Notes

INTRODUCTION

LOCATING THE TRADITION

  1. The scholarly literature on puja is not large, considering its importance within Hindu traditions. For descriptions of puja in various schools, see James Burgess, “The Ritual of Ramesvaram,” Indian Antiquary 12 (1883): 313-26; T. Goudriaan, “Vaikhanasa Daily Worship according to the handbooks of Atri, Bhrgu, Kasyapa, and Marici,” Indo-Iranian Journal 12 (1970): 161-215; R. V. Joshi, Le ritual de la devotion Krsnaite (Pondich6ry: Institut Fransais d’Indologie, 1959); K. Ranga chari, The Sii Vaisnava Brahmans (Madras: Government Press, 1931); and Mrs. Sinclair Stevenson, The Rites of the Twice-Born (London: Oxford University Press, 1920), pp. 368-400. The most useful translations of puja texts are those of Ηέΐέηβ Brunner-Lachaux, Somaiambhupaddhati, premiere partie (Pondichdry: Institut Franiais d’Indologie, 1963), and T. Goudriaan, Kasyapa’s Book of Wisdom (The Hague: Mouton, 1965). Two notable ethnographic accounts are Paul B. Courtright, “On This Holy Day in My Humble Way: Aspects of Puja,” in J. P. Waghome and N. Cutler (eds.), Gods of Flesh/Gods of Stone (Chambersburg, Pa.: Anima Publica tions, 1985), pp. 33-50; and Akos Ostor, The Play of the Gods (Chicago: Univer sity of Chicago Press, 1980).

  2. The best general overview of Śaiva siddhanta literature is Jan Gonda, Medi eval Religious Literature in Sanskrit (Wiesbaden: Otto Harrassowitz, 1977). While our knowledge of this corpus of texts has been substantially augmented since Gonda wrote, largely through the effort of scholars associated with the Institut Franfais d’Indologie in Pondichdry, there has not yet been any synthetic account incorporat ing this new information. I attempt to cite in these notes a large portion of recent scholarly work, so a diligent reader may follow up on topics of interest.

  3. I use the 1975 devanagari edition of Kamikagama edited by C. Swaminatha siva. This edition virtually repeats the 1909 grantha text edited by Mayilai Ala gappa Mudaliar, which in turn was based on seven separate manuscripts. I have also consulted a high-quality manuscript of KA formerly belonging to Sri Swaminatha Sivacarya of Tiruvatuturai matha, now held by the Institut Fran;ais d’Indologie at Pondichdry (T.298A). I use alternate readings based on this manuscript in a few cases where they seem clearly preferable. For Kriyakramadyotika, I follow the 1927 grantha edition (based on four manuscript sources) published by the Jfianasamband ham Press in Cidambaram, which includes Nirmalamani’s commentary. The 1967 edition of the South Indian Archakas Association follows the 1927 edition generally but contains numerous mistakes and adds many apocryphal passages. It is clear from a comparison of the texts that the 1927 edition of KKD is much truer to the text as it was available to Nirmalamani in the sixteenth century. A critical edition of this central Śaiva siddhanta ritual text, based on thirty manuscripts, is currently being prepared under the direction of S. S. Janaki at the Kuppuswami Sastri Research Institute, Madras.

166 * Notes to Introduction

  1. The authoritative dynastic history of the Colas remains K. A. Nilakantha Sas tri, The Colas (Madras: University of Madras, 1935).

  2. RSjaraja I adopted the practice of using a standardized meykkirtti to introduce his inscriptions. For one example, see E. Hultzsch, “Inscriptions at Mamallapu ram,” South Indian Inscriptions 1 (1890): 63-66.

  3. An excellent account of the temple, covering diverse topics, is J. M. Soma sundaram, The Great Temple at Tanjore (Madras: Solden & Co., 1935). For a re cent architectural description, see K. R. Srinivasan’s account in Michael W. Meis ter (ed.), Encyclopaedia of Indian Temple Architecture: South India, Lower

Dravidadesa (Delhi: Oxford University Press, 1983), pp. 234—41. 7. These extensive inscriptions were edited and translated by E. Hultzsch, “Inscrip tions of the Tanjavur Temple,” South Indian Inscriptions 2 (1891-1913). Useful studies based primarily on this epigraphic material include: R. Nagaswamy, “South Indian Temple—As an Employer,” Indian Economic and Social History Review 2 (1965): 367-72; K. A. Nilakantha Sastri, “The Economy of a South Indian Temple in the Cola Period,” in A. B. Dhruva (ed.), Malaviya Commemoration Volume (Benares: Benares Hindu University, 1932), pp. 305-19; George W. Spencer, ‘Temple Money-Lending and Livestock Redistribution in Early Tanjore," Indian Economic and Social History Review 5 (1968): 277-93; and George W. Spencer, “Religious Networks and Royal Influence in Eleventh-Century South India,” Jour nal of the Economic and Social History of the Orient 12 (1969): 42-56. 8. A local tradition, first reported by Bumell in 1879, claims that the liturgy of the RajaiSjeSvara temple followed the Mtdcufagama (the “crown” agama, one of the twenty-eight mulagamas). See A. C. Burnell, A Classified Index to the Sanskrit Mss. in the Palace of Tanjore (London: Trubner & Co., 1879). The published edi tion of Makufagama, however, is highly corrupt; a critical edition would be neces sary before a reasonable assessment of this claim could be made. See C. Svami natha Sivacarya (ed.), Makufagama, piirvabhaga (Madras: South Indian Archakas Association, 1977).

  1. Much of this brief depiction of “temple Hinduism” draws on the work of Ronald Inden, through his published and unpublished writings, classes at the Uni versity of Chicago between 1978 and 1980, and many discussions since. For a dis cussion of temple Hinduism as a transformation of previous ideological formations, see R. Inden, “The Ceremony of the Great Gift (Mahadana): Structure and His torical Context in Indian Ritual and Society,” Asie du Sud, traditions et change ments, Colloques Internationaux du CNRS, no. 582 (Paris: Editions du Centre na tional de la recherche scientifique, 1979), pp. 131-36. Nicholas Dirks, “Political Authority and Structural Change in Early South Indian History,” Indian Economic and Social History Review 13 (1976): 125-58, is also useful in specifying this trans formation. Other pertinent articles by Inden include “The Ceremonial Bath of the Hindu King of Kings” and “Imperial Formations, Imperial Puranas” (both un published).

  2. For accounts of the varied activities of medieval South Indian temples, see the well-titled essay by Κ. V. Soundara Rajan, “The Kaleidoscopic Activities of Medieval Temples in the Tamilnad,” Quarterly Journal of the Mythic Society 42 (1952): 87—101; and Burton Stein, “The Economic Function of a Medieval South Indian Temple,” Journal of Asian Studies 19 (1960): 163-76. Three important dis-

Notes to Introduction · 167

cussions of the general role of temple and temple ritual in medieval South Indian society and politics are: A. Appadurai and C. A. Breckenridge, “The South Indian Temple: Authority, Honour and Redistribution,” Contributions to Indian Sociology 10 (1976): 187-211; Nicholas B. Dirks, The HoUow Crown: Ethnohistory of an Indian Kingdom (Cambridge: Cambridge University Press, 1987); and Burton Stein, Peasant State and Society in Medieval South India (New Delhi: Oxford Uni versity Press, 1980).

  1. ΪΡ vol. 3 ch. 30. Discussed in Stella Kramrisch, The Hindu Temple (Delhi: Motilal Banarsidass, 1946-1976), vol. 1, pp. 261-70.

  2. VayusanMta, cited in VarQaSramacandrika, translated in SP 1 p. vi. 13. A list is given in Bhatt (ed.), RA 1, opposite p. xix.

  3. For an example, see AA 1.1.35—105. But agamas dispute among themselves which mouths emit which agamas. See Jean Filliozat, “Introduction: Les agamas Sivaites,” RA 1 pp. v-xv.

  4. E. Hultzsch, “The Pallava Inscriptions on the Kailasanatha Temple at Kan chipuram,” South Indian Inscriptions 1 (1890): 8-24.

  5. Bruno Dagens (trans.), “Introduction,” MM 1 pp. 1-7.

  6. See Ηέΐέηβ Brunner-Lachaux’s discussion of the relation between KA and MrA in her “Introduction,” MrA (Section des rites et section du comportement), pp. x-xix.

  7. Hara Prasad Sastri, A Catalogue of Palm-Letrf and Selected Paper Mss. Be longing to the Durbar Library, Nepal (Calcutta: Baptist Mission Press, 1905, 1915), vol. 2, p. xxiv.

  8. Narayanakantha quotes a verse of Utpaladeva, who lived probably in the mid ninth century; Ksemaraja’s commentary on the SvaT, composed in the eleventh century, in turn quotes Narayanakantha’s MrAV. See Bhatt (ed.), MPA 1 p. xii, and Brunner-Lachaux (trans.), MrA (Section des rites), p. vii and n. 6.

  9. Ηέΐέηε Brunner touches briefly on these points of disagreement in “Impor tance de la littdrature agamique pour l'6tude des religions vivantes de TInde,” In dologica Taurinensia 3-4 (1975—1976): 107—24. On the various lists of tattvas, see Bhatt, “Introduction,” MPA 1 pp. xxi-xxiv. Surendranath Dasgupta summarizes the philosophical positions of several agamas and related texts (including MPA, PA, V$A, and SPur) in A History of Indian Philosophy, Vol. 5: The Southern Schools of Saivism (Delhi: Motilal Banarsidass, 1922-1975).

  10. Compare Abhinavagupta’s account (in Tantraloka, ch. 36) of the division of agama-based Saivism into three propensities, as K. C. Pandey summarizes it:

[Siva] instructed the sage, Durvasas, to revive the Śaivagamic teaching. The sage accordingly divided all the saivagamas into 3 classes according as they taught mo nism, dualism or monism-cum-dualism, imparted their knowledge to his 3 mind-bora sons, Tryambaka, Amardaka, and Srinatha respectively, and charged each one of them separately with the mission of spreading the knowledge of their respective Agamas. Thus there came into existence three Śaiva Tantric Schools, each known by the name of the first earthly progenitor.

K. C. Pandey, Abhinavagupta: An Historical and Philosophical Study, Chow khamba Sanskrit Series, vol. 1 (Benares: Chowkhamba Sanskrit Series Office, 1935), p. 72.

168 • Notes to Introduction

  1. David N. Lorenzen discusses these and many other references to the four (most often) Śaiva schools. See The Kdpalikas and Kalamukhas: Two Lost Saivite Sects (Berkeley: University of California Press, 1972), pp. 1-12. For a general summary of several individual schools of Saivism, see also Pranabananda Jash, His

tory of Saivism (Calcutta: Roy and Chaudhury, 1974).

  1. Ηέΐέηε Brunner discusses the Śaiva siddhanta characterization of its main Śaiva competitor in “The PaSupatas as seen by the Śaivas,” Schriften zur Geschichte und Kultur des alien Orients 18 (1986): 513-20.

  2. Significant secondary sources concerning Śaiva siddhanta monastic lineages include: R. D. Baneiji, The Haihayas of Tripuri and Their Monuments, Memoirs of the Archaeological Survey of India, no. 23 (Calcutta: Government of India, 1931); V. V. Mirashi, “The Śaiva Acaryas of the Mattamayiira Clan,” Indian Historical Quarterly 26 (1950): 1-16; V. V. Mirashi (ed.), Inscriptions of the Kalachuri-Chedi Era, Corpus Inscriptionum Indicarum, vol. 4 (Ootacamund: Government Epigraph ist for India, 1955); J. Van Troy, “The Social Structure of the Śaiva-siddhantika Ascetics (700-1300 A.D.),” Indica 11 (1974): 77-86; B.G.L. Swamy, “The Golaki School of Saivism in the Tamil Country,” Journal of Indian History 53 (1975): 167-209; and Cynthia Talbot, “Golaki Matha Inscriptions from Andhra Pradesh: A Study of a Śaiva Monastic Lineage,” in Vajapeya: Essays on Evolution of Indian Art and Culture (Delhi: Agam Kala Prakashan, 1987), pp. 133-46. For Southeast Asia, see K. Bhattacharya, Les religions brahmaniques dans I’Ancien Cambodge (Paris: Ecole Franfaise d’ Extreme-Orient, 1961).

  3. As a tenth-century inscription from Madhya Pradesh refers to one important Śaiva siddhanta branch, the Mattamayiira. See F. Kielhorn, “A Stone Inscription from Ranod (Narod),” Epigraphia Indica 1 (1892): 351-61.

  4. As Brunner puts it, the paddhatis aim at “unification of the often contradic tory teachings of the agamas, and have performed their function so well that they have ended up somewhat eclipsing the agamas themselves” (“Importance de la litt&ature agamique,” p. 110).

  5. On AghoraSiva’s life, see Wayne Surdam, “South Indian Śaiva Rites of Initi ation: “The DHc§avidhi’ of AghoraSivacarya’s ‘Kriyakramadyotika’” (Ph.D. disserta tion, University of California, Berkeley, 1984), pp. xvi-xxi. AghoraSiva’s known works are listed in V. Raghavan, New Catalogus Catalogorum (Madras: University of Madras, 1968), vol. 1, pp. 58-59.1 have attempted to summarize the intellectual background of Aghorasiva in “Aghora&va’s Background,” in Dr. S. S. JanaM Sha shtyabdapurti Commemoration Volume (Madras: Kuppuswami Sastri Research In stitute, forthcoming).

  6. For two scholarly summaries of Śaiva siddhanta doctrine based on the paddhati literature, see Rohan A. Dunuwila, Śaiva Siddhanta Theology: A Context for Hindu-Christian Dialogue (Delhi: Motilal Banarsidass, 1985) (based primarily on Bhojadeva’s TP); and K. C. Pandey, An Outline of History of Śaiva Philosophy (Delhi: Motilal Banarsidass, 1986) (first printed in 1955 as Bhaskari, vol. 3). Jayan

dra Soni, Philosophical Anthropology in Śaiva Siddhanta (Delhi: Motilal Banar sidass, 1989), is based primarily on the works of the sixteenth-century Śaiva author Sivagrayogin.

  1. The best general account of the Tamil version of Śaiva siddhanta is M. Dha-

Notes to Introduction · 169

vamony. Love of God according to Śaiva Siddheinta (Oxford: Clarendon Press, 1971). Also useful are the collected lectures of V. A. Devasenapathy, with the excellent title Of Human Bondage and Divine Grace (Cidambaram: Annamalai University, 1963), and the study of K. Sivaraman, Saivism in Philosophical Per

spective (Delhi: Motilal Banarsidass, 1973). A primary agenda of most historical scholarship on Tamil Śaiva siddhanta works of the thirteenth century and later has been to develop a genealogy relating these texts directly to the works of the Tamil nayanmars of the sixth through ninth centuries. Such studies generally pay lip ser

vice to the role of the agamas in the philosophical system of Meykantar and his fol lowers, but they do not seriously investigate the connections between the Sanskrit dgama literature and the later Tamil works.

  1. The only critical work devoted to this, as far as I know, is the brief discus sion by Brunner, “Importance de la littirature agamique,” pp. 118-19. For another example of a Śaiva author working in both Sanskrit and Tamil, see Bruno Dagens’s discussion of the sixteenth-century commentator Vedajfiana/Maiaifiagatecikar, in “Introduction,” SPM, pp. 5-15. I am indebted to Dagens for leading me to recon sider the relation of Sanskrit and Tamil Śaiva siddhanta literature.

  2. On the poetry of the nayanmars, see especially the recent study and transla tions by Indira Peterson, Poems to Siva (Princeton: Princeton University Press, 1989).

  3. On the use of texts in contemporary temple practice, see Carl Gustav Diehl, Instrument and Purpose: Studies on Rites and Rituals of South India (Lund: C.W.K. Gleerup, 1956). For a portrait of a Śaiva priestly community in theological disarray, see Chris Fuller’s fascinating ethnography on the Madurai Mmak§i-Sundare£varar temple priests, Servants of the Goddess: The Priests of a South Indian Temple (Cambridge: Cambridge University Press, 1984).

  4. G. U. Pope (trans.), The nruvacagam, or “Sacred Utterances” of the Tamil Poet, Saint, and Sage ManHckavacagar (Oxford: Clarendon Press, 1900), p. lxxiv. The same position turns up repeatedly in general works on Saivism and Hinduism. For instances, see: M. Dhavamony, “Saivism: Śaiva Siddhanta,” in Mircea Eliade (ed.), Encyclopedia of Religion (New York: Macmillan Publishing Co., 1987), vol. 13, p. 11; and Thomas J. Hopkins, The Hindu Religious Tradition (Encino, Cal.: Dickenson Publishing Co., 1971), p. 118.

  5. For a more skeptical discussion of the question of integration between jnana and kriya, see Ηέΐέηβ Branner-Lachaux, “Introduction,” MrA (Section des rites), esp. pp. xlii-xliii.

  6. The texts set out an ideal program for worship that only the highly adept and devout could follow completely. Yet they also recognize and allow for individual variations and shortcuts in phrases repeated throughout, such as yathaiakti (in ac cord with one’s capacity).

  7. Agnipurana (ch. 72-106), LiAgapurana, and Sivapurana (Vayaviyasamhita ch. 16-20) clearly articulate siddhanta positions. (Brunner briefly discusses the pu ranic usage of agamic material in “Importance de la littirature agamique,” pp. 117 and 121.) Silpaiastras such as MM also reflect Śaiva siddhanta practice, but through the eyes of the sthapati rather than the acarya.

  8. I mention in the notes some of the disputes and differing formulations within

170 · Notes to Chapter One

the Śaiva siddhanta order. In her translation of SP, and particularly in vol. 3, Ηέΐέηβ Brunner-Lachaux presents a much more variegated portrait of Śaiva sid dhanta literature, focusing repeatedly on doctrinal discussion within the order.

CHAPTER ONE

RITUAL AND HUMAN POWERS

  1. Some Śaiva texts add “volition” (iccha) as a third inherent power of con sciousness.

  2. The agamas of course recognize that not everyone seeks mok$a in this lifetime; the world includes bubhukfus (“seekers of worldly enjoyments”) as well as mumukfus (“seekers of liberation”). But worldly powers and liberation do not lie along radi cally differing courses of action. Both result from the removal of fetters and the manifestation of inherent powers. Sadhakas (mantra-adepts), for instance, employ Śaiva ritual and knowledge to gain specific worldly powers. For a full discussion, see Ηέΐέηε Brunner, “Le sadhaka, personnage oubli6 du Sivaisme du Sud,” Journal asiatique 263 (1975): 411-43.

  3. Contrary to what many modern advocates of Śaiva siddhanta have held, the igamas do not portray themselves as derived from or in any way subordinate to the Vedas. On this matter, Ηέΐέηε Brunner’s “Le Śaiva-siddhanta, ‘Essence’ du Veda,” Indologica Taurinensia 8 (1980-1981): 51-66, is a welcome statement. For a some

what different formulation, see Richard Davis, “Cremation and Liberation: A Śaiva siddhanta Revision,” History of Religions 28 (1988): 37-53. Wayne Surdam, “The Vedicization of Śaiva Ritual,” in S. S. Janaki (ed.), Siva Temple and Temple Ritu als (Madras: Kuppuswami Sastri Research Institute, 1988), pp. 52-60, deals with the relationship as it has changed over time.

  1. MrAVD vidya 1.20, from Michel Hulin’s French translation. 5. Kamikagama refers to this useful technique as an analogy to the efficacy of “interior worship”:

Visualizing with his mind, he should daily worship ParameSvara in this imagined temple, using substances that are mentally constructed. As a guru meditates that he is Garuda and obtains the result of removing [poison, even though done mentally], so here also one obtains the desired benefits [resulting from worship], (KA 4.189- 90)

  1. This useful convention I owe to Ηέΐέηε Brunner-Lachaux, who has employed it in her translations and studies of saivagama.

  2. Nirmalamam understands the Śaiva form of meditation, in its concern to estab lish a relationship between meditator and divinity, to differ from Pataiijali yoga: ‘Toga is the close connection (sambandha) during meditation with the object of meditation; not, as in Patafijali, the condition of samadhr (KKDP, translated in Surdam, “South Indian Śaiva Rites,” p. lv, n. 102).

  3. KarA, cited by NirmaIamaQi in KKDP, and quoted in Janaki, ML p. 2. 9. For detailed discussion of mudras in Śaiva ritual, see the series of six articles by S. S. Janaki, “Śaiva Mudras I-VI,” Kalakshetra Quarterly 5-6 (1983-1984). Photographs of the mudras may be found in those articles and also in SP 1, Plate 1, “Mudra,” and in Surdam, “South Indian Śaiva Rites,” Appendix, pp. 276-313. In

Notes to Chapter One · 171

S. S. Janaki (ed.), ML, line drawings of mudrds following ML prescriptions are provided.

  1. V$A, from Brunner-Lachaux’s French translation, SP 1 p. xxxi. 11. On the grammar of Śaiva mantras, see Ηέΐέηβ Brunner-Lachaux’s brief dis cussion in SP 1 pp. xxxi-xxxii. Brunner-Lachaux also includes a list of the mantras employed in nityapuja in SP 1, Appendix 3, “Mantras utilises pendant Ie culte de Siva (d’apr&s Somasambhu).”

  2. Vedajfiana apparently quotes MPA here. Other authors construct similar scales of forms: “Proponents of various schools describe the inherent form of the highest mokfa in various ways, each in accord with its own views.” With this pref ace, Sivagrayogin considers, and criticizes, a whole series of alleged mok$a& held by other schools (&Pbh pp. 335-52), including the Sivasdmya views of AghoraSiva that I take here as normative.

  3. This includes both purva and uttara portions of the text. For a summary of the contents of KA, purvabhaga, see Bruno Dagens, “Analyse du Purvaka mikagama,” Bulletin de I’Ecole Frangaise d’Extreme-Orient 54 (1977): 1-38. Dagens’s rdsum6 is based on N. R. Bhatt’s Sanskrit upodghata, in KA, purvabhaga, pp. iii-xx. Two other agama r6sum6s are available: Ηέΐδηε Brunner, “Analyse du Kiranagama,” Journal asiatique 253 (1965): 309-28; and Brunner, “Analyse du Su prabhedagama,” Journal asiatique 255 (1967): 31-60. These two dgamas cover a more modest array of rituals but are noteworthy for containing all four padas of a proper agama.

  4. SupA, quoted in SASS p. 58.

  5. The dgamas generally present atmartha and pardrtha as two variants of a single scheme of piijd, and following their lead I will describe Śaiva daily worship in this study as a ritual unity. However, there are some practical differences be tween the two. Ηόΐέηε Brunner compares them in greater detail in “Atmarthapuja versus Pararthapuja in the Śaiva Tradition,” in T. Goudriaan (ed.), The Sanskrit Tradition and Tantrism (Leiden: E. J. Brill, 1990), pp. 4—23. She argues that the integral model of piijd assumed by later Śaiva literature in fact results historically from the combining of two originally distinct styles of worship, one an “old public worship” of a rudimentary sort, and the other an inner-oriented “private worship” maintained by yogic adepts.

  6. Frits Staal has discussed the “embedding” of ritual units within one another to constitute larger ritual wholes in the Vedic system: “Ritual Syntax,” in M. Naga tomi et al. (eds.), Sanskrit and Indian Studies: Essays in Honour of Daniel Η. H. Ingalls (Dordrecht: D. Reidel Publishing Co., 1980), pp. 119-42. As will be clear throughout this study, I depart strongly from Staal’s view that such embeddings are the empty exercise of syntactic elements without discursive significance.

  7. Svatantra, quoted by Nirmalamani in KKDP p. 293, cited in SP 3 p. 9. 18. To assist those who wish to reenact Śaiva piijd visually as well as mentally, the Kuppuswami Sastri Research Institute has recently made available an excellent videotape of K. A. Sabharatna Sivacarya performing an dgama-based worship of Siva, with comentaiy in English or Tamil. Readers wishing to obtain a copy may contact Ginni Ishimatsu, Department of South and Southeast Asian Studies, Univer sity of California, Berkeley, CA 94702, or Richard Davis, Department of Religious Studies, Yale University, New Haven, CT 06520.

172 · Notes to Chapter Two

CHAPTER TWO

OSCILLATION IN THE RITUAL UNIVERSE

  1. The classic 1918 essay by A. K. Coomaraswamy, “The Dance of Shiva,” in The Dance of Shiva (New Delhi: Sagar Publications, 1969), pp. 66-78, remains the best published account of the significance of Nafaraja as explicated by Tamil Śaiva siddhanta authors. On the origin of this iconic form, there is considerable scholarly debate. Starting about 970 C.E. the dowager Cola queen Sembiyan MahadevI, grandmother of Rajaraja, began placing Natarija in a prominent position on the exte

rior walls of the temples she sponsored, and I would date the elevation of Nafaraja to most-favored status among Cola images of Siva to her initiative. 2. The Śaiva scheme of tattvas clearly draws upon the Samkhya model of emanat ing prakrti but surpasses it by adding additional tattvas and an entire domain, the pure domain, not envisioned within Samkhya circles. See Dasgupta, History, Vol. 5: Southern Schools, pp. 164—70, for some discussion of Śaiva siddhanta in relation to Samkhya.

  1. On the emission of language and the alphabet, see KA 2.4—7. Ramakantha gives a more sophisticated treatment of language in the Śaiva siddhanta universe in his Nddakarika (with a commentary by AghoraSiva), included also in his KalAV (l.S). See Pierre-Sylvain Filliozat, “Les Nadakarika de RamakaQfha,” Bulletin de I’Ecole Frangaise d’Extreme-Orient 73 (1984): 223-35, for a translation and informed ex

plication.

  1. For a general study of nyasa as a ritual technique, see Andre Padoux, “Contri butions a l'6tude du mantraSastra,” Bulletin de I’Ecole Frangaise d’Extreme-Orient 67 (1980): 59-102.

  2. Further aspects of the five mantras are set forth by Brunner-Lachaux in SP 1, Appendix 6, ‘Xjuelques correspondences entre les visages de SadaSiva et Ie Cosmos.” 6. Some texts omit NETRA from the set of angamantras, which may account for its distinctive treatment in karunyasa and elsewhere.

  3. A much more detailed treatment of Siva’s angas may be found in Ηέΐέηβ Brun ner, “Les members de Siva,” Asiatische Studien 40 (1986): 89-132. 8. Mj-A kriya 3.3 offers an exception. Here the weapon mantra ASTRA should be placed on the index finger (tarjani, literally the “threatening finger”) because ASTRA is the mantra that “threatens intruders.”

  4. See Brunner-Lachaux, SP 1, Appendix 5, “Purification du corps grassier: les cinq mandala des 616ments,” for a r6sum6 of the features of each domain. 10. A shorter alternative method of bodily purification described in both KKD and KA involves visualizing the body as an upside-down tree and progressively de stroying it (KKD p. 59). See Brunner-Lachaux’s translation of Aghorasiva’s ac count in SP 1, Appendix 4, “Purification du corps grossier, autre methode.” 11. On the thirty-eight kalas and the kalamantras, see N. R. Bhatt’s summary in RA 1 pp. 25-28 and the table, “Les noms des Kala dans les textes,” opposite p. 28. Hie mantras are derived from five Vedic mantras in Taittiriya Aranydui 10.43-47, refracted and reformulated within a Śaiva grammar of mantras. 12. See N. R. Bhatt, “Introduction,” MPA 2 pp. xi-xvi; also Ηέΐέηε Brunner Lachaux, SP 3 p. 242 and Plate 8, “Distribution des mots du vyomavyapin entre les cinq kala selon !‘Adhvanyasa.”

Notes to Chapter Two · 173

  1. The classic secondary account of vastumandala is Stella Kramrisch, The Hindu Temple, vol. 1, pp. 19-97. Kramrisch utilizes Śaiva siddhanta texts, particularly IP, as well as texts belonging to many other schools, for her synthetic description. Some of the relevant chapters of IP are translated by Kramrisch in “ISanaSivagu

rudevapaddhati Kriyapada Chs. xxvi, xxvn,” Journal of the Indian Society of Ori ental Art 9 (1941): 151-93; and ‘Temple, Door, Throne, Etc.," Journal of the In dian Society of Oriented Art 10 (1942): 210-52 (which translates IP kriya ch. 5, 7, 12, and 13).

  1. The phrase is from Mark Twain, Following the Equator (Hartford: American Publishing Company, 1898), p. 504, quoted in Diana L. Eck, Banares: City of Light (Princeton: Princeton University Press, 1982), p. 19.

  2. Compare the model of a royal court presented in the Aparajitaprccha by the twelfth-century Gujarati author Bhuvanadeva and described in Ronald Inden, “Hi erarchies of Kings in Early Medieval India,” in T. N. Madan (ed.), Way of Life: King, Householder, Renouncer (Paris: Editions de la Maison des Sciences de 1’Homme, 1982), pp. 99-125.

  3. The architectural prescriptions of AA are lucidly summarized in Bruno Dagens, Les enseignements architecturaux de VAjitagama et du Rauravagama (Pondichdry: Institut Fran^ais d’Indologie, 1977), translated into English as Archi tecture in the Ajitagama and the RauravSgama (New Delhi: Sitaram Bhartia Insti tute of Scientific Research, 1984). See Table 5, “Disposition des Assesseurs selon YAjita” for a diagram of AA’s model temple topography.

  4. KA 4.225-69. The UKA ch. 4 also describes these diagrams, in most but not all cases congruent with the versions in the piuvabh&ga. A still more detailed ac count is found in RA 1 ch. 20-24.

  5. Compare the diagrams based on RA prescriptions, Plates 6-11, and those based on AA, Plates 2-8.

  6. A fine photograph of well-dressed Hvakumbha and vardhani may be seen in Bhatt (ed.), MrA, Plate 13, opposite p. 93.

  7. The procedure summarized here is based on “NavakalaSasnapanapaddhati,” a manuscript belonging to J. Visvanatha Gurukkal of Melmangalam and quoted in extenso by N. R. Bhatt in RA 1 pp. 92-93, n. 9.

  8. The adiiaivas constitute a special category of “original Śaiva” brahmans, made up of five gotras, who alone qualify to be priests in Śaiva siddhanta temples. For a full discussion, including intriguing remarks concerning their origin, see Ηέΐέηε Brunner, “Les categories sociales v6diques dans Ie Sivaisme du sud,” Jour

nal asiatique 252 (1964): 451-72.

  1. SupA, quoted in N. R. Bhatt, “What Is Śaivagama?” (pamphlet, no publica tion information), p. 2.

  2. APV, quoted in Bhatt, “What Is Śaivagama?” p. 3. For similar rules from other schools concerning differentiated temple access, see Ronald Inden, “The Temple and the Hindu Chain of Being,” Purusartha 8 (1985).

  3. KarA 24.95, quoted in K. A. Sabharatna Sivacarya, ‘XIicava Vimarcaoam," in S. S. Janaki (ed.), Siva Temple and Temple Rituals (Madras: Kuppuswami Sastri Research Institute, 1988), p. 93.

  4. KarA 141.1-2, quoted in Sabharatna Sivacarya, “Urcava Vimarcaoam,” p. 92.174 · Notes to Chapter Three

CHAPTER THREE

BECOMING Λ SIVA

  1. V. A. Devasenapathy cites the pleasing analogy of a prince raised among gypsies:

The soul is like a prince kidnapped in his infancy by gipsies and brought up by them in ignorance of his real identity. It is natural for the prince in such a state to behave in gipsy ways mistaking these as natural to him. But when the king comes and reveals his identity, the prince will give up his gipsy ways and conduct himself as befits a prince.

V. A. Devasenapathy, OfHuman Bondage and Divine Grace (Cidambaram: Anna malai University, 1963), p. 48.1 have not been able to locate the textual source of this simile.

  1. Śaiva texts refer to two other categories of fetters: tirobhava (or rodhiakti) and mahamaya. Tirobh&va, Siva’s power of obscuration, is a fetter only in a secon dary sense, says AghoraSiva, while mahamaya affects only beings such as the VidyeSvaras who dwell in the pure domain (^uddhadhvan) (TPV 17).

  2. The hierarchy of souls described in Chapter 1 is based on the presence or absence of these three fetters. As bound souls, sakalas, we are affected by all three categories of fetters (TP 8-15).

  3. The monist commentator Snkumara, discussing Bhojadeva’s TP, claims that the ripening of mala is brought about by the “fire of knowledge” (TPD 9), but AghoraSiva firmly denies this (TPV 15).

  4. Ηέΐέηβ Brunner-Lachaux discusses the Śaiva understanding of karman and its three types in SP 3 pp. xxii-xxvi.

  5. Śaiva authors differ from one another in their discussions of iaktinipata. SomaSambhu, for instance, identifies saktinipata with the act of initiation, while AghoraSiva and most subsequent authors view it as a necessary precursor to receiv ing initiation. For helpful discussions of this philosophically ambiguous phenome non, see Brunner-Lachaux, SP 3 p. viii, and Surdam, “South Indian Śaiva Rites,” pp. cxiv-cxxi.

  6. My descriptions of Śaiva diksas are based largely on the SP and KKD. For translations of samayadiksa procedures, see SP 3 pp. 2-110, and Surdam, “South Indian Śaiva Rites,” pp. 1-97.

  7. On the several stages of initiation and corresponding ritual competencies, see Brunner, “Le sadhaka,” Journal asiatique 263 (1975): 411-43. 9. Translations of vi§e$adiks& procedures: SP 3 pp. 112-56, and Surdam, “South Indian Śaiva Rites,” pp. 98-112. Brunner-Lachaux argues that samaya and visesa were initially two phases of a single ritual and were separated into distinct rituals at a fairly late date (SP 3 pp. xxx-xxxiii); these two in turn may have at an earlier date been separated from nirvanadlksa, the originally unitary Śaiva diksa. To gain an idea of the variety of Śaiva treatments of dikfa generally, see N. R. Bhatt, “Introduc tion,” MPA 2 pp. xviii-xxiii.

  8. See SP 3 pp. 157—426, and Surdam, “South Indian Śaiva Rites,” pp. 113— 271. The amount of discussion SomaSambhu and AghoraSiva devote to each of the three types of dik$a offer a good initial index of their importance in the Śaiva scheme

Notes to Chapter Four · 175

of things: in Surdam’s translation of KKD, samayadikja gets 96 pages, viiesadiksa 14 pages, and nirvanadlksd 1S8 pages.

  1. Texts refer to this method only as tadanadi, “the set of actions beginning with striking.” For a more detailed discussion of “transportation,” see SP 3 pp. 118-25, and Plates 1 and 2, “Transport de Yatman depuis Ie Coeur du disciple jusqu’k la matrice de VagTsvan.”

  2. The method described here, focusing on five kolas, is only one of six possi ble “paths” (adhvan) along which initiation may proceed: mantras, words, pho nemes, worlds, tattvas, and kalas. As the quotation from UKA below indicates, however, Śaiva siddhanta understands the kalas to encompass the constituents of the other paths, and hence the method of kola purification offers a more comprehen sive means of dikja (at least from the time of SomaSambhu). On the sixfold path, see Brunner-Lachaux, SP 3 pp. xiii-xxii.

  3. In SP 3, Brunner-Lachaux provides charts summarizing these englobements: see Plate 5, “Englobement des rdalitds de TUniverse par Ies cinq kalS’\ Plate 6, “Ripartition des tattva selon Ies cinq kalS’·, and Plates 7 A-E, displaying the worlds encompassed within each of the five kalas.

  4. After nirvanadikjd, a Śaiva initiate may undergo special additional consecra tions (abhifeku) that grant him particular capacities for action. Most common is the “priestly consecration” (acaryabhiseka), enabling the recipient to act as priest in per forming rituals on behalf of others. Another consecration grants special mantra

powers to the adept (sadhakabhiseka). For Soma£ambhu’s account of these conse crations, see SP 3 pp. 455-524.

  1. Nirmalamani distinguishes the efficacy of atmaSuddhi according to the wor shiper’s stage of initiation. For one who has undergone samayadiksa but not nirvdnadiksa, atmaxuddhi destroys any obstructions that prevent mantras from achieving their purposes; for those who have undergone nirvdnadiksa, by contrast, it helps complete initiation by destroying all newly arisen karman (KKDP p. 60).

  2. See SP 1, Appendix 5, “Purification du corps grossier,” for a chart of the gunas located in each of the five bhutas.

  3. See Brunner-Lachaux’s detailed discussion of this parallelism, in SP 3 pp. 396-405.

  4. SomaSambhu’s prescriptions for antye$ti are translated in SP 3 pp. 567-618. For a general discussion of this ritual in the Śaiva system, see Davis, “Cremation and Liberation.”

CHAPTER FOUR

SUMMONING THE LORD

  1. MrAVD vidya 3.8-9, following Michel Hulin’s French translation, MrA (Doc trine et Yoga), p. 121.

  2. SvdA vidya 4.3, quoted in §PM 1.2.

  3. On the nonactivity of the muktatman, see MrAV vidya 2.29 and $SPbh p. 37. 4. V$A 1.22, quoted in SPM p. 56, n. 2.

  4. AcA, quoted in SPbh p. 63.

  5. In his “Compendium of All Schools” (SDS), the monist author Madhava dis-

176 · Notes to Chapter Four

cusses the views of Śaiva siddhanta (iaivadariana) in chapter 7. He bases his de scription primarily on the MfA and Narayanakantha’s commentary MfAV, on Bho jadeva’s TP and AghoraSiva’s TPV; he also employs KirA, PA, SP, and a very few other texts. The passage here relies on MrAV 3.8. See Ηέΐέηβ Brunner’s expert trans lation, “Un chapitre du Sarvadarianasanigraham. Le Śaivadariana,” in Michel Strick man (ed.), Tantric and Taoist Studies in Honour of R. A. Stein (Brussels: Institut Beige des Hautes Etudes Chinoises, 1981), pp. 96-140.

  1. For general treatments of the lithic forms and iconographic representations of SadaSiva, see Brijendra Nath Sharma, Iconography of SadMiva (Delhi: Abhinav Publications, 1976), and Thomas S. Maxwell, “The Five Aspects of Siva (In The ory, Iconography and Architecture),” Arts International 25 (1982): 41-57.

  2. The list of angas is consistent, but the attributes vary somewhat from text to text. Compare for instance the discussions in SP 1.3.72-74 and KKD pp. 108-9. As noted in the previous chapter and implied in the passage quoted here, the angas are closely related to Ihe six divine qualities (guna) of Siva, acquired also by the initiate in dlkfd. See Brunner’s two long footnotes on the subject, SP 3 pp. 396-405.

  3. The set of weapons held by Sadasiva varies slightly from text to text. KA specifies two possible sets (4.332-34), while AghoraSiva gives three alternatives (KKD pp. 98-99). SP 1, Plate 7, “Ayudha de SadaSiva,” lists eight different sets of SadaSiva’s weapons attested in diverse agamas. However different, these sets gen erally encompass all the various weapons associated with Siva in his iconic

(MaheSvara) forms.

  1. Retellings of Siva’s puranic escapades are conveniently available in Wendy Doniger O’Flaherty, Siva: The Erotic Ascetic (Oxford: Oxford University Press, 1973), and Stella Kramrisch, The Presence of Hiva (Princeton: Princeton University Press, 1981).

  2. RA 2 ch. 35. In his “Preface,” N. R. Bhatt concisely summarizes the MaheSvara forms described in various agama texts (RA 2 pp. ix-xi). Also in this volume may be found drawings of these forms based on agama prescription.

  3. VA 63.4-5, quoted in SPM 5.151-54. As Dagens notes, the passage Vedajfiana cites does not include the expected pentad of samharanuirtis. 13. See for instance SP 3.1.27-28, where the guru identifies himself with all five manifestations of Siva.

  4. For a further discussion of the relationship between “mental” and “concrete” divine images in Śaiva worship, see Ηέΐέηε Brunner, “L’Image Divine dans Ie Culte Agamique de Siva: Rapport entre 1’image mentale et Ie support concrete du culte,” L’Image Divine. Culte et Meditation dans VHindomsme (Paris: Editions du Centre national de la recherche scientifique, 1990), pp. 9-29.

  5. This is not to say that all lingas are the same. MM goes on to describe a great many categories of lingas, made of different substances and according to different dimensions, appropriate for various types of worshipers (MM 33.72-92).

  6. Ηόΐέηε Brunner, “Toujours Ie Niskala-linga,” Journal asiatique 256 (1968): 445-47. For a fuller account linking this tripartite scheme to temple iconography and layout, see Doris Meth Srinivasan, “From Transcendency to Materiality: Para Siva, Sadasiva, and MaheSa in Indian Art,” Artibus Asiae 50, (1990): 108-42. I thank the author for sending me proofs of her essay at the time I was revising this chapter.

Notes to Chapter Four · 177

  1. See for instance KKD ^ivapratisthavidhi, p. 2. Similar sets of eighteen rites are prescribed for the establishment of an image (pnuimapratiffha) (KA 68.1-6). 18. For a more detailed examination of the reuse of Puru§a imagery in an early Vaisnava establishment ritual, see Shantanu Phukan, “From the Idol to the Icon: The Transformation in the Pratima Prati§thana Ceremony” (unpublished). 19. The majority of agamas holds that the divine throne extends from Earth up to fuddhavidya (KA 4.312, SP 1.3.S6, AA 20.157). Another school of thought, ad hered to by KirA and Narayanakantha, argues that the divine throne extends all the way up to the Saktitattva (MrAV kriya 3.12).

  2. See Figure 9 here and the more detailed chart of throne stages in SP 1, Plate 5, “Asana-puja.”

  3. The identity of this particular Sakti is not certain. She is variously assimi lated to Kurwjalinl, KriyaSakti, and IcchaSakti. (See Brunner’s discussion, SP 1 p. 156.) What is never in dispute is that it must be a differentiated form of Sakti at the base of the divine throne.

  4. The ndga-like description of Ananta in KA calls to mind another famous An anta: the cosmic snake who serves as the couch of Visnu-Narayana. Hie VidyeSvara Ananta is never identified with Visnu’s great servant, but the snake imagery is fre quent enough to suggest that the throne of Visnu has been incorporated into Siva’s more encompassing divine throne. (Brunner raises the question of Ananta’s iden tity, in SP 1 pp. 158-60.)

  5. See Brunner’s list of the Saktis and the VidyeSvaras, together with her dis cussion of the identity of these Saktis, SP 1 pp. 166-68. It is noteworthy that the names of the Saktis “are the feminine equivalents of the divine names associated with Vamadeva in the Taittiriya Aranyaka mantra, whose division served to form the 38 kola of SadaSiva.”

  6. These three categories generally include the tattvas of the impure domain as follows:

iiva—maya

vidya—six tattvas from kala to puru$a

atman—twenty-four tattvas from prakrti to Earth

These three tattvas also pervade the body: see SP 3, Plate 14, “Decoupage du corps par Ies trois tattva.”

  1. As an alternative method of visualizing the divine throne, Kamikagama de scribes it as a lotus that unfolds as the worshiper constructs it.

The pedestal is fixed upon all paths (adhvan). Its great feet are the four ages of the world. Its bulb is the element Earth. The upraised stalk constitutes the tattvas up to Time. It is decorated with thorns which are the fifty kinds of emotions. Its main node is the mayatattva. The broad lotus flower is iuddhavidya. The petals are the VidyeSvaras. This lotus is ornamented with stamens which are the Saktis, and made splendid by a pericarp and seed where the couple Siva and Sakti are located. (KA 4.316-19)

Here again the form of the divine throne encompasses both all the constituents of the impure domain (worlds, ages, tattvas, states of emotion) and the powers or agents of Siva through which he governs (VidyeSvaras, eight Saktis, Siva and Sakti).

178 · Notes to Chapter Five

(Compare KirA kriya 2.19-22 and SSV 48.) Like the divine throne of five stages, tbis lotus-form throne is a comprehensive ritual construction embodying diva’s ac tive presence in the world.

  1. I follow here the account in AA, which differs somewhat from KA and KKD, primarily because the description in AA seems to me clearer. The main difference lies in the distinction between siik$mamiirti and vidyadeha.

  2. According to Appayadiksita, still other constituents of reality are contained in the vidyadeha: the kolas, adhvans, bhuvanas, agamas, and so on ($AC pp. 56-59). The vidyadeha is of course an inclusive body. Appaya’s emphasis on this point here, however, may relate to his monistic perspective. If Siva is considered the material as well as instrumental cause of the world, then his SadaSiva form can and should include all manifest realities as well as his instrumentalities.

  3. There are in fact several formulations of the MOLA mantra, according to the Siddhantabodha, appropriate to different categories of worshipers. See Brunner Lachaux, “Introduction,” SP 1 pp. xxxii-xxxiii. The name PRASADA for this mantra suggests a homology between mantra and temple (also prasada). Brunner-Lachaux points to this when describing the final dissolution of the mantra into ParamaSi va: ‘This is the summit of the PRASADA: the temple, which is at the same time the tem ple of the body, the complete domain of manifestation, and the mantra that con structs it in ascending it” (SP 1 p. 186, n. 1).

  4. According to Anantasambhu, each portion of the mantra is characterized by a name, a form, an illumination or color (abha), a pervasion, a location, a duration of pronunciation (m&tra), a path, and a presiding lord (SSV 37). SP 1, Plate 6, “Recitation du Prasada-mantra,” gives a relatively complete depiction of the man

tra portions. One may also compare the three additional diagrams of mantra por tions, based on three different texts, in SP 3, Plates 11-13.

  1. This point was emphasized to me by Sabharatna Sivacarya (depicted in these photographs), who spoke of ascending pronunciation in invocation as the single most difficult ritual procedure in daily worship. The coordination of simultaneous actions requires much concentration and training.

CHAPTER FIVE

RELATIONS OF WORSHIP

  1. This episode is the frame-story of the MataAgaparameivaragama, the teach ings given by ParameSvara to MataAga: see MPA vidya 1.1-34. 2. There are two main types of arghya: common (samanyarghya) and special (yiiefarghiya). Common arghya is used generally for rites directed toward subordi nate deities, while special arghya must be used in worshiping Siva. As Brunner Lachaux puts it, “The difference between these two arghya& resides in their com position, but particularly in the nature of the mantras that one recites over the recipient” (SP 1 p. 138).

  2. KA 5.33-36 gives several alternative sets of substances for preparing arghya, though the group of eight listed here is the most common set.

  3. Nirmalamani quotes Bhojadeva in describing the formation of the cow mudra: “When the fingers are woven together by joining little finger with ring finger, and

Notes to Chapter Five · 179

middle finger with index finger, the cow mudra will resemble the udder of a cow” CKKDP p. 46).

  1. The divyamudra employed in this divyadr$fi is described in ML pp. 19-20, quoted in KKDP pp. 130-31. See the photograph of divyadrsti employed upon enter ing the temple, in ML p. 27.

  2. Attributed to KA by SASS p. 109. Cited in SP 1 p. 101.

  3. The themes may vary in proportion according to the situation of worship. Texts like KKD that deal with atmartha worship in a private shrine generally stress acts of hospitality, while KA and other texts prescribing public (parartha) temple worship give greater emphasis to Siva’s lordship. For a fuller account of these di

vergences between atmartha and parartha, see Brunner, “Atmarthapfija versus Panirthapfija,” pp. 7-12.

  1. See also the lists of twelve, sixteen, and twenty-four upacaras given by KarA, in SP 1, Appendix 7, “Upacara du culte de Siva.”

  2. J. Filliozat and P. Z. Pattabinunin, Parures divines du Sud de Vbide (Pondi chiry: Institut Frangais d’Indologie, 1966), photographically documents the crowns and ornaments used to decorate the IiAgas and images in South Indian temples. 10. IP p. 56, quoted in SP 1 p. 213.

  3. Some texts provide suitable panegyric. In KKD (pp. 121-23), for instance, AghoraSiva quotes the stotra& from RA, KirA, KalA, and MPA. 12. Much of this discussion of nirmalya is based on $AC pp. 111-26. This was repeated in NIlakantha’s Kriyasara (assuming that §AC predates Kriyasara), which Ηόΐέηε Brunner translates and discusses in her article “De la consommation du nirmalya de Siva,” Journal asiatique 256 (1969): 213-63.

  4. Unnamed purana, cited in SAC p. 120.

  5. Readers versed in the Tamil Śaiva bhakti tradition will recognize Canda also as one of the sixty-three Śaiva nayanmars, whose story of devotion to Siva is narrated by Cekkilar in chapter 20 of the Periyapuranam. Eric af Edholm considers more fully the relation of the Tamil nayanmarCanta. and the agama deity Canda in “Canda and the Sacrificial Remnants: A Contribution to Indian Gastrotheology,” Indologica Taurinensia 12 (1984): 75-91.

  6. Nirmalamani points out that candapuja is obligatory for the siddhanta wor shiper, while it is prohibited to followers of other Śaiva schools (KKDP p. 132). The significant role of Canda and the notion that nirmalya is too pure for human con sumption appear to be ritual features distinctive to Śaiva siddhanta. 16. SupA, quoted in SAC p. 113.

  7. ML p. 13, quoted in KKDP p. 27.

  8. ML pp. 9-10, quoted in KKDP p. 64.