pandita who is described as the chief of the Kalamukhas (Kālāmukha-mukhyarum), a life-long

2. S. Ritti and G.C. Shelke (eds.), Inscriptions from Nanded District (Nanded: Yashwant Mahavidyalaya, 1968). Throughout the present paper Sanskrit-based names have been transliterated in their corrected Sanskrit forms except where inappropriate. 3. Ibid., nos. 10, 13, 24, 28.

APPENDIX B 227

brahmacārī (paramanaisthikarum), and ascetic (tapodhanara)." A long Sanskrit inscription from Hottul in the same taluk, datcd sometime in the reign of Somesvara II (1068-1076 A.D.) of the Western Cälukyas, mentions the construction of a temple of Siddheśvara at the capital city of Kalyāna, made before the guru Candrarāsi.5 The names of the temple and preceptor suggest that the latter was a Kālāmukha. Two inscriptions from Sagroli in Biloli Taluk of the same Nanded District, dated in the reign of Vikramāditya VI, mention a priest named Brahmarāśi-pandita (1110 A.D.) or Bammarāsi-pandita (1115 A.D.) who is described as the head of the temple of Sarveśvara in Savaravalli.6 The names again suggest that he was a Kālāmukha. The most interesting of the Kālāmukha inscriptions found at sitcs in Nanded District comes from Ardhäpur in Nanded Taluk. It is written in Sanskrit, as opposed to the Kannada of most Kālāmukha inscriptions, and is dated in a year equivalent to 1192 A.D. during the reign of Mallugi II of the Seunas and the Ratta chief Ballala.7 After a long description of the family of Ballala, it registers some grants of land and money for offerings to the temples of Mallikārjuna and Sarveśvara. The donce was Kālāmukha preceptor named Brahmarāśi, the disciple of Kalyānaśakti, who was the disciple of Kāleśvara. Brahmarāśi's "son" Bhogarāśi is also mentioned. Brahmarasi is described as "possessing the yogic virtues of restraints, disciplines, internal scriptural repetition, meditation, concentration, constant silence, murmured prayer, and yogic ecstasy; skilled in the whole collection of agamas and arts; a preceptor grounded in the initiation of the Kālamukhas; devoted to the worship of the lotus feet of the god Sri-Tripurantaka; the chief of the seventy-seven monasteries; and faithful to his vow."'8 Perhaps the most interesting of these epithets is that of "chief of the seventy-seven monasteries." This is virtually identical to the epithet "preceptor of the seventy-seven places" used to describe two Kālāmukha

4. Ibid., no. 3. 5. Ibid., no. 33. 6. Ibid., nos. 14, 18. 7. Ibid., no. 40. 8. yama-niyama-svā dhya ya-dhyāna-dhārana-monā susthā na-japa-samá dhi- sila-guna-sampan(n)a(h) sakalāgama-kalā-kalāpe kusala(h) kālāmukha- diksā-prutist(h)-åcårya(h) sri-Tripuråntaka-deva-pådāravimdavamdanarata(h) saptādhika-saptati-mathānām adhipo vrati.

228 APPENDIX B

preceptors from Tilivalli in Dharwar District. It seems likely that all these preceptors belong to the same branch of the Kālāmukhas: the Kittagāve- santati of the Parvatävali of the Sakti-parisad. The Tilivalli inscription and other records relating to this newly identified branch of the Kālāmukhas will be discussed in more detail below. A Sanskrit inscription from the village of Hippalgãon (Pippalagrāma) in Osmanabad District of southern Maharashtra, issued in 1099 A.D. during the reign of the Western Cālukya Vikramāditya VI, reveals the existence of still another santati of the Parvatāvali and Śakti-parisad.9 This is called the Siva-lingi-santana. The inscription registers a donation to the Sarasvati-mandapa of this place by Bhima-nätha, the chief minister of the king. The gift was entrusted to the Kālāmukha leader Tatpurusa- pandita, the "younger brother" of Trilocanäcärya and the disciple of Kriyāśakti-pandita-deva. Kriyāśakti was the disciple of Vādi-Rudra- pandita. The virtues of all four preceptors, and of Tatpurusa's own disciple, Srikantha, are recounted in some detail. Several are given the usual list of yogic accomplishments (yama-niyama etc.) and are said to be masters of the orthodox darsana-s of Nyāya, Vaiśesika, Sāmkhya, and Mīmāmsā in addition to being followers of the Lākulāgama or Lakulīśvarāgama. Although these several different spiritual lineages of Kālāmukha preceptors from Nanded and Osmanabad Districts were unknown previously, their presence in this part of southwestern Maharashtra during the eleventh and twelfth centuries is not surprising. The dynasties which controlled this region during this period were the Western Cālukyas followed by the Seunas. Both dynasties, especially the Western Cālukyas, were generous patrons of the Kälämukhas. It is worth noting that the zone of greatest Kälämukha strength in Karnataka-approximately Shimoga, Dharwar and Belgaum Districts-was an area of nearly constant conflict between the Seunas and the Hoysalas, who were only rarely patrons of the Kälämukhas, during the whole of the thirteenth century. By the end of this century Ala-ud-din Khilji had begun to bring the territory of the Seunas under his control, which can only have made the situation worse for the Kālāmukha temples and monasteries.10

9. V.B. Kotte, "Ganeshvati Inscription of the Time of Chalukya Tribhuvanamalla, Year 24, "Epigraphia Indica, vol. 38 (1970 [1981]), pp. 289-304. 10. Perhaps the clearest discussion of this historical situation is that of A.V. Narasimha Murthy, The Sevunas of Devagiri (Mysore: Rao and Raghavan, 1971).

APPENDIX B 229

During the eleventh and early part of the twelfth century, parts of the present state of Andhra Pradesh also fell within the territory of the Western Calukyas. Not surprisingly we find that kings and feudatories of this dynasty made grants of Kālāmukha temples and monasteries located in this region. The important Mallikārjuna monastery at the famous pilgrimage center of Srisaila or Śrīparvata in Kurnool District was controlled by Kālāmukha preceptors of the Parvatāvali of the Sakti- parisad. Earlier, in the tenth century, a king of the Eastern Cālukyas made a grant to the temple of Samastabhuvanāśraya at Vijayavātī (Bezwada) on the Krishna River. The temple was at that time controlled by Kālāmukha preceptors of the Simha-parisad.11 Recently, a number of new Kālāmukha sites in Andhra Pradesh have come to light from inscriptions published by the Andhra Pradesh Archaeological Department. At Sanigaram in Karimnagar District two inscriptions have becn found which register grants to the temple of Kommēśvara and its preceptor Lokābharana-pandita in 1128 A.D. and to the temple of Partheśvaradeva and its preceptor Candräbharana-pandita in 1149 A.D. respectively.12 The first record refers to the Western Cālukya king Bhūlokamalladeva (Someśvara III), his feudatory the Kākatīya Prolalarasar, and his feudatory the chief Repola Kuruvarasar. The second record refers to the Western Cälukya king Jagadekamalladeva (III) and the same two feudatories. The names of the gods and preceptors suggest that they were Kālāmukhas. More evidence from Karimnagar District comes from Vemulavāda. An inscription of 1083 A.D., issued during the reign of the Western Cālukya king Tribhuvanamalladeva (Vikramāditya VI), describes the construction of a temple of Rājādityeśvara by a feudatory of the king and some gifts made to the temple for the upkeep of ascetics belonging to the Bhujangāvali.13 The Bhujangāvali, as noted above, was a major subdivision of the Sakti-parisad of the Kālāmukhas.14 Another inscription from the same town, issued in 1108 A.D. during the reign of the same king, registers a gift for the temple of the god Harikeśvaradeva entrusted to the preceptor Brahmeśvara-pandita of the "Simgha

11. See above-pp. 136-37, 141-43, 165. 12. Epigraphia Andhrica, vol: 4, ed. P.V. Parabrahma Sestry (Hyderabad: Government of Andhra Pradesh, 1975), pp. 67-89. 13. Inscriptions of Andhra Pradesh: Karimnagar District, ed. P.V. Parabrahma Sastry (Hyderabad: Government of Andhra Pradesh, 1974), no. 13 (also p. xv). 14. See above-pp. 98-99, 108, 138-40, 156.

230 APPENDIX B

Parise" (Simha-parisad).15 This shows that representatives of this second major division of the Kālamukhas also has an establishment at this site. A number of Kālāmukha inscriptions have been found in Warangal District of Andhra Pradesh.16 A recently published inscription from Hanumakonda (Hanan-Konda, Anamkonda), although fragmentary and not dated, records a gift entrusted to Sureśvara-pandita, the preceptor of the Mūleśvara temple of Śrīvarddhanadakota. He belonged to the Bhujangāvali (of the Sakti-parisad) and is described as "sakala- Lakulasiddhānta ... pradīpam."17 Lakulīśa is the well-known founder of the Pasupatas, hence their doctrine can be called Lākula-siddhānta. Another inscription from this district was found at Ayanavolu. It registers a grant to a preceptor named Lakulesvara-pandita, and is dated in 1118 A.D. during the reign of the generous patron of the Kālāmukhas, Vikramāditya VI of the Western Cālukyas.18 Several recently published inscriptions which seem to pertain to the Kālāmukhas have also been found at sites in Cuddapah District of Andhra Pradesh.19 Other such inscriptions from sites in this state have been noted or edited in several state government publications.20 Particularly interesting is a mural painting of the Pāśupata-Kālāmukha

15. Inscriptions of Andhra Pradesh: Karimnagar District, no. 21. A Simha-parisad inscription of 1122 A.D.from Onikeri, North Kanara District is noted in Archaeological Survey of India, Annual Report on Indian Epigraphy for 1972-73 (Delhi, 1983), stone No. 77. 16. See above, pp. 136-37. 17. Inscriptions of Andhra Pradesh: Warangal District, ed. N. Venkataramanayya (Hyderabad: Government of Andhra Pradesh, 1974), No. 113. 18. Ibid., no. 35. 19. Inscriptions of Andhra Pradesh: Cuddapah District, ed. P.V. Parabrahma Sastry (Hyderabad: Government of Andhra Pradesh, 1977-78), Part I, nos. 129, 162, 140, 142; Part II, no. 1. 20. See N. Venkataramanayya, Andhra Pradesh: Report on Epigraphy for 1966 (Hyderabad: Government of Andhra Pradesh, [1972]), nos. 130, 131, 132; P.V. Parabrahma Sastry (ed.), Departmeni of Archaeology and Museums, Andhra Pradesh: Annual Report on Epigraphy, 1967 (Hyderabad: Government of Andhra Pradesh, 1975), no. 406; Mallampalli Somasekhara Sarma, Corpus of Inscriptions in the Telangana Districts, Part 4 (Hyderabad: Government of Andhra Pradesh, 1973), nos. 3, 6, 7, 10, 11, 12, 13, 19, 20, 21; P. Sreenivasachar, Kannada Inscriptions of Andhra Pradesh (Hyderabad: Government of Andhra Pradesh, 1961), nos. 30. 50. See also Epigraphia Indica, 37 (1967), 41-44 and ibid., 36(1965), 69-74. See also Archaeological Survey of India, Annual Report on Indian Epigraphy for 1961-62 (Delhi, 1966), stone no. 94; ARIE for 1962-63 (1967), stone nos. 186, 199, 200.

APPENDIX B 231

founder saint Lakulisa found at Alladurg in Medak District.21 The region in which the Kalamukhas exercised their maximum influence was the northwestern part of the present state of Karnataka, approximately the territory comprising Shimoga, Dharwar and Belgaum Districts. A number of recently published inscriptions from these districts deserve some discussion. One interesting series of inscriptions refers to a line of preceptors associated with the Rameśvara temple at Rämatirtha in the northern Athani Taluk of Belgaum District. The earliest of these records was issued in 1115 A.D. during the reign of Vikramäditya VI.22 It was found in the compound of the Rämeśvara temple itself. It registers a grant to this god which was entrusted to Sivaśakti-pandita, described as a devotee of the god Nemeśvara, an upholder of the preceptor-lineage (guru-kula) of Gotakanura, and endowed with several of the yogic virtues often attributed to Kālāmukha preceptors. This Gotakanura scems to be the same as the site of Kokatnur also in Athani Taluk. An inscription from Kokatnur issucd in 1169 A.D. during the reign of Rāyamuräri Sovideva of the Kalacuris records a gift made to Vidyārāśi-pandita-deva, the precepior of. the temple of Svayambhu Visveśvara at the agrahāra Korgotanura (i.c. Kokatnur).23 A record found at Umaräni in the former Jath State of Sholapur District, Maharashtra, a site not far from Rāmatīrtha, registers a donation to a temple of Hemmeśvara in Umarāni, entrusted to Candräbharana-pandita-deva, the preceptor of the temple of Svayambhu Rāmeśvara and disciple of Lokābharana-pandita-deva. The inscription is dated in a Saka year probably corresponding to 1141 A.D.24 That this temple of Rāmeśvara is in fact the temple in Rămatīrtha is confirmed by a grant found there dated in a year equivalent to 1167 A.D. during the reign of the Kalacuri king Bijjala (c. 1145-1167 A.D.).25 At the instance of Bijjala, the mahāmandaleśvara Permādi made a grant for the renovation of the Svayambhu Rāmeśvara temple. The grant was entrusted to the same Candrabharana, disciple of Lokābharana.

21. Abdul Waheed Khan, A Monography on the Discovery of Mural Paintings of Kalyani Calukyas at Alladurg (Hyderabad: Govemment of Andhra Pradesh, 1968), p. 4 and plate I. 22. Karnatak Inscriptions, vol. 2, ed. R.S. Panchamukhi (Dharwar: Kannada Rescarch Institute, 1951), no. 13 of 1940-41. 23. Ibid., no. 25 of 1940-41. 24. Ibid., no. 14 of 1940-41. 25. Ibid., no. 24 of 1940-41.

232 APPENDIX B

Two other inscriptions from Belgaum District deserve mention. The first is from Gölinalli, Khanapur Taluk, and is a composite grant with dates equivalent to 1160, 1163 and 1173 A.D.26 A subordinate of the Kadamba chief Sivacitta Vira Permādideva had a temple of Someśvara built in 1160 and entrusted it to Śrotriyadeva. In 1163 a grant was made to the same temple, entrusted to Srotriyadeva's disciple Saktiśivadeva. The second inscription is from Saundatti, Saundatti Taluk, and bears dates equivalent to 1048 and 1087 A.D.27 In the former year, during the reign of the Western Cālukya king Someśvara I, a Ratta chief named Ankarasa built a temple of Ankesvara in Saundatti and made some gifts for its upkeep. These gifts were entrusted to Someśvara-pandita of the puliyarasa monastery. In 1087 another Ratta chief renewed the gifts and entrusted them to Someśvara-pandita's disciple Jñānaśakti. This Puliyarasa monastery was located in the town of Hūli, ancient Püli, also in Belgaum District. Several earlier published inscriptions relating to this line of preceptors have been found in Hüli itself.28 Dharwar District is particularly rich in Kālāmukha sites. Many new inscriptions relating to previously known and unknown sites in the district have been published, or briefly noted, over the past twenty years. A short inscription found at Anaji in Hirekerur Taluk records a gift of land to a temple connected with the "Sasti-parise."29 This must be the Sakti-parisad. The chief interest of the inscription is its early date. It mentions a local ruler of Banaväsi province who is known from other records to have been a subordinate of the Rastrakūta king Krishna II (c. 880-914 A.D.). This makes the inscription the earliest Sakti-parisad epigraph and one of the earliest of all Kālāmukha epigraphs. Only the 810 and 806 A.D. grants from Nandi Hill in Kolar District, Karnataka, are significantly older.30

26. Karnatak Inscriptions Series, vol. 5, ed. B.R. Gopal (Dharwar: Kannada Research Institute [1969]), no. 64. 27. Ibid., no. 52. 28. See above-pp. 150-53. 29. Karnatak Inscriptions Series, vol. 6, ed. B.R. Gopal (Dharwar: Kannada Research .Institute, 1973), no. 3. 30. See above-pp. 160-61, 141-42. But see also the newly published ninth century inscription from Basavanahalli, Mysore District, in K.V. Ramesh, Inscriptions of the Western Gangas (Delhi: Indian Council of Historical Research, 1984), no. 104. This record mentions a Lakulisvara-bhatāra, disciple of Śnkantha-pandita and sthānika of the Ereyange'svara temple, without calling them specifically either Pasupatas or Kālāmukhas. There is a Lakulisa sculpture in the Lokeśvara, now Virūpaksa, temple

APPENDIX B 233

Two inscriptions from the town of Hangal in the same Dharwar District, both dated in years equivalent to 1119 A.D. during the reign of Vikramāditya VI, mention probable Kālāmukha temples and preceptors. One registers a gift to the Mulasthāna of the Billeśvara temple at Pāntipura entrusted to a Devaräi-pandita.31 The second registers another gift, this one to the temple of Teligeśvara attached to the temple of Cikēśvara of Pāntipura, entrusted to Tribhuvanaśakti-pandita, the "brother" of Sakalēśvara-pandita.32 All these preceptors are said to possess the yogic virtues often listed in Kālāmukha inscriptions. At the village of Kummur in Byadgi Taluk of Dharwar District an inscription of 1195 A.D., issued during the reign of Someśvara IV of the Western Cālukyas, records a grant made to the temple of Rāmanātha there.33 The donee was a preceptor named Kedāra-pandita, the disciple of Mallikārjuna-deva, the disciple of Vāmaśakti-deva, the disciple of Vaktraśiva-panđita. Kedāra-pandita is given the usual yogic virtues. Another genealogy of probable Kālämukha preceptors is found in an undated inscription of the twelfth century from Kanivi Siddhagēri in Hirekerur Taluk.34 It records a gift to the temple of Jamadagni Rāmeśvaradeva entrusted to the sthānācārya Rāmaśakti-deva, who seems to be the disciple of Mānikarāśi-pandita, the disciple of Rāmaśakti. Two early thirteenth century inscriptions which seem to have originally come from a site then called Manigundage, but are now deposited in the Kannada Research Institute, Dharwar, amply describe the activities and virtues of a preceptor named Candrabhūsana who was in charge of the temples of Grämeśvara and Siddheśvara in Manigundage.35 Both inscriptions are dated with reference to the reign of the Kadamba chief Jayakesi III of Goa: one in the years 1203, 1211 and 1215 A.D., and the second in 1216 A.D. On the earliest date, the settis of Manigundage, at the request of Chandrabhūsana, built a Siva temple named after this Kadamba chief. A record of 1207 A.D. which refers

at Pattadakal in Bijapur District built in the time of the early Cālukya Vikramāditya II (733-45 A.D.). See Epigraphia Indica, 32 (1957), 179. 31. Karnatak Inscriptions Series, vol. 5, No. 18. 32. Ibid., no. 95. 33. Ibid., vol. 6, no. 33. 34. Ibid., no. 7. 35. Ibid., vol. 5, nos. 32, 67.

234 APPENDIX B

to the rule of this chief mentions a gift made to Sakaleśvara-pandita, a disciple of Someśvara-pandita, who was associated with a Siddheśvara- matha. The find spot of the inscription is unknown.36 Perhaps the most important of the Kālāmukha inscriptions to be published during the last twenty years comes from the site of Tilivalli in the Hangal Taluk of Dharwar District.37 It is dated in the Saka year 1160, corresponding to 1239 A.D., during the reign of the Seuna of Yädava king Singhana II. It records, in mixed Sanskrit and Kannada, the construction of a temple dedicated to Siva as Sāvanteśvara by two sons of an officer of the king and the donation of a village to this temple. The village was entrusted to the Kālāmukha preceptor Rudraśakti, whose spiritual genealogy is described in some detail. Rudraśakti was the disciple of Saktideva, the disciple of Sarveśvara-deva, the disciple of Kumāra, the disciple of Gokarana, the disciple of Vidyābharana-deva. The praise of the first preceptor, Vidyabharana, begins with the usual list of yogic virtues found in Kālāmukha inscriptions. He is further called the preceptor of the seventy-seven places (saptāhattaristhānācāryyarum), a master of the four Vedas and other traditional subjects of Sanskrit learning, the upholder, of the doctrines of the Lakulāgama, a worshipper of the god Sri-Siddhanatha or Siddheśvara of Kabbina-sirivura, a teacher of the Kittakröma monastery (Kittakröma-mathattadesika-padah), and a leader of the "Kālāmukha" (Kālāmukhasyāgranī). The fourth preceptor of the lineage, Sarveśvara, was associated with the god Kotīśvara. known from another source to be a temple located at Devasthāna-Hakkalu in or near Kuppatūr in Shimoga District. The fifth preceptor, Saktideva, is the guru of the donce Rudraśakti. His praiscs begin with the same list of yogic virtues as that given to Vidyābharana and continue with similar phrases describing Saktideva as the preceptor of the seventy-seven places, the master of the four Vedas and other traditional subjects of Sanskrit learning, and the upholder of the Lākulāgama. He is associated with the god Srī-Kotinātha, presumably that of Devasthäna-Hakkalu, and with the god Svayambhu of Kalla- Mulugunda. Finally Saktideva is given the curious epithet Ekköti- cakravartti. The editor of the inscription, B.R. Gopal, argues that the term

36. Ibid., no. 31. 37. Ibid., no. 42 (also p. xxxvi).

APPENDIX B 235

Ekkoti is another designation of the Kālāmukhas.38 A previously published inscription of 1231 A.D. mentions, without the full gencalogy, a preceptor who must be the same Rudrasakti and also his "younger brother" Sarveśvaraśakti-deva, who is called the preceptor of the seventy-seven places, as the head of the Kotisvara temple at Devasthäna-Hakkalu in or near Kuppatūr in the Shimoga District.39 This inscription names several other temples as being attached to the Kotśvara temple including Siddhanātha of Kabbina-Sirivūr, Svayambhu of Mulugunda, Rāmanātha of Kurivade, Grāmeśvara (?Brahmeśvara) of Abbalūr (Ablūr), the Mūlasthāna Vośavanteśvara (Sāvanteśvara) of Tiluvalli (Tilivalli) the Caitrapura of Devangiri, the Mūlasthāna of Hānungal (Hangal), and Rāmanātha of Kuppatūr. Some slight evidence suggests that this lineage might be connected with the Mūvara-köncya- santati of the Parvatävali of the Sakti-parisad, whose headquarters was at Belagave in Shimoga District. Much stronger evidence, however, connects the Tilivalli and Devasthana-Hakkalu preceptors with the Kālāmukhas of the newly identified Kittagāvc-santati of the Parvatāvali of the Sakti-parisad whose other main centers appear to have been at Haralahalli in Haveri Taluk of Dharwar District and at Rattihalli in Herekerur Taluk of the same District. The Mallikārjuna and Sarveśvara temples at Ardhäpur in Nanded District, Maharashtra, discusscd abovc, probably belonged to the same Kittagäve-santati since the preceptor Brahmarāsi, who apparently headed these temples in 1192 A.D., is also called "chief of the seventy-scven monasteries." Two inscriptions found at Haralahalli in Haveri Taluk in Dharwar District register gifts to the temple of Someśvara or Svayambhu Somanätha.40 The donce in both cascs was the preceptor Kalyānasakti- deva, who is described as being a "Kālamukha" of the Kittagāve- santati and the Parvatavali of the Sakti-parisad. One mentions the Kalacuri

38. The term ekkôti appears again in a twelfth century inscription from Kambadahalli in Nagamangala Taluk of Mandhya District. See the new Epigraphia Carnatica (Mysore: · University of Mysore, 1972), vol. 7, no. 31 (old edition, vol. 14, No. 131). An inscription of 1150 A.D. from Belvantara, Dharwar District, which mentions an Ekköticakravarti preceptor of the god Svayambhudēva of Mulungunda is noted in Archacological Survey of India, ARIE for 1965-66 (Delhi, 1970), stone no. 385. 39. Epigraphia Carnatica, ed. B.L. Rice (Bangalore: Mysore Govemment Press, 1886-1904), vol. 8, Sb 275. The inscription is discussed above-pp. 100-01, 129-30. 40. Karnatak Inscriptions Series, vol. 6, nos. 39 and 74 respectively.

236 APPENDIX B

king Āhavamalla (1161-1183 A.D.) and bears dates equivalent to 1181 and 1188 A.D. The second record mentions the Gutta (Gupta) chief Vira Vikramäditya-deva, without naming any overlord, and bears -a date equivalent to 1188 A.D. The religious virtues of a Kalyānaśakti and his spiritual predecessors are described at considerable length in both inscriptions. Kalyānaśakti was the disciple of Sarveśa, the disciple of Amrtarāśi. A later record of Vīra Vikramāditya-deva, dated in 1213 A.D., registers another gift entrusted to the sthānācārya of the Somanātha temple named Trilocana-deva.41 Two newly published inscriptions form Rattihalli in Hirekerur Taluk of Dharwar District also register gifts to a long spiritual lineage of Kālāmukha preceptors of this same Kittagave-santati. The first of these inscriptions is dated in a year equivalent to 1174 A.D. during the reign of the Kalacuri king Rāyamurāri Sovideva and his Kadamba subordinate Pāndyadēvarasa.42 It registers a gift to the temple of Kadambesvara of Rattapalli, entrusted to Kalyānaśakti-pandita of a gurukula of the "Kālamukhas." He must have been quite aged at the time since both his disciple and disciple's disciple are mentioned in the record. The full spiritual genealogy given includes ten generations of preceptors who must have extended over more than two-hundred years. The genealogy begins with Vämadeva-yati, the guru of Vādi-Rudra-yati, the guru of Kriyāśakti (I), the guru of Amrtarāśi-yatīsa, the guru of Śrīkantha-deva, the guru of Kriyāśakti (II), the guru of Sarveśvarāya, the guru of the donee Kalyānaśakti-vratipati, the guru of Murujāvidēva-vratipati (I), the guru of Lakulīśvara-deva. The second grant is dated in a year equivalent to 1238 A.D. during the reign of the Seuna or Yädava king Singhana II and his Kadamba subordinates.43 The genealogy of the Kālāmukha preceptors is mostly copied from the earlier grant. The name of Lakulīśvara's disciple, Murajāvi (II), who is the donee, is. added, and the names of Kriyāśakti (I) and Amrtarāśi are inadvertently omitted. Vādi-Rudra-yati is called Vadi-Rudrasakti. The grant also adds the important information not found in the earlier record, that these Kālāmukha preceptors belonged to the Parvatāvali and Kittagāve-(santati). In addition, the donee

41. Ibid., no. 75. 42. Ibid., no. 37. 43. Ibid., no. 44. See also B.R. Gopal and S. Ritti, Summaries of Inscriptions (1943-44 to 1949-50). (Dharwar: Kannada Research Institute, 1966), nos. 25 and 26 of 1948-49.

APPENDIX B 237

Murujāvi (II) is given the title rājaguru, showing that he had important political patronage and influcnce either with the Kadamba feudatory or with the Seuna king himself. This title was in fact claimed by several Kālāmukha preceptors during the reigns of various Western Cālukya. Kalacuri and Seuna kings.44 The preceptor Murujävi II of Rattihalli is given the same epithet, "preceptor of the seventy-seven places" (saptāhattari-sthānācārya), that we met in the epigraphical elogies of the Kālāmukha lineages of the Sāvanteśvara temple at Tilivalli in Dharwar District, the Kotīśvara temple at Devasthāna-Hakkalu (Kuppatūr) in Shimoga District (together with several associated temples in this region), and the temples of Mallikärjuna and Sarveśvara at Ardhāpur in Nanded District. The preceptor Vidyäbharana of Tilivalli is given the additional epithet of "teacher of the Kittakroma monastery" which again makes his connection with the Kittagave-santati highly probable. This Kittagāve-santati of the Parvatāvali and Sakti-parisad was previously unknown to modern scholars. Taken together, the evidence from Rattihalli, Tilivaļli, Devasthāna-Hakkalu, Ardhāpur, and from Haralahalli in Dharwar District shows that the Kittagave-santati was a major institutional subdivision of the Kālāmukhas, perhaps rivaled only by the Mūvaraköneya-santati centered at Belagāve in Shimoga District.45 A collection of the available information about the known organizational divisions of the Kālāmukhas and their principal centers yields the following diagram on the next page. Academic research on the Kālāmukhas is still in its initial stages, and a great deal more work needs to be done, particularly in two areas. First is the problem of the sources of financial and social support for these

44. See above-pp. 122-25, 163. 45. Other probable Kälāmukha related inscriptions of somewhat lesser interest which are not noted either in my The Kāpālikas and Kālāmukhas or elsewhere in this appendix include the following: Gopal and Ritti, Summaries ... , nos. 7 and 25 of 1943-44; nos. 8, 11, 13 and 17 of 1948-49, nos. 3, 1, 4 and 7 of 1949-50. Some of these inscriptions have been published in Karnatak Inscriptions, vol. 4, ed. A.M. Annigeri (Dharwar: Kannada Research Institute, 1961), nos. 40, 39, 22 and 3. See also the new Epigraphia Carnatica, vol. 3, nos. 239, 241, 294; vol. 4, nos. 114, 145; vol. 5, no. 320. See also Archaeological Survey of India, ARIE for 1958-59 (Delhi, 1963, stone no. 59 (Nidanegila); ARIE for 1960-61 (1964), stone nos. 454, 468; ARIE for 1962-63 (1967), stone nos. 796, 797; ARIE for 1965-66 (1970), stone no. 398; ARIE for 1967-68 (1979), stone nos. 330, 331, 333; ARIE for 1970-71 (1979), stone no. 137; ARIE for 1972-73 (1983), stone nos. 64, 65, 80.

238 APPENDIX B

Kālāmukhas

Śakti-parisad Simha-parisad (Andhra Pradesh, Karnataka)

Parvatāvali Bhujangāvali

Ittige-santati (Mattikote, Shimoga District)

Mūvara-koneya-santati Belleya-santāna (Belagāve, Shimoga (Puligere, Dharwar District) District) Kittāgāve-santati Sivalingi-santati (Tilivalli, etc., Dharwar (Pippaligrāma, District) Osmanabad District)

preceptors and their monasteries and temples. The inscriptions contain considerable information relevant to this subjcct, but R.N. Nandi's pioncering studies scem to be the only serious attempt so far at analyzing the available data.46 As it has evolved, Nandi's position stresses the importance of large land-holding and industrial temple-estates as the foundation of the Kālāmukha movement and the subsequent rejection of this temple-estate system by the Viraśaivas, who were comprised of and led by Smārta Brahmans, farmers and artisans. These groups, he argues, had interests incompatible with those of the large temples. Nandi's thesis is an attractive one, but he fails to weight the influence of contingent historical events such as the fall of the Western Cālukyas of Kalyāna and the later takeover of the region by a Muslim dynasty. There is also a need for a more exhaustive and detailed examination of the empirical evidence.

46. R.N. Nandi, Religious Institutions and Cults of the Deccan (Delhi: Motilal, 1973), pp. 76-90; "Origin and Nature of Saivite Monasticism: The Case of the Kalamukhas" in R.S. Sharma and V. Jha, Indian Society: Historical Probings (New Delhi: People's Publishing House, 1974), pp. 190-201; "Origin of the Virasaiva Movement," The Indian Historical Review, 2 (1975), 32-46; and his "Introduction" to the new edition of S.C. Nandimath, A Handbook of Virasaivism (Delhi: Motilal, 1979).

APPENDIX B 239

Second is the problem of determining the process of takeover of the Kālāmukha temples by the Virasaivas and the nature and extent of the historical connection between the Kalāmukhas and the Virasaivas. I would like to discuss this latter problem a little further. By the middle of the thirteenth century, inscriptions referring to Kālāmukha preceptors suddenly, and somewhat mysteriously, become very scarce. The latest record which clearly mentions Kālāmukha preceptors is an inscription from Devasthāna-Hakkalu (Kuppatūr) in Shimoga District dated in a year equivalent to 1280 A.D. The next oldest scems to be a record of 1252 A.D. from Munavalli in Belgaum District.47 Both pertain to the reigns of Seuna kings. There follows over one-hundred years of virtual epigraphical silence from all the Kālāmukha sites. Then inscriptions reappear at many of these sites starting from about the end of the fourteenth century. By this time, however, most of the temples and monasteries are in the hands of Viraśaiva preccptors. The Kālāmukhas, who were the dominant religious force in northern Karnataka and many surrounding areas up to at least 1200 A.D., have simply

47. Ed. and trans: B.L. Rice, Epigraphia Carnatica, old vol. 8, Sb 275, and ed. and trans. J.F. Fleet, Journal of the Bombay Branch of the Royal Asiatic Society, 12 (1876), 1-50, respectively. The second inscription was re-edited in Karnatak Inscriptions, vol. 1, ed. R.S. Panchamukhi (Dharwar: Kannada Research Office, 1941). See abovc-pp. 100, 123. See also the epigraph of 1285 A.D. from Tumkur District mentioning a supporter of the Lākula system in B.L. Rice, Epigraphia Carnatica, old vol. 12, Tiptur 12. Sec also an epigraph of 1252 A.D. from Tumkur District mentioning a Šivayōgi Cikka Aghamme-deva and his spiritual son Cikka-deva in K.A. Nilakanta Sastri, Epigraphia Carnatica, old vol. 16 (1958), Gb 76. K.A. Nilakanta Sastri (Ibid., p. 19) thinks that the donees may be Lingayats, but it is more likely that they are Kālāmukhas. Similarly, see an epigraph of 1292 A.D. from Hassan District giving a long account of the life and miracles of a priest named Parvataiya, related to the god Mallikārjuna of Śrīparvata, in M.H. Krishna, Epigraphia Carnatica, old vol. 15 (1943), HN 112. An inscription of 1368 A.D. from the Madhukēśvara temple in Banavasi, North Kanara District, registers a grant to Bikkidēva, son of Lakulēśvaradēva. Another from the same site and same date registers a grant to Cikkadēva, son of Lākulisvara, who are likely the same two persons (see Vasundhara Filliozat, L'Epigraphie de Vijayanagar du debut a 1377 [Paris: École francaise d'extreme-orient, 1973], nos. 88, 90). These priests are said to be rāyarājagurus and to have "the lustre of the autumn moon which swells the ocean of the Siva doctrine (siva-samaya)." It is more likely that they were Pāsupatas than Kālāmukhas. The Pasupatas were still active at Tiruvānaikka, Trichinopoly District, in 1584 A.D. (Sce T.V. Mahalingam, "A Family of Pāsupata Grhasthas at Jambukēśvaram," Journal of Oriental Research, Madras, 25 [1957], 79-85).

240 APPENDIX B

disappeared.48 To take the example of the Savantesvara temple at Tilivalli in Dharwar District, we find a fifteenth or sixteenth century inscription at this site which records the installation of an image (?linga) of the Viraśaiva Basaveśvara.49 Similarly, the famous site of the Mallikārjuna temple at Śriśailam in Andhra Pradesh becomes the site of the Viraśaiva monastery called Bhiksāvrtti, first mentioned in a grant of 1385 A.D. and described at length in a 1506 A.D. Srīśailam copper plate inscription.50 Cynthia Talbot has shown that this Bhiksavrtti Matha also had some historical connection with the earlier Golaki Matha at Śrīsailam. The Golaki Matha priests seem to have belonged to the Saiva Siddhānta sect.51 How can the disappearance of the Kālāmukhas be accounted for? Political, social and religious factors must all have been involved. Without a great deal of further research in both epigraphical and literary

48. A somewhat similar situation occurs in the case of the well-known former Pasupata temple of Eklingji near Udaipur in Rajasthan. An inscription of 971 A.D. found there relates the history of the Pasupata founder-saint Lakulisa (ed. D.R. Bhankarkar, JBBRAS, 22 [1904-07], 151-65). The Pasupatas continued to exist in this region until at least the end of the thirteenth century. When the traditional Sanskrit history of the Eklingji temple was written, however, the Pasupatas were not even mentioned. See Ekalinga-māhātmyam, ed. Premlatā Śarmā (Delhi: Motilal, 1976). 49. Gopal and Ritti, Summaries ... , no. 53 of 1945-46. 50. See Cynthia Talbot, "Gölaki Matha Inscriptions from Andhra: A Study of a Saiva Monastic Lineage," in A.M. Shastri and R.K. Sharma (eds.), Bajapeya (KD. Bajpai Felicitation Volume) (Delhi: Agam Kala Prakashan, 1987), pp. 133-46. The two inscriptions are edited in South Indian Inscriptions, vol. 16, 7 and in Epigraphia Andharica, vol. 1, ed. N. Venkataramanayya (Hyderabad: Andhra Pradesh, Director of Archaeology and Museums, 1969), pp. 44-51, respectively. Other examples of this transition are cited above-pp. 167-72. A work which might shed more light on the transition is the Mallikārjuna-panditārādhyacaritramu by Pālkuri Somanātha, Virasaiva text of about the thirteenth century written in Telugu. See M. Rama Rao, The Temples of Srisailam (Hyderabad: Govt. of Andhra Pradesh, 1967). Talbot (p. 137, also note 8) suggests that the preceptor Parama-siva of the Kalmatha-who was the donee of a 1293 A.D. grant (APRE 244 of 1966) from Lingala, Kurnool District, to the god Mallikārjuna of Srisailam-may have been a Kālāmukha. The late date makes this unlikely, although the identification of the Kalmatha with the Kallu and Śilā mathas may well be correct. Talbot (pp. 140-41) also tentatively identifies the Asamkhyāta Māhēsvaras found at Śrisailam in 1313 A.D. (SII, 10, 504) and 1315 A.D. (ARIE 41 of 1915) with the Virasaivas. These seem to be the earliest records (also SII, 10, 503 of 1312) found at Śrisailam proper, suggesting that earlier inscriptions may have been purposely destroyed (see Talbot, p. 140). 51. See Talbot, pp. 134-41.

APPENDIX B 241

sources, coupled with fieldwork at the temple site themselves, however, any analysis of this question must remain highly speculative. One preliminary question which is somewhat more tractable, concerns the possible historical continuity between the Kälāmukhas and the Viraśaivas. In other words, to what extent is it possible to characterize Viraśaiva beliefs and practices as an outgrowth or further development of Kālāmukha-Pāsúpata beliefs and practices? Writing twenty odd years ago, I argued that "the evidence suggests that the Viraśaiva movement was a reformist schism from the Kälāmukha church with Basava cast in the role of Luther."52 Today, on second thought, I think this opinion needs some revision. It cannot be denied that there is at least some historical continuity between the two movements.53 To begin with, both the Kalämukhas and the Viraśaivas are worshippers of Siva. Both place special emphasis on the worship of this god in the form of a linga. Kālāmukha preceptors are sometimes called jangamas or jangama-lingas ("moving lingas") as are Viraśaiva preceptors. The geographical spread and concentration of the two movements is nearly identical, and many former Kālāmukha temples were subsequently occupied by the Virasaivas. The intellectual centers of both the movements are large monasteries. There are even a few doctrinal similarities between Pasupata and Viraśaiva theology. There is, therefore, a strong likelihood that the jangamas to whom Basava, then said to be a minister of the Kalacuri king Bijjala (c. 1145-1167 A.D.), supposedly gave large donations from the State treasury were none other than the Kālāmukhas. In addition, we have the ambiguous figure of Ekāntada Rāmayya. He is described as a great Saiva saint, without specifically calling him a Viraśaiva, in Viraśaiva literature. The story of his victory over some Jains in the time of Bijjala is described in both the Canna-Basava Purana and in an undated inscription from the Someśvara temple at Ablūr in Dharwar District of about 1200 A.D.54 In this inscription Ekāntada Rāmayya is said to have delivered a sermon in the Brahmeśvara temple

52. See above-pp. 167-68. 53. The assertions of this paragraph are more fully documented above-pp. 167-72. 54. See above-pp. 169-70 for further details. The Ablür inscriptions have been edited by J.F. Fleet, Epigraphia Indica, 5(1898-99), 213-65 (n.b. no. E). P.B. Desai has edited some further Ablür inscriptions relating to Ekantada Ramayya and others in Epigraphia Indica 29 (1951), 139-44. See also Gopal and Ritti, Summaries ... , nos. 40, 50-54 of 1948-49. See also Archaeological Survey of India, ARIE for 1958-59 (Delhi, 1963), stone no. 543.

242 APPENDIX B at Ablür, which is known from other records to have been controlled by the Kālāmukhas until at least 1144 A.D. Was Ekāntada Rāmayya at that time a Kälamukha? There is really no way of knowing. The Someśvara temple at Ablūr also contains relief sculptures depicting Ektäntada Rāmayya's victory over the Jainas together with scenes from the legends of the Virasaiva saints Jēdara Dāsimayya and Siriyālasetti and Cangālavve. Ekāntada Rāmayya is also mentioned in several Viraśaiva inscriptions from other sites in Karnataka. A grant of uncertain date from Kempanapu in Mysore District mentions the Vīraśaiva teacher Ekānta Basavēśvara, described as a descendant of Ekānta Rāmeśvara, probably the same as Ekāntada Rāmayya.55 Another inscription from Mandhya District, Karnataka, includes Ekāntada Rāmayya's name in a list of Viraśaiva saints. The record bears a date equivalent to either 1280 or 1305 A.D., about 150 years after his famous victory over the Jains.56 Nonetheless, the differences between the Kālāmukha and Viraśaiva movements are even more striking. On the whole, Virasáiva theology and Kālāmukha or, more accurately, Pāsupata theology bear little resemblance to each other. The characteristic sat-sthala doctrine of the Virasaivas may have become popular only some time after Basava, but in any case it displays little similarity to the pentads of the Pāsupata texts. More important, the entire atmosphere of devotionalism and of religious and social non-conformism found in the early Virasaiva vacanas has no observable point of contact with the orthodox scholasticism purveyed in the Kālämukha monasteries. It can even be said the two movements represent antipodes of Indian intellectual and religious tradition the Brahmanic and the anti-Brahmanic, the scholastic and the devotional, Sanskrit learning and vernacular poetic inspiration, pan-Indian culture and regional culture, social and spiritual hierarchy and social and spiritual equality. Even if we admit the existence of some elements of historical continuity between the Kālāmukhas and early Viraśaivas, there is no doubt that Basava and his fellow Viraaivas revolutionalized many of the cultural and social values of the earlier movement. Virasaivism represented not "a reformist schism of the Kālämukha church" but rather its overthrow. The extent to which later Viraśaivas subsequently fell back into a religious and social conservatism similar to that of the Kālāmukhas is another question best left to others to discuss.

55. Epigraphia Carnatica, vol. 6 (new edition), no. 144 (pp. 86, 632). 56. Epigraphia Carnatica, vol. 7 (new edition), no. 13 (pp. 197, 558).

David N. Lorenzen has been a professor in the Center of Asian and African Studies of E1 Colegio de Mexico in Mexico City since 1970. In 1989 he was a visiting professor at the University of Iowa. He has made numerous research trips to India under the auspices of both the India- Mexico cultural exchange agreement and the American Institute of Indian Studies. His recent publications include several articles on the Kabir Panth, on Gupta history, and a forthcoming book on Kabir legends. He is currently editor of the Spanish language journal, Estudios de Asia y Africa, published by EI Colegio de Mexico.