cvacchnda tantra

8/7/2019 cvacchnda tantra

1/510

9THE UN IVERSITY OF CH ICAGO

Birth Daceuthor

7Jvv ~br4m-J - ~ t tAU:U ~"Z:wc .U a f 1Title of Dissertation

~ ~. ~I~~ 'P ~~7;;,.~_jDepartment or School Degree

PH.D. 19V~ r-~_~Convocation

Permission is herewith granted to the University of Chicago to make copies of the above tide, at itsdiscretion, upon the request of individuals or institutions and at their expense.

.~~~ Signature of writer

Extensive Quotation or Further Reproduction of This Material by Persons orAgencies Other than the University of Chicago May Not Be Made without the ExpressPermission of the Writer.

Date f i lmed Number of pagesNOTE:


2/510

TH E UN IVERSITY O F CHICA GO

THE SVACCHANDATANTRAMH ISTORY AND STRUCTURE OF A SA lVA SCR IPTURE

V OLUME O NE

A D ISSER TA TIO N SU BM ITTED TOTHE FACULTY OF THE DIV ISION OF THE HUM AN IT IES

IN CAND ICACY FOR THE DEGREE OF

DOCTOR O F PH ILO SO PHY

D EPA RTM EN T O F SO UTH A SIA N LA NG UA GES A ND C IV IL IZA TIO NS

BYW ILLIAM JAM ES ARRAJ

CH ICAGO , ILL INO ISA UG UST . 1988


3/510

ACKNOWLEDGMENTS

Research for this study was supported in part by fellowshipsfrom the American Institute of Indian Studies, Fulbright-Hays, andWolfson College, Oxford. I would like to thank the members of mydissertation committee, David Gitomer and Edward Dimock, andespecially my Sanskrit teachers, J.A.B. van Buitenen t, AlexisSanderson, and my dissertation chairman and adviser, EdwinGerow.

11


4/510

TABLE OF CONTENTS

PageACKNOWLEDGMENTS iiChapter

I. INTRODUCTION 1I .1 HISTORYAND STRUCTUREOFSVACCHANDATANTRAM 1

I. 1.1 Internal and External HistoricalEvidence 1

1.1.2 Methods of Textual Criticism 211.1. 3 Tradition Criticism 281.1.4 Source Criticism 531.1.5 Redaction Criticism 63I . 1. 6 Reconstructed History of the Text .. 69

1.2 THECOMMENTARY"UDDYOTAW'OFK$EMARAJAl:f 78

I. 2.1 A Proponent of Monistic Saivism 78I. 2. 2 Interpretation and Reinterpretation

of Ritual 901.2.3 Commentatorial Techniques 961.2.4 Harmonization and CommentatorialGoals 112

II. SUMMARIES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 124II.1 BOOK1 12511.2 BOOK2 13411.3 BOOK3 141

11 1


5/510

TABLE OF CONTENTS-CONTINUED

Chapter PageII. 4 BOOK4 147

11.4.1 The Initiation Liturgy 147II. 4.2 The Rites of Joining. . . . . . . . . . . .. 15311.4. 3 Concluding Rites and

Consecration . . . . . . . . .. 16211.5 BOOK5 168II. 6 BOOK6 181II.7 BOOK7 190II.8 BOOK8 204II. 9 BOOK 9 21011.10 BOOK 10 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 218

II .10.1 Worlds in the Bottom Half ofthe Egg of Brahma. . .. 218II .10.2 The Ritual of Purification. . . . . . . . .. 22911.10.3 The Worlds in the Top Half ofthe Egg of Brahma. . .. 234

II.I0A The Sarpkhya Worlds 24011.10.5 The Saiva Worlds 274

VOLUM E TW OII .11 BOOK 11 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 298

II .11.1 The Emanation of the Planesof the Universe . . . . . .. 290II .11.2 The Maintenance and Reabsorptionof the Planes 323

I I .12 B O O K 12 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 334II .12.1 The Nature of the Lower

Planes 334

IV


6/510

TABLE OF CONTENTS-CONTINUED

Chapter PageII.12.2 Meditations on the Planes. . . . .. 341

II.13 BOOK 13 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 35711.14 BOOK 14 362II.15 BOOK 15 . . . . . . . . . . . . . . 366

III . TRAN SLAT ION 373!Ii.1 PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373III. 2 BOOK 1 [vss.1-45a, pp .1-40J . . . . . . . . .. 376III. 3 BOOK 5 [vss.1-43a, pp .1-37] 409III.4 BOOK 12 [vss.1-83a, pp.1-35] 435III. 5 BOOK 13 [vss.1-46a, pp. 89-108] 457III.6 BOOK 15 [vss.1-23, pp.125-139] 471

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 482

v


7/510

CHAPTER I

INTRODUCT ION

1.1 HISTORYAND STRUCTURE OF SVACCHANOATANTRAM

I.1.1 Internal and External Historical EvidenceThis dissertation has as its subject the text published under

the name of the SvacchandatantramJ with the commentaryUddyota.(1Jby K~emaraja1;l.l Svacchanda-bhairavah, who forms theritual subject of the SvacchandatantramJ and in his differentforms, its narrator and supreme deity, has served as the basis forthe title of this text. Manuscript lists and catalogues record thesame text under variant names: trSvacchanda-hhairavlJ-tllntramJ /11trSvacchanda-hhairava-maha-tantramJ /11 and 68vacchanda-lalita-hhairava-tantram. "2 However, the "Svatantratantram'" and

1PaI)QitMadhusudan Kaul Shastri, ed., TheSWacchanda=tantra with Commentary by KslJemaraja, 7 vols.,Kashmir Series 0: Texts and studies 31, 38, 44, 48, 51, 53, 56(Bombay: Nirnaya Sagar Press, 1921-1935). (All unmarked textreferences are to this edition by book (pataJam) and page or versenumber. The text in this edition contains 3648.5 verses, almost allslokiiQ.) Now reprinted in two volumes with notes, indices and newintrcduction by Pt. Vrajavallabha Dwivedi, ed., TheSvacchandantantram with Commentary 'Udyota~byK~araJ5carya8 Partmal Sanskrit Series 16 (Delhi: ParimalPublications, 1985)., 2 For Svacchanda-bhairava-tantram, v., for example, RamaSankar Tripathi, lJescriptive Catalogue of SsmsZrit Manuscripts inBaJkwadaLibrary; Ebarat Kala lIbavana library and SamskritMllha- Vidyalaya Library; Banaras Hindu UniverSity Sanskrit Series 6(Banaras: Banaras Hindu University, 1971), mss.no.C4400,p.776; for

1


8/510

2

NSvacchanda-paddhati./l" "noted in manuscript catalogues representindependent works belonging to a separate tradition not based onthe Svacchandatantram.1

In Nepal and northern India, catalogues record numerousmanuscripts of Svacchandatantram written in devanagan: 1arada"and newsri scripts. Moreover, the devanagari and saradamanuscripts appear to contain both the text and the commentaryof Ksernarajah. Presuming a Sdrl1dl1 autograph of Ksernarajah, it isprobable that sarada archetypes underlie the devanagarimanuscripts. The newarimanuscripts transmitted withoutcommentary, however, potentially represent a version of the textdiffering from that utilized by Ksemarajah.

The printed edition in the Kashmir Series of Texts andStudies, based, according to the editors' notice, on three fairlyuniform sarada manuscripts, has been utilized as the textusreceptus for this study. 2 Although the printed text serves as agenerally reliable and workable edition, a twofold critical editionremains a desideratum.f First, an edition should be prepared ofthe text and commentary of Ksernarajah, using additionalSVacchanda-bhairava-maha-tantram, v. Tripathi, mss. no.3865,p.776; for SVacchanda-ia/ita-bhairava-tantram., v. The Nepal-German Manuscript Preservation Project, Kaisar Library,Kathmandu, tantra mss.no.68; cf. BvaccIJanda-/aJita-bhairava-stotrem in NepJaraJakayavirapustakaJayastha-hastaJikhitasamastapustaJranarp Sarpk$1ptasiicipatram"PuratpattvaprakaSanamala 18 (Kathmandu: Joragal)eSa Press, 1963),p.167.

1For information about these, SVacchandatantram" and textswith related names, v. Gopinath Kaviraj, Tantrika S8hitya, HindiSamiti Granthamal! 200 (Lucsnow: Bhargava Bha~l)a Press, 1972),pp.720-721.2 On the mss. used for the printed work, v. Shastri, ed.,SWacchanda=Tantra 1: i-ii.3 I am now preparing an historical-critical study of the text ofSVacchandatantram.


9/510

3

manuscripts, and following the standard principles of textualcriticism. Second, using primarily the neweri manuscripts, oneshould prepare an edition of the text transmitted withoutcommentary .

This second edition can not follow the principles used for thefirst, suitable for works of single author, but rather must followthose procedures elaborated for the anonymous, collectivelyauthored works of ancient Indian literature. For remarks in thecommentary of Ksernarajah and internal literary criticism indicatethat the text of Svacchandatantram had undergone a long andcomplicated process of transmission before his comrnentatorialwork.! In editing such anonymous works, the canons forestablishing the earliest, correct reading must be supplemented bythe systematic collection of significant variants reflectingsuccessive, independent versions. 2

Preliminarily, the distribution of extant manuscripts supportsarguments for a north Indian origin of the text ofSvacchandatantram:1 in agreement with what is otherwise knownabout the geographical popularity of the cult of Bhairavah.Linguistic evidence based on an analysis of manuscripts, however,at present. does not permit an additional specification of theprovenance of this text. For this specification, the non-standard,or better, non-Paninian Sanskrit, designated by the later traditionas aiSa.fJ:1.e . the usage of i S a l : t , the lord, in the corpus of extant

1V., esp., his remarks inbk.14, p.120. There K~marajal)states that he has used old manuscripts to remedy the corruptreadings and interpolations found in the text.2 For a summary discussion of the differences between the twotypes of editions, v, Oskar von Hintiber, -Remarks on the Problemsof Textual Criticism in Editing Anonymous Sanskrit Literature." inProceedings of the First Symposium of Nepali and GermanSanskritists 1978' (Kathmandu: Institute of Sanskrit Studies,Tribhuvan University. DangNepal Research Centre, 1980), pp.28-40.


10/510

4texts would first have to be thoroughly collected and analyzed. 1Such work might enable the isolating of distinctive features of a"tantric Sanskrit, .. which, at least by frequency, might bedifferentiated from those irregularities shared with other forms ofnon-standard usage. variously designated as Buddhist hybrid, epic,puranic, or vernacular Sanskrit. 2 Even after such a collection offorms, however, any subsequent geographical localizing of theprovenance of Svacchandatantram on the basis of features shared,for instance. with a middle Indie dialect. would be unlikely, due tothe difficult in localizing the known middle Indic dialectsthemselves. Currently, the manuscript evidence merely confirmswhat the substantive history of the content of Saiva texts of thisgenus show: namely. an assimilation and reworking by morecultivated milieus. And the progressive excision of deviant formsby successive generations of copyists or commentators merelyreflects this process.S In sum, therefore, linguistic evidencecorroborates the other evidence pointing to an origin in a non-Pat;linian, non-sm5rta brahmanical dominated tradition.

Besides manuscripts and linguistic analysis, information foundin other related texts offers another potential source of evidenceabout the origin and history of Svacchanda tan tram. In the self-representation of these expressly &iva traditions, works such asSvacchandatantram constitute the scriptural corpus of revelation

1In his commentary, K~marajal) regularly notes these non-standard forms, and glosses them with the correct usage.2 On the need for this work, v. Albrecht Wezler, review of The$atsahasra Samhft4 by J.A. Schotermann. ZeJtschrfft derJJeutschen Morgenliindischen f1esellschaft 135, 1 (1985): 143-146.3 A tension exists between a conservative respect for scripture,and a need. to adapt it either linguistically or ideologically to supportaltered circumstances. As an example of an ingenious solution thatmaintains the literal wording but with a trans.ormed sense, v. infra1.2.3.


11/510

5

that fufills the function in the Saiva community that Vedic textsfulfill in the explicitly brahmanical smsrts traditions. And, atsome time during the growth and transmission of this Saivatradition, as it developed an independent identity, it appliedclassificatory schemata to its scriptures. Saiva circles subsequentlyused and transmitted these schemata in several different ways:they either inserted these schemata into the body of existing texts,used them to elaborate and to incorporate "new" texts, ortransmitted them concurrently in oral form. 1

One popular schema organized the scriptures in threebranches according to their revealing deity, doctrine, and numberof texts: ten dualistic scriptures revealed by SivaJ:t, eighteendualistic-non-dualistic revealed by Rudrah, and sixty-four non-dualistic revealed by Bhairavah. In later presentations of thispattern, the Svacchandatantram.. revealed by Bhairavah,naturally, falls among the non-dualistic group of sixty-fourscriptures.Z The historical value of the doctrinal classification,specifically, the labeling of Svacchandatantram as a non-dualscripture, becomes immediately suspect, however, by the veryfact that Ksemarajah explicity states that he composed hiscommentary to refute dualistic commentaries on

1For information on the Classificatory schemata used intantrtc texts, v. Teun Goudrlaan and SanjUltta Gupta, Hindu Tantrlcand Sa7rtaLiterature (Wiesbaden: Otto Harrassowitz, 1981), pp.1-31.2 V. K.C. Pandey, A.bhinavagupta. An Historical andPhJlosopbical study; 2d ed. The Chowkhamba Sanskrit Studies 1(Varanasi: Chowkhamba, 1963): 138ft; Raniero Gno11.trans . Luredelle SacreScritture (Tl11'1tralo.ka)di A.bbintJvagupta ..ClassicidelleReligioni, Sezione prima (Torino: Unione Tipografico-EditriceTorinese,1972), p.70, n.17. The best summary ot the traditional sources isgiven in the commentary by Jayarathal:,l on TsntraJoJrsQ.. in Par::u;litMadhusudan Kaul Shastri, ed., The Tantr5lo.ka of A.bhinava Buptalwith commentsry- by R~jana.kaJaytJr6tbal Kashmir Series of Textsand Studies, 23 (Allahabad: The Indian Press, 1918), 1, vs.18: 35ft.


12/510

6Svacchandatantram.1 And as will be shown, Svacchandatantramexhibits affinities with a wide spectrum of Saiva and non-Saivatexts, and lacks an explicit dualistic, qualified non-dualistic, ornon-dualistic credo. Clearly, therefore, for Svt!lcchandaU1ntram,this particular schema represents an ex post facto superimpositionreflective only of the later transmission of the text and not thecircumstances of its early composition.

Similarly, the number sixty-four associated with the non-dualistic scriptures revealed by Bhairavah, which is then dividedinto eight subgroups, appears to be a factitious and retrospectiveconstruction. It may have been based in part on traditionalnumerical associations, such as the eight bhairavah, expanded bycorrelation with the sixty-four yoginyah associated withBhairavah, 2 Though Saiva tradition expressly associatedSvacchandatantram with a subset of eight specifically Svacchanda-bhairava tantrdpi in this list of sixty-four, Ksemarajah in hiscommentary, which is full of citations and references, does notextensively quote these works.:S This pattern argues that this

1 V. the concluding verses to his commentary, in bk..1S, p.146.2 On the sixty-four yoginyah, v. Marie-Therese de Mallmann,Les enseignements iconographiques de J~ni-Purana, Annales duMusee G; -tmet 67 (Paris: Presses Universltaires de France, 19(3):chap.8, esp., p.176, n.9, referring to the sixty-four bhairavah. Asshe notes, only the eight main bhairaval) appear to be well defined.For a list and descrtptton of the sixty-tour bnairavan, v. GopinathRao, EJements of Hindu lconography2, 1 (Madras: The Law PrintingHouse, 1916): 180tt.3 In his commentary on Tantraloka,(J" Jayarathal), in Shastri,M.K., ed., Tantr51okaQ, 1: 42. quotes the Srlka.(1thldescribing thisfirst group of eight as an expansion of Svacchanda-bhairvaI:t:trsvacchandobl1aira vaSca.r;1{1a,(Jrrodha unmattabhairava,(J / asitangomenoccnusme kapaliiast8thaiva ce // ete svacchandarupastu.bahuri1~a bl1a.sitap/ 11 1 v. also bk.l, p.l0, where K~marajal)recognizes this tradition. by quoting the sarvavira~l, associatingSvacchandatantram, with the first four texts in this group, theCaJ")f/a.fJ,rodha.(1, Unmattabl1airavaQ, and others. In his


13/510

7association derived more from later. systematic considerations thanintrinsic historical and material connections. If additional texts arediscovered in manuscript and studied, however, perhaps moresubstantive reasons for their affiliation by these related patternswill come to light. Nevertheless, just as it lacks a non-dualisticcredo, so Svacchandatantram itself does not refer to this schemaof sixty-four Bhairava 0:- eight Svacchanda-bhairava tantrt3lJi1

Another widespread classificatory schema divides the Saivascriptures into various streams (srott3J'pSJ)of scriptural revelationwhich emanate from the faces of Sival;t in his various forms. 2Thus, in addition to assigning Svacchandatantram to the non-dualistic branch of sixty-four Bhairava scriptures, Saiva tradition,as followed by Ksemarajah, further categorizes Svacchandatantramin this related schema as belonging to the dak,SJ".paJ. e., right or

commentary, however, K~arajal;l quotes only a Rurusarpgraha(1.bk.l0, p.402, (which may by identical to the Satva Rauravagamap)on the names of rudrah, and more significantly, anUcchu~abhajrava(1. bk.7, p.315, bk.l0, p.382, giving the samequote twice, whtch states that only one pure reality exists.

1In bk.10, pp.S11ff. the text does mention the ten dualisttc,and the eighteen, dualistic-non-dualistic texts, representing thetwenty-eight traditional Saiva scriptures. For a rejection of thehistorical significance of this drvision, into dualtsttc, and non-dualist-dualistic, v. Helene Brunner-Lachaux, "On the Classificationof SivaTantras," in Summaries of Paper~ International Association ofSanskrit Studies, Fifth World Sanskrit Conference (Varanasi:Rashtriya Sanskrit Sansthan, 1981), pp.128-129.2 In the basic form of this model, the Saiva scriptures areemitted from the five faces of Sadasival;l; v. J for a collection of theseclass1fications,M. Arunachalam. Thesajvagama~ Peeps into TamilCulture.6, studies in the Agamas 1 (Madras: Gandhi Vidyalayam,1983), pp.20ff, 99ff. This schema represents the Saiva counterpart ofthe Vedic model in which the different Vedah and other forms ofknowledge are emitted from the four faces of Brahma. For ane~ample, v. Kr~l)aSailkaraSastri, et al., eds.SrimI'Jdbh5gavl1tapuraJ;1.:Jm3 (Varanasi: Samsara Press, 1965), bk.12, vss.34-48: 454ff.


14/510

8

southern, stream of scriptures.! Using an adage evidently popularin his own tradition, Ksemarajah ranks the various streams ofscripture in the order, Saivasiddhantin, vama.1 i. e., left ornothern, da.k~jna.1and Kaula.2 Svacchandatantram itself does notallude to this structure, and Ksemarajah explains it more fullyonly in his commentary on the Netratantram.1 which describes thepresiding deity of each stream: Sadasiva}:l for the Saivasiddhantin,Tumburuh for the left, Bhairavah for the right, and Bhairavahonce again for the Kaula. 3

The historical value of this schema is weakened, however,because only a single text survives that can be putativelyattributed to the left stream.f Just as the preceding schemata, soalso this model apparently reflects more the conditions of the latertransmission of texts in the various Salva schools than the actualcircumstances of their composition. In particular, by dropping thevama.1 this schema becomes just another expression of the lateropposition between the dualistic Saiva-siddhantin and non-dualisticschools of Saivism represented by Ksemarajah. As will be examinedin full, much of Svacchandatantram.1 however, appears to predate

1V. bk.1, p.8.2 V. hk.ll, p.55:'" Saivadvamam tu dak$Jnam/da.k~natparata~ .kauJamkauJatparataram nahi. -Saiva, asK~marajal) explains, refers here to the Saiva-siddhantin texts.3 V. Pa!)


15/510

9the split between these dualistic and non-dualistic schools. And, aswill be discussed, the Kaula stream, though closely related to thedak~.inaas in this schema, only appears as an independenttradition considerably after the formation of the other scripturalbranches. 1

The Svacchandatantra_m~ in characterizing itself, uses twosimpler schemata instead of referring to any of the precedingpattern of streams found in many scriptures and commentators.First, in the opening book, the text introduces the commonplacethat this particular tantram represents a condensation of a muchmore extensive and primordial "Ur-tentrem. "The revelation ofthis particular text, in its accessible and abbreviated form,constitutes a particular act of grace. 2 Behind the theological clichelies a possible historical kernel, for in his commentary,Ksemarajah refers to larger texts specified only as "brbet-tantra{1J:....which apparently, in extenso, supplement the datagiven by Svacchandatantram."5 Perhaps this citation warrants,with caution, the supposition that larger texts of this particulargenre were extant at his time, and were to be taken as thecontext for explicating parts of Svacchandatantram. But the vaguereference gives pause, being inconsistent with Ksernaraja's usualcommentatorial practice of citing texts by title. 4

1V. infra II.IS, the summary of bk.lS.2 V. bk.l, pp.9-10, esp. vss.,6-7. Cf. Shastr!, M.K., ed.,jantraJoka!J~ 36, vss.l-l0, for a similar version of the condensation

of revelation based, according to Abhinavaguptal:1, on theSiddhayogeSvarimatam. For a Pur8I:licversion of this idea, v. LudoRocher, ThePur5Qa~ A History of Indian Uterature 2, :3(Wiesbaden: Otto Harrassowitz, 1986), pp.46-47.3 V., for example, bk..l0, p.223, referring to a Brhattantram~and hk.l0, p.267, p.440, referring to unspecified Brhattantrs.{2l:4 This may also represent K~araja's shorthand for a generalcross-reference to other Saiva texts discussing the same topics: note


16/510


17/510

11

designate primary Saiva scriptures. it is often noted that theirusage, at least in part, differs regionally: agamap is the normalterm in the south, tantram in the north. 1Athough thisdifferentiation may correctly reflect the later history of the Saivaschools. in which the dualistic Saiva-siddhantin based on thetwenty-eight agamap became dominant in the south, and the non-dualistic in Kasmir, it in no way reflects the religious conditions ofearlier perioda.Z In addition, Svacchandatantram~ like the otheragama.(18uses both terms in references interchangeably.SFurthermore, even if this distinction were accepted, the use of theterm, tentrsm, would merely argue for an unspecified northIndian origin for Svacchandatantram.

In its opening dialogue, the Svacchandatantram offersanother self-characterization as the compendium of the greatscripture {mahatantramj that has four seats or bases(catu~pi.tham).4 This refers 1:0 another classificatory system,which. as explained by Ksemarajah quoting the SarvavIra.{J1groupsscriptures by one of four topics that they may predominantlytreat, or for which they may serve as the basis: mantrap orformula; vidy~ or (female) formula; mt!J{1aJamr diagram; and

1V., for example, Gonda, Medieval ReligiousLiterature inSans.kri~ p.2, n.5.2 On this misconception, v., for example, AlexisSanderson,review of Mrgendragamal by MichelHulin, Bulletin of the School ofOriental and African Studies 46 pt.1 (1983): 161.3 V. bk.4~ p.13~ for example, where Svacchandatantram refersto other agamaQl or bk.4, p.l0l, to other tantrani. Cf. on this topicHelene Brunner-Lachaux, ""LesMembres de Siva,'" AsfatischenStudien 40, 2 (1986): 89, n.l.4 Bt..l, p.l0.


18/510


19/510

13

serve as the authority and basis for the later datable andidentifiable Saiva sectarian authors. And by default, their datesoften serve as the terminus ad quem fer the Saiva scriptures.Accordingly, in the case of Svacchandatantram this terminus isthe tenth century and Abhinavaguptah, and probably prior to thehistorical figures in his sampradaya!1 or teaching lineage.1Supplementary evidence must then help decide a suitable terminusoh quo.

At the outset. it should not be assumed that the compositionof anonymous scriptures necessarily terminates at the time whenmore cultivated and historically self-conscious milieus begintheoretically expounding and elaborating them in works of aqualitatively different nature. A continuing parallel process offabrication, recomposition, and transmission of scriptures, if not inthe same, at least in other circles, should be presumed. And thisprocess argues for prudence in pushing back the terminus ab quofor anonymous texts by a set amount determined by reasoningabout the time needed for an evolution, as it were, of ideas to amore complex level, or rather, assmilation by a more sophisticatedmilieu. Nonetheless, in the case of SvacchtJndt1tantram~ identicalreasoning argues for a considerable prehistory. if not antiquity. inwhich the text could acquire prestige sufficient to motivatenumerous commentaries, and even colonization by competingsects.

In the absence of direct evidence such as citations (prior tothose of the Saiva authors discussed previously), or inscriptions,dating a text such as Svacchandatantram depends on the dating,or periodization, of the genus of texts to which it is commonly

1The first historical figure in the non-dual lineage, for whomscriptures such as SVacc.handatantram are the supposed scripturalbasis is Vasuguptah, who is the author -transmitter of the sourcetext of this school. the $iVt!15iltrm:u; dated to the beginning of theninth century. V. Pandey, Abhjnavagupta~ p.1S4.


20/510

14

attributed. Inscriptional evidence for other Saiva tantra,pi fallswithin the span proposed for the composition ofSvacchandatantram: in the eighth century before the historicalfigures in the tradition of Ksemarajah. 1A century earlier, theexistence of Saiva scriptures, if not the agama.{1oflater traditioncan be inferred from the South Indian inscription of the Pallavaruler Narasimhavarman II, in which he refers to the Saiva-siddhanta1;l.2 Certainly, the organized cult of Siva1;l, and, inparticular, Saiva sects, such as the Pasupatah, whose practicesand beliefs have strongly influenced works likeSvacchandatantram, predate this time by at least anothercentury.J And though from other evidence of Saiva religiousactivity in earlier periods the concomitant existence of oral or evenwritten traditions may be inferred, this evidence cannot warrantretroactive assumption of a corpus of Saiva scriptures at thisearlier period. More reasonably, their appearance can not be

1For the Cambodian inscriptions that attest the existence ofSalva tantralJifrom the begi!lnlng of the ninth century, v. AdhlrChakravarti, "New Light on Saiva Tantrika Texts Known in AncientCambodia," Journal 01 the Asiatic Society (Bengal)15, 1-4 (1973): 1-10.2 E. Hultzsch, ed. and trans., South Indian Inscrtpttons: 1 inArchmological Survey of India n.s.3 (1890; reprint, Varanasi, Delhi:Indological BookHouse, 1972): 12, 13-14, vS.S. On the dates ofNarasirphavarman, c.690-91- c.728-29, v. T.V. Mahalingam,Kancipuram in Early South Indian History (Bombay: Asia PublishingHouse, 1969), pp.109ff.:5 The Pasupatal) can be dated in the most conservative viewto the preceding two centuries. For a reexamination of the debate ontheir date and origin, v. V.C. Srivastava, "The Antiquity of thePasupata Sect, ., in K. C . Chattopadhyaya Memorial votume;Department of Ancient History, Culture & Arch.'iEOlogyAllahabad:Allahabad University, 1975). pp.109-125, especially p.111: "Therecannot be any doubt that the Pasupata sect was a well-establishedsystem in India in the early mediaeval period (7th-12th cent.A.D.)."


21/510

15

placed before the efflorescence of Saivism produced by thepatronage accorded by dominant Saiva dynasties in the seventhand eighth century. 1

To complement historical evidence such as inscriptions,anonymous works of Sanskrit literature such asSvacchandatantram can be analyzed for isolated internal clues totheir provenance that potentially can be dated or located in anexternal context. Identifiable and specific historical referencesnaturally merit first consideration. Unfortunately,Svacchandatantram~ with its predominantly ritual andmeditational content, does not refer to specific historical figures orevents, nor does it associate its deity with a definite pilgrimagecenter (p.t . tham). 2

Comparable to references to historical events, persons, andplaces, literal citations from other known and datable works canprovide reliable information about the anonymous text in whichthey are embedded. As might be expected. however. in ananonymous and collective composition that purports to be revealedscripture, paraphrase is the rule and quotation the exception. 3

1For literature, v. Fran~ois Gros, "Towards Reading theTevaram," introduction to Tevaram. Hymnes Sivaltes du PaysTamoul T.V. GopalIyer and Francois Gros. eds. (Pondiehery:Institut Irancais d'indologie, 1984), esp., pp.l-liii.

2 cr., for an example of such a reference used in identifyingthe provenance of a text, J.A. Schotermann, ed., The Yonitantra(NewDelhi: Manohar Publications, 1980), pp.3ff.3 Cf. in reference to the Pur8J:llctexts, Ludwik Sternbach, TheManava-dharmasastra I-II~ and the Bhavi~yaPuraQa(Varanasi: AllIndia Kashi Raj Trust, 1974). He notes (p.4) how exceptional it isthat the BhaVJ~yapurQamecopies verbatim the smrtitexts.


22/510

16

Next in importance as internal indicators come iconographicand architectural details. 1Many agam5{1have extensive sectionson temple construction, installation of images, and similarsubjects, since they functioned in large part as guidebooks fortemple priests. While sharing many other topics with these texts,Svacchandatantram~ however, treats these subjects only inpassmg.

First, for iconographic evidence, though not matching exactlythe meditational form of Svacchende-bhalravah described inSvacchandatantram~ a bronze figure of Svacchanda-bhairavah inunion with Bhairavi has been found at Kangra and dated to thetwelfth century. 2 This iconographic evidence corroborates otherevidence locating the worship of Bhairavah as the supreme form ofSivaJ:!in north India, and specifically the northwest Himalayanregion.3 The composition of Svacchandatantram as the principalscripture associated with the worship of Svacchanda-bhairevahmight then be inferred to have occurred in this region. This cannot be concluded with certainty, however, for evidence attests the

1Cf. Teun Goudriaan, KaSyapa- : s - Book of Wisdom (Kasyapa-.Jiianak5.(1(1a.tJ).A Ritual Handboo.kof the Vai./rhanasas(Utrecht)Disputationes Rheno-Trajectinae 10 (The Hague: Mouton & ce., 1965)p.10, n.13: "The best way for finding some chronological clues is toconsider the arch1tectural and iconographic date offered by thetext."2 B.Ch. Chhabra, Svachchhanda-Bhairavi Bronze Image fromKangra, Journal of Indian History42, 1 (April 1964): 145-148. Cf. thedescnptton of SVacchanda's meditatlonal form in bk.2, pp.51ff.3 V. Brunner-Lachaux, "Les Membres de Siva," p.114, " ... celledu Bhairava a cinque tetes qui est au centre du culte du sivaismedu Nord." V. also Alexis Sanderson, "S&vism in Kashmir," in TheEncyclopedia of Religion 13 (New York: Macmillan, 1987): 17: "... inKashmir '" the principal Saiva cult was ... the worship ofSvacchandabhairava and his consort AghoreSvari ......


23/510


24/510

18

date or provenance of the text.1 Moreover, rather than withexternal structures, such as a Bhairava shrine or temple, here, asin its other statements about monuments or iconography,Svacchandatantram may be more concerned with internal modesof worship and meditation. Accordingly, later in the text, afterprescribing construction of a ritual diagram {ma.pt;lalam}withcolored powders, Svacchandatantram repeats its instructions toperform the same worship at the ctdoors. " of the diagram. 2

Svacchandatantram contains another potential indicator of itsorigin when it describes the construction of amulets written onbirch bark (bbt:lrjapatraD!j.'5 Since birch bark was used as awriting material primarily in the northwest Himalayan region,Schrader, for example, argued that the Paficaratrin AhirhudhnyaSaJ71hjta~which describes the same kind of amulets, had beencomposed in Kasmir. 4 But as reported by Buhler, the use of birchbark for amulets and letters had a much wider distribution. 5Therefore, though providing another clue as to the provenance of

1For an exhaustive list of the distribution of this pattern in314 temples, v. von Stietencron, Ganga und Yamun~ pp.16ff.2 Bk.5, p.23.~ Bk.9, pp.S4ff.4 F. otto Schrader, Introduction to the Paiicaratra and theAhirbudhnya SaqJhit8 (1916.Reprint. Madras: The Adyar Ubraryand Research Centre, 1973), p.11!.5 Johann GeorgBUhler, Indian Paleography;Appendix to TheIndian Antiquary33 (1904t ed. John Fleet, p.93: -According to astatement made to me by Bhall Daji, birch bark mss. occur also inOrissa, and amulets, written on Bhllrja. are still used throughout allthe Aryan districts of India."


25/510

19

Svacchandatantram~ the use of birch bark does not prove aKasmiri or northwest Himalayan origin.1

Furthermore, even if this information pointed unmistakablyto a Kasmiri origin, it would not provide a certain starting pointfor the evaluation of Svacchandatantram as a whole. For in ananonymous work of this size, presumably transmitted in differentregions over a long period, a detail permits the dating or locatingof only a particular topical section of the work; its importance forthe remaining sections must be corroborated by additionalevidence.

Thus, isolated evidence for date and provenance found inspecific sections of an anonymous and extensive scripture likeSvacchandatantram must be evaluated more conservatively thanthat found in the unified work of a single, known author. In thesame way, judging manuscript readings or interpreting singleverses and larger sections in a composite work, requires anawareness of the intention of a lineage of distinct authors, and anunderstanding of the text's unique history of composition andtransmission.First, the text of Svacchandatantram has, as its basiccompositional or organizing structure, a division into books(pata1am). Just like the choice of tantram to designate the workas a whole, the choice of the term pa!alam~ used throughout thegenre to designate sections of a text, does not appear significant. 2

1V. also Sanjukta Gupta's introduction to the Lak$Il1iTantra~in Sanjukta Gupta, trans., Lak$Il1i Tantra. A Pancaratra Text(Leiden: E.J. Brill, 1972), p.xxt, against arguments for provenancebased on the used of birchbark amulets.. 2 On the historical significance of the term ...pa talam inVedicntual literature, v. Asko Parpola, trans. J TheSrautasiitras ofL5ty5yana and lJrabyayalJa and their cammenteries;Commentationes Humanarum Litterarum, Societas ScientiarumFennica, 42, 2 (Helsinki, 1965),pp.43ff; for references to theetymology and used of the term tantram in medicalllterature, v.


26/510

20

For the most part, these books coincide with the introduction ofmajor new topics by the other main organizational structure ofthe text, the frame dialogue between Bhairavah, who in this workassumes the role of promulgator, and his consort goddess, whoplays the interlocutor. In addition, the goddess lists, in the firstbook, the major topics to be covered in the work. 1Manyanonymous, encyclopaedic works like puralJi have this short tableof contents called a suci fJ or anukramalJJKa.2

The actual number of topics treated in the text exceeds notonly the number of books but also those topics listed separately inthe opening anukra.'11a{1ika. Thus, in his commentary on thisinitial iH1ukramt?'J1JlK~ Ksemarajah attempts to demonstrate thatthe listed topics include and, thereby, imply the remaining topics.Similarly, in the fourth book, before discussing a series ofpreviously unenumerated topics, the text supplies anotheranukramalJJKa.:5 While the anukramaJ?1Ka omits many topics,almost every topic is marked by the framing dialogue betweenBhairavah and the goddess. And where even this dialogue islacking, Ksemarajah supplies it, by marking the beginning and endof added topics. 4 He confirms, thereby, the perception that theReinhold F.G. Mtiller, "Worterheft zu einigen Ausdrticken indischerMedizin," Mitteilungen des Instituts filr Orientforschung, DeutscheAkademie der Wissenschaften zu Berlin Institut ftir Orientforschung8 , 1 ( 1 9 6 1 ) : 1 0 2 .

1Bt..1, pp.U-12.2 On this organizational device in the puraJ:JJ: v. GiorgioBonazzoli, "Schemes in the Puranas," Pura}Ja 24, 1 (January 1982):149ff.: 5 Bk.4, pp.142-143.4 Cf., for example, hk.5, p.l0, where K~marajal:l rationalizesthe deficient dialogue introducing a topic without an appropriatequestion as due to the supernumerary grace of the lord:

Ir anuJighrk~orbhagavtyaasayaparit~yapraSnjt8mapjpadadidik~J'!1 niriipayi~yan .... R


27/510

21

dialogue frame forms the primary and even indispensable structureof the text.

Although these internal organizing structures, do not, inthemselves, further specify the provenance and date of the text ofSvacchandatantram3 they do clearly identify its genre, and suggestthe kind of historical and structural analysis that would furtherunravel its history. Along with the non-Paninian Sanskrit andsimple verses, the lack of a strict internally coherent organizationmarks the text as the collective and cumulative product of an oraland vernacular rather than literary tradition. In particular, thedialogue frame has permitted associated topics to be easily added tothe text. While devices like the anukrama1Jika indicate the work ofmore careful redactors operating with traditional patterns,inconsistencies throughout the work point to the absence of a singleauthor or editor.1.1.2 Methods of Textual Criticism

Works carefully composed by known authors in documentedcircumstances, naturally permit a systematic textual andcontextual interpretation not available for collective andanonymous texts. in order to compensate for this absence oftextual cohesion and supplementary biographical and sectariandata, anonymous compilations must be critically analyzed bothinternally and comparatively. Once this analysis has identifiedindividual textual components and thus furnished a context forinterpretation, the particular process of their historical compilationcan be studied to provide insight into the systematic significance ofthe text.

Provisionally, three interrelated stages of this analysis can bedistinguished: first, literary criticism, or the hypotheticalidentification of written documents used to compile the text;second. tradition criticism, or the examination of the traditionsunderlying the text and its sources; and third, redaction criticism,


28/510

22

or the study of the processes by which documentary sources andtraditions have been woven together to produce successive versionsof the extant text. This analysis may assume that in constructinga text like Svacchandatantram~ compilers not only have drawn onwritten documents, or sources, but also have used varioustradition-complexes or residual collections of material of varyingages which had been handed down partly orally, and partly inwritten form. These traditions that form the context, i.e., thedirect or mediated background accompanying the constructing andreworking of the text, may be identified as its underlying strata.!

Redaction criticism may assume, as well, that foranonymous Sanskrit sectarian literature the composition,transmission, and recitation of texts do not fall into distinct phasesbut overlap and intermingle. 2 Thus, the extant text ofSvacchandatantram resulted not only from successive but alsofrom interdependent activities of compilation, redaction,interpolation, revision or normalizing correction, emendation, andcommentary. Accordingly, in place of a single author or editor,stands a Saiva tradition, composed of generations of teachers,ascetics, officiating ritualists, meditative practitioners,commentators, initiated patrons .. and scribes.

This literary, tradition, and redaction criticism serve asprimary methods for the internal analysis of a single text. Forcomparative analysis, many model studies already exist, whichprovide examples of the methods to be employed when examining

1For this terminology, adopted here with modifications, v.R.E. Oements, "Pentateuchal Problems," in Tradition andInterpreUltionl ed. G.W. Anderson (Oxford:Clarendon Press, 1979),esp., pp.96-97.2 On the overlapping of these processes, and the intermixtureof oral and written tradition in the puraQJ: cf. GiorgioBonazzoli,"Composition, Transmission and Recitation of the Puranas," Purana25, 2 (July 1983): 254-280, and esp. 279. . .


29/510

23

multiples versions of, for example, cosmological sections, found inmany texts. 1By reconstructing the historical context andideological basis for the agreements and differences among versions,this comparative criticism of anonyomous Sanskrit literature hasbeen able to generate, at least in outline, doctrinal or sectarianhistory. 2

Material from Svacchandatantram has already beenextensively cited in the study of certain Saiva rituals.:3 Evenwithout a specific interest in applying the comparative method forhistorical purposes, cross-referencing and accumulation of commonmaterial remain prerequisites for understanding any particularSaiva scripture, especially in its ritual descriptions that are oftenat once, technical, stereotyped, laconic, and obscure. And thetraditional commentators themselves, such as Ksemarajah, havepursued and relied on this type of comparative explanation.

Both the non-systematic explicative use of comparative data,and the methods of systematic historical and comparative analysiscan be applied, without alteration, to comparable sections of the

1For the programmatic account of this comparative historicalmethod, v. Paul Hacker, "ZUI:'Methode der geschichtlichenErforschung der anonymen Sanskritliteratur des Hinduismus," inPaul Hacker Kleine scanaea , hrsg. Lambert SChm1thausen,Glasenapp-Stiftung 15 (Wiesbaden: Franz Steiner, 1978), pp.8-17. Fora paradigmatic application, v. Klaus Rtiping, Amrtementhene undKllrma-Avatara~ Schriftenreihe des SUdasien-Instituts der trnrversitatHeidelberg (Wiesbaden: Otto Harrassowitz, 1970).

2 Hacker,"Zur Methode der geschichtlichen Erforschung deranonymen Sanskritliteratur des Hinduismus," p.14: "Und da direktehistoriscl1e Zeugnisse meist rehlen~ ist die Textgescl1ic.hte bee.,allgemeiner gesprocl1en~ die Methode der Vergleichs derMehrfachtiberlieferungen, oft ~YJ8ardas einzige wissenschaftlicheEr.kenntnismittel fiJr die geschichtlichen Vorgiinge. ".: 5 V. Helene Brunner-Lachaux. trans., SomaambhupaddhatJ;trotsieme partie, Publications de l'institut francais d'indologie .25, 3(Pondichery: lnstitut trancats d'indologie, 1977).


30/510


31/510


32/510

26

Sanskrit literature. The first rule used in comparing multipleversions of a tale, or the like, is that the shorter version is theolder.1The correlative of this principle assumes the continuousaccretion of anonymous works of Sanskrit literature. Despite cogentarguments against the universal application of this principle, in theabsence of counter-evidence, which is obviously often difficult tolocate, criticism must exclude questions of excision and subtractionand proceed with this working rule. 2 And, as will be shown, thetext of Svacchandatantram would appear to confirm this rule, byshowing repeated evidence of interpolation and of the extension ofearlier systems. In some cases, however, the text qualifiesdescriptions as summary, and refers along with Ksernarajah tomore comprehensive accounts in other books of Svacchandatantramand in other texts.3 This suggests that compilers and redactorscould freely condense and reorganize source material, and cautionsagainst the uncritical use of this nonetheless basic principle.

A second rule, or a variation on the first, indicates that thesimpler version precedes the more complex. 4 In

1This principle especially forms the basis of the critical editionof the Mah5bharatam; v. Mahesh M. Mehta, TheMahabharaia. Astudy of the Critical Edition, Bharatiya Vidya Series 34 (Bombay:Bharatiya Vidya Bhavan, 1976), pp.37ff.2 On the possibility of excision, v. Mehta, TheMah5bharata~pp.44ff.: 3 For the references to other texts, v. supra the discussion onthe BrhattantrQi; for typical cross-references, v., for example,

bk.5, p.25, vs.38: tr S8J71puJayet ... pilrvoktena vidhanena .... IFFor an example of an alternative procedure marked by the text asan abbreviated description, trsamk$epeQa"v. bk.4, vs.404b, p.256.4 V. on the uncritical use of this principle, Rocher, ThePura.lJa~p.102: ..... Iagree with the more cautious and open-minded approach, that 'when a series from simple to complex isconsidered providing the chronological framework. a counterargument that with the passage of time, the same complex situationwould get simplified also requires to be carefuly considered.'"


33/510

27Svacchandatantram~ while many complicated rituals appear tohave been built up out of simple rites, other rites appear to besecondary simplifications of more complicated procedures.1 In somecases, technical terms forming part of a prior complex systemreoccur in simpler contexts. Here only careful examination of allthe evidence about the relations of the traditions, and mostimportantly, the possibility that these terms might have been "de-systematized" and "re-contextualized," can correct a prematureattribution of priority according to the rule that simpler is earlier.

This caution reflects a third working rule that considers asmooth, unitary, and homogeneous treatment of a subject to belater than a traditional, digressive, less logically composedversion. 2 This principle, in turn, relies on the evidence ofredaction criticism, which looks for specific signs of textualreworking. Most often, anonymous texts develop through limitedoperations of compiling, redacting, and localized retouche Athorough recasting characterizes the work of a later, educatedmilieu and the systematizing of individual authors.

This principle, in turn, reflects a fourth working rule, that aversion closer to a Panil,lian norm is later.:3 Corresponding to this

1For an examination, in srauta ritual texts, of theconstruction of complex rituals through the recursive embedding ofsmaller rites, v. Frits Staal, "Ritual Syntax," in Sanskrit and IndianStudies~Essays in Honour of lJanie/H.H. Ingall~ ed. M. Nagatomi, etal. (Dordrecht: E. Reidel Publishing Company, 1980), pp.119-142.2 On this principles in relation to the Upanisads, v. Joachim

Friedrich Sprockhoff, SaI!U1y8sa.Quellenstudiensur As.k~ imHinduismu~ Abhandlungen fUr die Kunde des Morgenlandes 42,1(Wiesbaden: Franz Steiner, 1976). pp.20-21.:3 V. Friedrich otto Schrader, "The Kashmir Recension of theBhagavadgita," in Friedrich Otto Schrader Kleine Scbriften~ hrsg.Joachim Friedrich Sprockhoff, Glasenapp-Stiftung 19 (Wiesbaden:Franz steiner, 1983). p.182: "Experience in the field of Indianphilology has yielded the rule that of two recensions of the samework. the longer one is more often than not also the later one. And


34/510

28

presumed typical pattern of assimilation by more educated, orbrahmanical circles, comes a fifth principle, that the pronouncedsectarian version is normally the later.1 The same sectarianmotives responsible for creating a particular recension of a work,have undoubtedly stimulated sects to construct sectarian variantsof particular rituals, myths, or entire genera of Iiterature.Z Whilesectarian rivalry encourages imitation and thereby innovation,sectarian identity and self-consciousness simultaneously contributesto a reverential and conservative attitude toward their ownscriptures or their particular recensions. This conservatism, inturn, moulds the redactional process, favoring accumulation,accommodation, and adjustment at the price of repetition,contradiction, and disorder.1.1. 3 Tradition Criticism

Working with such rules, internal literary and redactioncriticism can attempt to reconstruct the historical development ofthe Svacchandatantram by carefully analyzing its structure.Before such an analysis, however, a preliminary enumeration ofthe traditions that have influenced the text would be useful. Suchequally, the approach to pa1'.lini'sstandard, in pre-classical works, asagainst nonconformance with tt, Is prima facie a sign of lateness."

1Cf. Friedrich otto Schrader, "Das $a~titantra," 1914, reprintin Friedrich Otto Schrader Kleine Schriften~ nrsg. Joachim FriedrichSprockhoff, Glasenapp-Stiftung 19 (Wiesbaden: Franz Steiner, 1983),p,J10: Ir ... Ahirbudhnya-Samhit~ ein sus 60 Adhyyas .bestehendesSlokawer.k~dessenrelBtiv hobes Alter sic.hU.B. .bekundet durchseiner auffa11endunsektarischen Charuter. ROn the history andshortcomtng of the sectarian Is later reasoning, v. Rocher, ThePur5?111~ pp.18ff.

2 The isvaragita offers a good example of this process; as notedby P.E. Dumont, trans., L'isvaragita. Le Chant de Sival te.zteextrsit du Kilrma-pur8iJa (Baltimore: The Johns Hopkins Press,1933). p.8: IrL'lsvaragita apparalt done comme une oeuvre compositeet surtout comme une imitation de 1a.Bbagavacigita.P


35/510

29tradition criticism, even in a general form, provides the generalsupport and direction for more specific hypotheses about the actualdocumentary sources employed in constructing the text and theidentity of its compilers and redactors. Once again, internalreferences to tradition found in the text, and the citations andcross-references made by Ksernarajah serve as the starting point,offering the initial blueprint for this tradition criticism. Some ofthese traditions undoubtedly affected the compilers ofSvacchandatantram directly, in the form of source texts andtraditional teaching; others perhaps affected them only indirectlythrough these sources, though later educated redactors andcommentators clearly possessed first hand knowledge of multipletraditions.

Vedic and brahrnanical tradition, specifically Srauta andsmsrte precepts and practice, form the first stratum in thehistory of SvacchandataJ1tram. In contrast to, and in the likenessof, this tradition, sectarian Saivism has formed its own scriptures.rituals, and prescriptive identity. Svacchandatantram rejects thecornerstone of smsrte tradition, the var!la system and thepreeminence of the brahmanah, which is replaced with theorganized community of Saiva initiates embracing all of society.1Itremains possible, however, that this rejection may have had onlya ritual extension, which left the social order intact, and whichthus presumed complimentary coexistence with the smsrttradition. 2

1The elimination of possible rebirths through the Saivainitiation that bestows liberation (njrva?1adjk~hincludes purifyingrebirth or current status as a branmanah. In order to remove thisstatus, the ritual proceeds by purifying each of forty-eight rites thatconfer and accompany brahrnana existence, and thus presumes ak.nowledgeof dharma texts. (V. hk..10, pp.1S9ff, and infra section11.10.2.)

2 At the end of the initiation in bk.4, for example, whenenumerating the obligations of the new Saiva initiates, the text


36/510


37/510

31

(homa.fJ)~and the Iike, actually developed in reaction tocomparable tantric and agamic rites. 1

This problem reflects the larger difficulty surrounding thedevelopment of Vedic tradition. For at the periphery of srauta andsmsrte practice, but nevertheless in the Vedic tradition asrepresented particularly by texts like the Satarudriyam associatedwith the YaJurveda.fJ~are found many antecedents to the practicespresented by Svacchanda tan tram. These texts may be seen eitheras already reflecting extra-Vedic influence, or as documenting thoseportions of the Vedic tradition later elaborated in sectariantradition.Sections of these peripheral "Vedic" texts and theircorresponding traditions, though closely related to the firststratum, may be considered a second Rudra stratum in thehistory of our text because of their less clear origin and historicalassociation. Specifically~ part of the Satarudriyam has provided theBahurUpa formula of sakala Svacchenda-bhairavah, used in hisprimary rituals throughout Svacchandatantram.2 In general, thecemetery and crossroads rites performed to Svacchanda-bhairavahcontinue many of the rites inearlier srauta texts directed to

1V. Vishwambhar Sharan Pathak, Smarta Religious Tradition(Meerut: Kusumanjal1 Prakashan, 1987)~pp.33.ff.2 On the Satarudriyam~ v. Jan Gonda, "The Satarudriya," in

Sanskrit and Indian Studi~ Essays in Honour of .oamei H. H. Ingalls,eel. M. Nagatomi, et a1. (Dordrecht: E. ReidelPublishing Company,1980)~pp.75-91. The source for the thirty-two syllable Bahurtlpaformula is Ml1itr~YlllJi StJJ'l1hit~2~9, 10.(V. Leopoldvon SchrCEder,hrsg. ~Maitray8lJi SaJ'l1hit~ 2 (Leipzig:Otto Harrassowitz, 1923)~pp.130.) Later Upani~ds like the MahanarayalJ8 also contain thisformula. (V. Jean Varenne, La Maha Naraya(1a UpanJ~dlPublications de l'institut de civilisation indienne 11 (Paris: Institut decivilisation indienne, 1960), vss.282-283, pp.70-71.) Cf. also infra thediscussionof the Pasupatal:,lfor the Bahurupa formula.


38/510


39/510


40/510

34

like Svacchandatantram remains difficult to isolate.! Somecontinuity can be supposed between ascetic milieus cognizant ofsuch meta-ritualistic speculations in the Brahmanas and in olderUpanisads and the bearers of Saiva traditions. 2

Closely related to the preceding strata, brahmanical sastra.piconstitutue a fourth stratum influential in the development ofSvacchandatantram. Lacking organized expository or polemicalsections, Svl1cchl1ndaantram does not show direct evidence ofborrowing from such texts. In their developed formula practice,technical terms, and verbal cosmology, however, the Saivascriptures display strong similarities to speculative grammaticalthought as epitomized in the work of Bhertrharih.J Later Saivaauthors in Kasmir, of course, were fully versed in this thought,and if one accepts an early date for Bhartrharih, then theinfluence of his school on the anonymous Saiva texts as well seems

1For representative discussions of the Pranavah, showing thelater widespread, cross-sectarian, nature of this practice, cf. PanditDhundhiraja Sastrl, 00., Pr81JavaKaJpafrom Sri SkandapurajJa withthe commentary Pra(1avaKaJpaPrakaSaPrajJava BlJashyaby PanditBangadharendra Sarasva0; The Chowkhamba Sanskrit Series 418(Benares: The Chowkhamba Sanskrit Series Office, Vidya Vilas Press,1933); for a complementary work and study, cf. Patrick Olivelle,..Pr~avamimamsa A Newly Discovered Work of VidyaraI:lya, ..Annals of the BhandarJrar Oriental Research Institute 62 (1-4, 1981):11-101.Accordingly, in bk.5, p.108, K~marajal) states thatSvacchandataptram has asserted the identity of its Pranavah andk.nowledgeof Sivab, inorder to differentiate it from the similarPranavah found in other traditions {trVl!Xla8aruavai.s.{1avadi~urupasamye pi ... 'l

2 V. also infra on the pasupatal;t._ : 5 y. David Seyforth Ruegg, "La speculation linguistique dans lesAgama Sivaites," inDavid Seyforth Ruegg, Contributions a l'l1istoirede la phiJosophielinguistique indienn~ Publications de l'institut decivilisation indienne 1 (Paris: Institut de civilisation indienne 1959),pp.101-116.


41/510

35

likely.1Alternatively, this influence could have been mediated bya secondary tradition, or the similarities could derive from pre-scriptural sources known both to Bhartrharih and the compilers ofthe Saiva scriptures.Z

Like grammatical teaching, logic also had strong associationswith Saivas.;3 Svacchandatantram mentions logic several times,but only to single it out for condemnation, beyond even its usualrejection of rival doctrines.f In addition, astronomical andastrological learning {jyotii)iastram}. herbal and anatomicalteaching (ayurveda~/ and terms from musical treatises occur inSvacchandatantram~ but perhaps had a general influence mediated

1Accepting the date proposed for Bhartr'harih by Frauwallner,450-510A.D., would place him at the earliest limit for thedevelopment of the Saiva scriptures. (Y. Erich Frauwallner,"Landmarks in the History of Indian Logic," in Erich Frauwallner,Kleine Sc.IJriften~hrsg. Gerhard Oberhammer und Ernst Steinkellner,Glasenapp-Stiftung 22 (Wiesbaden: Franz Steiner, 1982), pp.856ff).Ksemarajah, for example, quotes Bhartrharih in bk.12, p.6.

2 On the unclarified relations between sectarian Saivas andBhartrharih, v. Ruegg, contrtbuttons, pp.62-63. V. infra on thename Svacchanda-bhairavab for another sign of possible influencefrom the tradition of Bhartr'harth.3 On the connections between Pasupata Saivism and the laterNaiyayikab, v. Gerhard Oberhammer, Wahrbeit und Transzendenz.em Beitrag sur Spiritualitat des Nyaya~ verorrentncnungen derKommissionfUr Sprachen und Kulturen SUdasiens 18. OsterreichischeAkademie der Wissenschaften, philosophisch-historische Klasse,Sitzungsberichte 424 (Wien: Verlag der Osterreichischen Akademieder Wissenschaften, 1984), esp., pp.67ff. Oberhammer notes (p.67):"Fest steat sc.IJlie.DJjclJ~aJJNyaya necn Pak~lasvamin~ soweit erii.berliefert ist~ sivaitisch ge'Wordenist. n(Onp.vii, Oberhammerdates Pak~ilasvamin Vatsyayana, to c. fifth century A.D.).4 Thus, for example, among the characteristics that disqualifya person as a goodmaster, Svacchandatantram includes (Bk.1,vs.26, p.23) being a logician (tukilra.(1). For an extendedcondemnation of logicas a useless discipline, v. bk.11, pp.98-100.


42/510


43/510


44/510


45/510

39

Next in the inclusive hierarchy of sects listed in theSvacchandatantram, immediately before the properly Saivatraditions, comes the Paiicaratra}:l, which forms the seventh andmost important non-Saiva stratum to be considered in thedevelopment of texts like Svacchandatantram.1 Both Saiva andPaficaratra scriptures share many features and topics, andsectarian rivalry characterized by concurrent imitation certainlyaccounts for much of this similarity. 2 The M~ha.bh~rClammentions the Paficaratra sect, and some scholars, accordingly,have dated Paficaratra texts earlier than the Saiva agama}:l.3 Inthat case, the Paficaratra tradition might have played a seminalrole in the development of Saiva traditions. If these early datesare rejected, however, then the similarites between Saiva andPaficaratra scriptures must be examined on a case by case basis.An important contribution of the Paiicaratra}:l might have been in

1V. bk..10. p.283. vs.676, for one reference to the Pal)caratraband Ksernaraja's categorization of their doctrine.2 Schrader. Introduction to the PaiIcaratra3 thought thatmany of the Pancaratra Sarphitas were composed in north India and~ifically Kasmir. (V. pp.18-22, 110ft.) For an introduction tosatva and P~ficaratra scriptures, and a comparative study of thecosmology, v. Vrajavallabha Dwividi, "Vaisnavesu taditaresucagem~u sadadhvevtmarsah," in Tantra-Yatral pp.14ff.3 V. Schrader, Introduction to the pa.ficaratra3 pp.16-18 for arefutation of the arguments based evidence in the Mal1abl1aratam.In the introduction to his edition of the Jl!Jyakl1yl!JsaJ71l1it~.Khrishnamacharya dated the text on palaaographical grounds to 450

A . ! > . . . : (V.:.Em~ar Krishnamacharya, ed., Jayakl1yasaJ'!1l1itaofPancaratra Agarna3 Gaekwad'sOriental Series 54 (1931. Reprint.Baroda: Oriental Institute. 1967), p.34.) Gonda (Medieval ReligiousLiterature in Sanskrit3 pp.53-54) rejects this dating, and places thecomposition of the Pancaratra sarphitas between 600 and 850A.D.


46/510

40

the modification of the Sarpkhya cosmology through the addition ofMaya as an essential phase in the theistic scheme of emanation.1

Next, in a direct line of tradition stands the eighth stratumof the Pasupatah, which Svacchandatantram recognizes as thesystem ranking right beneath itself in the inclusive hierarchy ofsystems.Z The Pasupatah form the oldest known Saiva sect, and

1V. Schrader, Introduction to the pfjcaratra~ p.74; cf. PaulHacker, Vi"varta. Studien zur Geschichte der illusionistischenKosmologie und Erkenntnistheorie der Inder; Abhandlungen derGeistes- und Sozialwissenschaftlichen Klasse, Mainz, Akademie derWissenschaften und der Literatur 5 (1953), p.235, and Paul Hacker,"Eigenttimlichkeiten der Lehre und Terrninologie Sailkaras," PaulHacker Kleine Schriften~ hrsg. Lambert Schmithausen, Glasenapp-Stiftung 15 (Wiesbaden: Franz Steiner, 1978), pp.97-98.

2 V., for example, bk.ll, pp.52-54. Cf. the summary of thissection, which considers whether it contains an interpolation.The text itself actually states that the Vaimalah perform theKapala observance. In this case, they might tentatively be identifiedas Kapa.likal). The Karukal), called in other sources, the Ka.ruka-siddhantins, might then be later identified according to thehypothesis proposed by R.G. Bhandarkar, as the KzHamukhas. Sincethe text mentions the Saivas immediately afterward, it would thenbe distinguishing, as Bhandarkar noted in other later medievalsources, four Saiva sects, the Pasupatah, the Kalamukhal), theKapalikal), and the Saivah proper.The close historical and social connections among thePasupatal), the Kalamukhal), the Kapalika.l:l, attested by theinterchangeability of epithets, would account for Ksemaraja's hereclassing the MausuU~l:l,KarukaI:1, and vaimetan together as varietiesof pasupatal:l. (v. R.G. Bhandarkar, vetsnevtsm, Saivism, and MinorReligious Systems (1913. Reprint. New Delhi: Asian EducationalServices, 1983), 91, pp.169-173.) David Lorenzen rejectsBhandarkar's derivation of the Karukal) from a disciple of Lakulesah,but accepts their identification with the KalamukhaI:1, and confirmstheir close relation in practice and ritual with the pasupatah. OnLakulesah, v. David Lorenzen, TheKapJikas and Kalamukhas(Berkeley: University of California Press, 1972), pp.175ff; on theKar'ukah, or Karuka-stddhantrns, p.182; pp.175-176; and on theKalamukhal:l, pp.97ff, especially p.107. On the location and variousdesignations of Karohanam, v. Minoru Hara, "MiscellaneaPasupatica," pp.58ff.


47/510


48/510

42

lord ( pa ti .f J) 3 bonded being or animal (pasu.fJ)3 and the bond(pasa.{1). l Buildingon this inherited symbolism, the Saiva initiationliturgically acts out the cutting of the initiate's bond by the powerof Siva}:!incarnate in the master and the rest of the ritualparaphenalia. In particular, the Pasupata scriptures prescribe therecitation of the Pranavah and of the Bahurupa formula, which asthe formula of Svacchanda-bhairavah, constitutes the center ofmany of the practices set out by Svacchandatantram.2 ThePasupata theologicalcharacterization of the highest lord (isvara.fJ)as the cause (kara{1arp) who acts without any external constraintsthrough his own desire would seem to have provided the doctrinal

1On the smearing with ashes or the ash bath, v.Psupatasiitram? I, 2; it is performed by the ascetic Saiva devotee,according to Ksemarajah, bk.2, p.ro, on the triad pati/13 and soforth, v. Sastri, Pasupata Sutras with Pancharthabhasya o rKaundinya, I, 1, pp.5ff.

2 V. for the Pranava meditation, v. Sastri, Pasupata Sutraswith Pancharthabhasya of Keundinye, 5, 24, p.124; for the use ofthe formula, "OlP aghorebhyoh ... ." 3, 21-26, pp.89ff, andSvacchanda tan trsm, bk.1, pp.34ff. On the designation Bahurflpah,v. Sastri, Pasupata Sutras with Pancharthabhasya of Kaundinya, I,17, p.39.As indicated by the PaiupatasiitramJ and by the statement ofSVacchandatantram itself, which calls this formula (Bk.10,p.36) theAghorah king of formulae (Aghora/1 mantraraja./J). SVacchanda-bhairava's formula is the identitical to the formula of Aghorah, oneof Siva's five faces, whose formulae are drawn from the TaittiriyaAra.lJyakaJ 10, 43-47. This identity is confirmed by the designation~ghoreSvari (Bk.l, p.61) for SVacchanda's Devi. As the supremeSiva}:l,however. SVacchanda}:lstill retains Aghorah as one of his fivefaces, though Ksemarajah apparently does not indicate (Bk.1, p.45)the continued use of the Bahurupa formula for his worship. (On theformulae of the five faces, v. the notes in Bhatt's edition of theRauravagama? N.R. Bhatt, ed., Rauravagama 1, Publications del'institut Irancais d'indologie18 (Pondicher'y: Institut francaisd'indologie, 1961):22ft.)


49/510

43

basis for the very name of "Svacchanda't-bhairavah.! Though thesePasupata scriptures can not be dated with certainty before theearliest stages in the compiling of the sectarian Saiva scriptures. itremains nevertheless. highly probable that antecedent Pasupata

1V. on the freedom of action of the lord. Sastri. PasupataSutras with Pancharthabhasya of Kaundinya, 2, 6, p.60; for the useof cause (.kara.(lam) to designate isvara.(1,v. Minoru Hara, "Nakullsa-pasupata-Darsanam." Indo-Iranian Journa12, 1 (1958):24, andn.102. In bk.11. vss.316ft. pp.170tf, the text characterizes the actionof the lord in language very similar to that of Kaundiya'sPaflcarthabh~yam.ThePasupata-Naiyayikal) Bhasarvajnal) expresses this theologyin language even closer to that implied by the name of Svacchanda-bhairavah. As summarized by Gerhard Oberhammer(Oberhammer, Wahrheit und rrenssenaenz, p.89): "Versucht manniimlich jenen Begriff zu Iinden, in dem sich das Wesendieseshochsten Atma verdicnten und aus dem alle diese enderenBestimrnungen als notwendige Bedingungen seiner MCiglichkeitenabgeleitet werden kb"nnen, so ist es der eines absolute, souveren,d.h. unabhangig wirkenden t/rheber (svatantrah karta) allerW.irklichkeit. nIn this interpretation, "svacchanda-" in the nameSvacchanda-bhairava1"_lmay be taken as a synonym of "svatantra-."Alternatively or complementartly, the conceptual basis for thisname may have been supplied by the linguistic theology ofBhartrharih. There, similarly on the basis of the Paniniancharacterization of the agent (.karta) as autonomous (svatan treh),the lord, as supreme agent, has autonomy (svatantryam).as hisfirst and highest (saktib). Similarly, in the sarva theologyofemanation represented by svsccaendetsntrem, the lord firstspontaneously manifests through his desire (iccha) which is his basicpower or goddess.While the text does not refer to this conceptdirectly, one section (Bk.7, p.318,vs.255) does characterize therealized adept who has become identified with Svacchanda-bhairavah as autonomous and possessingundiminished capability(asvatantro 'luptasamarthyas-")(OnBhartrharlh v. Ruegg, Contributions, p.66, and K. A.Subramania Iyer, A Study of the Vakyapadiya in the light of theAncient Commentaries, DeccanCollegeBuildingCentenary and SilverJubilee Series 68 (Poena: DeccanCollegePostgraduate and ResearchInstitute, 1969),pp.110ff;On the concept of the svetentrii saktiiJ, v.GopinathKaviraj, Aspects of indian Thought (Burdwan: TheUniversity of Burdwan, 1966),pp.220-223;for Ksemaraja'sexplanation of the title v. his commentary in bk.1, pp.9-10.)


50/510

44

circles constitued the proximate and immediate source for many ofthe practices described in Svacchandatantram.1 The Pasupatatradition, therefore, has apparently transmitted the Vedicforrnulee, discussed previously, and the ascetic traditions offormula meditation taught in Svacchandatantram. In addition, itappears to have first incorporated and adapted the Sarpkhyasystem that serves as the foundation for the scripturalmetaphysics of sectarian Saivism. Svacchandatantram apparentlyhas incorporated not only the Sarpkhya system through thePasupatal,l but even parts of their polemic against theSarpkhyam. 2

1Kaundinya's Pancrthabh~yam on Pasupatasutram isusually dated to c. 600A.D., and the 6a.(1akarikawith the Ratna,/ikcommentary attributed to Bhasarvajfiah to c. 950A.D. (V. Hara,"Nakulisa-Pasupate-Darsanam, " pp.9-11; Oberhammer, Wahrheitund Transsendenz; pp.107-108, n.170 and n.171.) Other anonymoustexts, such as the Athervehres Upanisad, the Linga- and Kilrma-Puranas also contain Pasupata materials. The Linga, for example,records the Bahurtlpa formula. (V. J.L.Shastri, ed., Liligt1p,ura.(1aofSage /(r.S1JaDvaipayana Vyasa with Sans./crit Commentary Sivato~i.(1jof 6a.(1eSaNatu (Delhi:MotHalBanarsidass, 1980),2, 26, 6, p.176.) Inthe absence of precise dates and clear sectarian history, however,the exact sources and lines of transmission of Pasupata doctrines tothe compilers of Svacchandatt!Jmtram can not be determined.

2 V. for the general incorporation of the Sarpkhyam by thePasupatal)" Friedrich August Schultz, Die philosophisch-theologischeLehren des Pasupa ta-Systems nach dem Pancarthabha,sya und derRatna,/l:ka (Walldorf-Hessen:Verlag fUr Orientkunde Dr. H. Vorndran,1958), pp.27ff; on the polemics between the Pasupata-Naiyayikal:land SaJp.khyam, v. Minoru Hara, "Pasupata and Sarpkhya-Yoga,"Journal of Oriental Research (Madras) 34-35 (1973): 76-87, andGeorgeChemparathy, "The Testimony of the Yuktidipika Concerningthe Isvara Doctrine of the Pasupatas and Vatsestsas," WienerZeitschrift ru r die Kunde Sild- und Ostasiens 9 (1965): 119-146, esp.123. There Chemparathy notes that the Yuktidlplkakara's explicitlyrejects the puruseh as an agent. Svacchandatantram apparentlycontains directs echoes of this debate. In bk.12, p.31, the textattributes the identical assertion of non-agency to the Sarpkhyam asthe characteristic of their liberation, and rejects it as delusional; theself at this stage, the text confirms, remains subject to lsvara}:l.


51/510

45

In contrast to the Pasupatah, the Saiva scriptures assign apreeminent and indispensable place to their ritual of djk~a orinitiation, though many of its promised benefits incorporate thestated results of Pasupata practice. 1Socially and historically, thetwo branches of Saivism differ since the Pasupatah appear to havebeen exclusively brahmana ascetics, while at least by its owndeclarations, Svacchandatantram professes to be open to all ranksof initiates from all groups of society. 2 But the ascetic and

1For a declaration that only Saiva initiation confers liberation,v. bk.l0, vss.704-705, pp.296-297.Unlike lower ranked and more distant systems, such as theBuddhists, Svacchandatantram states that the followers of thePasupata and Kapalika observance, though not able to obtainliberation through their practice, nevertheless do not become subjectonce again to transmigration. Rather they proceed, after subsequentpurification, from their comparatively high level of attainment, theplane of Isvarah, to liberation. (V. bk.11, p.103). Similarly, thevaimaran, who followthe Kapalika observance, and who are closestto the Saivah proper, apparently obtain their slightly h~gher statusdue to their purification by an initiation related to the Saivainitiation. (V. bk.ll, p.53, vs.73, and commentary.)Similarly, in the description of the cosmology in bk.10,(pp.252-253), the text mentions, without sectarian designation,liIigam worshippers who inhabit the much lower world of Rudrah,still located at the plane of earth in the egg of Brahma. The textapparently alludes here to early or causal Rudrah or Sival)worshippers, prior to or outside of the organized Saiva sects such asthe pasupatal). Though attaining only this low level and failing toattain liberation because they are deprived of the knowledgeconferred by initiation, (dj.k~a.Jfianavihjna(1):1hey, nevertheless, thetext asserts, do not reenter transmigration, but throughsupplementat grace proceed upward to the abode of the supremeSiva}:l.

2 A large part of Pasupata practice requires concealing theirstatus as celibate brahrnanah and engaging in seeminglyreprehensible practices in order to obtain merit by unjustifiedcondemnation. (V. Minoru Hara, "Transfer of Merit," Adyar Li,braryBulletin 31-32 (1967-68):397ff.). Both Ksemarajah and the text ofSvacchandatantram confirm this brahmanical association. Whenlisting the brahmal).ical rites that require purification during


52/510


53/510

47

While the text often praises the superiority of Saivaknowledge or initiation over the non-Saiva, and sometimes overthe Pasupata}:l, it never refers to the Siddhantah or to othersubsidiary Saiva sects or traditions. Similarly, it never refers toother scriptures (agamafJ or tentrssu) except in general terrns.JNevertheless, though not the self-professed document of any Saivaschool, many sections of Svacchandatantram parallel the structureand content of the agama{1 that form the scriptural corpus of the

Cf. HilkoWiardo Schomerus, Arunentis Sivaji'1anasiddhjyar 1,Bertrage zur SUdasien-ForschungSUdasien-Institut UniversitatHeidelberg49a (Wiesbaden: Franz Steiner Verlag, 1981):esp. 2-3ff,42ft, for the identification of various sarva sects and their mostimportant opponents. This thirteenth century Tamil Saiva-siddhantincompendium successively classifiestraditions] with appropriatenames for each group, as Saiva versus non-Saiva, Siddanta-relatedversus non-Siddhantin related Saiva, and so forth. In contrast, a listin the early Tamil epic Ma(lin1ekalal~ written from a Buddhistviewpoint, enumerates in its list of doctrines only a singlesaivavada.(1. "The ambiguity in the use of the term existed at the time ofKsemarajah. When commenting on the passage praising thesuperiority of the Saivab over the PasupataQ, he notes, that it maydesignate either, as the context implies in another list of traditionshe quotes, either the Saiva-siddhantin based on agama.(1Jor, as inthe passage in sveccnenaetentrem, the Saiva tradition in general.(V. bk.11, pp.54-55) In a comparable list of doctrines in Shastri,M.K., ed., Tantraloka.(1J (12, vs.320), Abhinavaguptah substitutesthe term SiddhantaQ for the ambiguous Saiva, and here andelsewhere in this section (for example, vs.317) uses Saiva in ageneral sense contrasted with non-Barvah such as the Vaisnavah., (For citations from texts using SatYa to designate on'e'of fourSaiva sects, v. Lorenzen, The Kylikas ant?Kalamukha~ pp.1-12.V. also on the use of the terms Saiva and Saiva-siddhantab,Brunner-Lachaux, "Importance de la litterature agamique, ,.pp.114ff.)

1For example, the text (Bk.I, vs.13) simply characterizes thegoodmaster, as "knowing the scriptures of Sivab(sivasastravidhanaJ.na-jJ ,. the bad master as "delighting in otherscriptures (anyasastrara te-). "


54/510

4 8

dualistic Saiva-siddhantins.1 Moreover, by sharing and affirmingSaiva initiation ritual and coordinated cosmology, these sectionsimply that when Svacchandatantram was compiled, there existed.if not a socially and canonically defined sect of Saiva}:!, then atleast interacting lineages of ritualists and of teachers, andrelatively stable collections of doctrines and of practices. And thecompilers of 5'vacchandatantram drew on this tradition, but notexclusively. This lack of exclusivity, in turn, might further implythat their sectarian consciousness had not yet crystallized intotheological tenets and set liturgical prescriptions, nor reached thematurity marked by later sub-sects and secondary traditions. 2

Svacchandatantram differs from the preceding Saiva agamaJ;by sometimes designating or identifying the supreme Sival;1of thegam{i as Svacchanda -bhairavah , In addition to substitutingSvacchanda-bhairavah for Sivah in standard Saiva ritual,cosmology, and meditation, the text describes special modes ofworshipping Svacchanda -bhairavah. By occupying the place of thesupreme Sivah, Svacchanda-bhairavah incorporates Siva's retinue,and instead of forming an aspect of Sivah, receives instead his own

1The long notes to Helene Brunner-Lachaux's translation andstudy of tjle SomasambhupaddhatJ~, troisieme partie, whichdiscusses Saiva initiation, frequently compare different statements ofBvacchandatantram and the statements of Bvacchandatantram withthose of other Saiva gamb and the Siddhantin commentators.2 Cf.Brunner-Lachaux, Somasambhupaddhati, troisieme partie,pp.548-549: "t/ne lecture attentive de tous les Agama parvenusjusqu~ nous et de leur commentaires permettrait peut-etre dereconneltre un stade du sivaisme iigamique ou l'oppositJonSiddhanta/ &colesdu Nord titait ignoree, soit que les Tantra dun'aie.nt connus, soit que les differences entre cesTantra (enparticulier les Bhairavatantra, devenus Tantra du daksinamarga) etlesTantra ouAgama de l'&cole du Sud rr'eierrt pas ete senties comrnedes opposit/ons fondamentales.


55/510


56/510

50

identifies the adept (sadhaka_(J) who undertakes the special worshipof Svacchanda-bhairava}:l, as a performer of the great observance

For the special use by KapalikaQ of human flesh and bdellium(guggulum)7 v. Lorenzen, The Kaplikas and Kalamukhas, pp.26-27;sveccnenaatantrsm prescribes them as substances for oblating inbk.6, pp.146-147. V. Lorenzen, pp.20ff, for the association ofworshippers of Bhairavah with the KapalikaQ. As he notes, in Barra'sHarsacaritam7 there is a Kapalika ascetic named Bhairavacarya:"Bhairavacarya's name indicates he worshipped Siva as Bhairava,the form of the god held in especial esteem by tantric groups suchas the Kapalikas." Cf. also pp.32ff, where Lorenzen recounts thelegends of Sarnkara's encounter with Kapalika ascetics named Ugra-Bhairava and Unmatta-Bhairava. In these legends the Kapalikaascetic has a form identical to that of the manifest Bhairavahhimself.On the cult of the eight bhairavah, and its connection with theKapalikaQ and burning ground, v. P.H. Pott, Yoga and Yantra,Koninklijk Institut voor Taal-, Lang- en Volkenkunde, TranslationSeries 8 (The Hague: Martinus Nijhoff, 1966), esp., pp.82-83. "Theeight bhairavas ... form a series which occurs with only slightvariations in more than one text, primarily in the texts which havegreat authority among the Kapalikas or Aghori, followers of the godBhairava. "

Moreover, the name Kotaraksah, for the terrifying aspect ofSvacchanda-bhairavah (Bk.9, pp.28-29), might suggest that heforms the male counterpart to the terrifying form of Devi,designated most often Camuz:lQa, who is frequently described asko!arak$i. Camunda, in turn, is part of a terrifying group of eightmothers connected with the cremation ground; this would thenassociate Kotaraksah with the forms of Bhairavah worshipped insidethe circle of mothers (matrkah which are discussed in bk.l, pp.28-31, and bk.l0, pp.418-421. V. also Pott, p.86: " ... the eightBhairavas can be conceived as the krodha- aspects of the eightguardians of the world, while the eight Matrkas thus form theirsaktis. (V. on CamulJQa, Mallmann, Les enseignementsiconographique~ pp.153-157.)For references to entering another's body and possession(aveSab~ among the Pasupatal), and its later development as aform of mystical union documented in other Bhairava texts, v.Minoru Hara, review of Le Bbakt~ by Lilian Silburn, Indo-IranianJournaJ9, 3 (1966): 212, n.3. Hara suggests that the mystical fusionof later non-dualistic Saivism, expressed by terms such asusamaveSa!Jnrepresents a reinterpretation of Pasupata-Kapalikaconcepts of possession or entering another's body (aveia!J)'


57/510

51

(rnehsvratsm) traditionally connected with the Kapalikal;..l Inaddition, like the Kapalikah, the Svacchandatantram rejects theapplicability of caste, and other distinctions of purity or impurity,and even perhaps in antinomian fashion, any moral restrictions forthe realized adept.j' Perhaps acknowledging its closeness to thistradition, in its hierarchy of other schools, the text places thosePasupatah who perform the Kapala observance at the level rightbefore the Saiva proper. 3

1Bk.9, pp.35-36., Alluding to the Puranic origin myth of thernsnsvretsrn, where sivah atones for cutting off Brahma's head byroaming the earth as Bhairavah in the form of a Kapalikal),Ksemarajah explains that the adept pursues this vow in order toeffect his imitation, sc. identity with the supreme Bhatravah whichwill guarantee the efficacy of his ritual performance:ubhasmoddhlliitatvdin sdhakasya mantrradhanvasaremensvretttvemustemete eva parabhairavanukaritvenapitedebiurnsruno vighna naiyanti. n (On the mahavratam and itsconnection with the Kapalikal), v. Lorenzen, TheKapaJjkasandKalmukh


58/510

52

Historical evidence suggests that the Kapalikah and relatedBhairava sects were the original propagators and transmitters ofsexual rituals that later were incorporated in the tantric varietiesof many traditions. 1 In Saiva tradition, the later elaboratesymbolic doctrine and ritual which developed surrounding theworship of Bhairva or Bhairavi in these sexual rites, and whichwas transmitted secretly within circles of adepts, became knownas the Kaula tradition. 2 This later Kaula tradition, closely relatedto the preceding Kapalika stratum, represents the eleventh andlast stratum that left its mark on Svacchandatantranl. Althoughthis text does not describe sexual rituals explicitly, its last bookexplains both the code language (chummaka.p) used by circles ofadepts, and the inner significance of the encounter (meJakam)between the adepts and female deities or partners (yoginyaJ;).:5

1For the association of Kapalikal) and sexual rites, v.Lorenzen, The Kpllkas and Klmukhas.. pp.88ff.On the cemetery cult of Bhairavah and the Kapalikal'),as thesource for later Buddhist sexual tantric practices, v. Shin'ichi Tsuda,"A Critical Tantrism," Memoirs of the Research Department of theToyo Bunko 36 (1978), pp.172-174; cf. also DavidSnellgrove, Indo-Tibetan Buddhism. Indian Buddhists and Their Tibetan successors: 1(Boston: Shambhala, 1987): 152ft. For the association in the Buddhisttradition, of non-sexual practices with Bhairavah, similar to thosedescribed by ~vacchandatantram v. the entry "Bairaba" in PaulDemievtlle, ed., H6b6girin 1 (A-Bombai) (Tokyo:Maison franco-japonais, 1929): 47.

2 On the nature and practices of the Kaula tradition and itsdevelopment from the Kapalika cult, v. Alexis Sanderson, "Purityand Power among the Brahmans of Kashmir," in Michael Carrithers,et al., ed., The Category of the Person. Anthropology; Philosophy;HIstory (Cambridge: CambridgeUniversity Press, 1985), pp.202ff. (V.for the Kaula as a secret tradition within SaiVism, p.215, n.125; forthe Kapalika-Kaulaorigin of corresponding Buddhist practices, v.p.214, n.106.): 5 Cf. infra section II.15.


59/510


60/510

54

criticism, naturally, can not afford to ignore the slightestirregularity that might hint at the hand of a redactor, and thusruns the constant risk of over-interpreting and multiplying thenumber of sources, recensions, and compilers.1As a corrective tothis tendency, a schematic and simplying presention of the moresecure results of this analysis will precede a detailed discussion ofmethods and illustration of their application. 2

The main deity, title, and dialogue participants of thepresent recension of Svacchandatantram strongly support thehypothesis that at its basis lies a written document from the"Bhairava" tradition. This document, the "Bhairava" source, as itwere, likely recorded practices associated with the worship of

1For an example of a methodic attempt to read a text as aunity, in reaction to this type of criticism. v. GeorgFeuerstein. TheYoga-Sutra of PatafiJal~ An Exercise in the Methodology of TextualAnalysis (NewDelhi: Arnold Heinemann, 1979). Feuersteinprogrammatically argues (pp.39-40): "Past scholars, on the whole,proceeded from the assumption that the Yoga-Sutra cannot possiblybe a single homogeneous textual entity owing to its apparentdisorganisation. In the followingI shall commence from the oppositeend as it were by presupposing the perfect homogeneity of the text.In other words. I shall look for points which seem to contradict, orat least seriously challenge, this basic working hypothesis. In thisway I hope to avoid the fallacy common to all attempts of textualcriticism so far, namely to cut more and more slices from the cakeuntil it simply vanishes out of sight and nothing but disconnectedfragments - quite meaningless in themselves - are left behind." Healso attempts to isolate this critical tendency as a cultural basis:..... before one contrives distortions, interpolations, etc., one mustask onself seriously whether what seems to be 'corrupt', 'confused'or 'patched together' is not merely the result of an unwarranteddemand for absolute logical consistency characteristic of our specificthought pattern." As an exemplar in Indology of the dissecting typeof text criticism. Feuerstein appropriately cites Friedrich Weller'scritical study of the KathopanJ$ad: Friedrich Weller, Versuch einerKritik der Kaf,hopani.sad, DeutscheAkademie der Wissenschaften zuBerlin, Institut fUr Orientforschung. Veroffentlichung 12 (Berlin:Akademie-Verlag, 1953).

2 V. infra in the summaries of individual books.


61/510

55Bhairavah, or specifically. Svacchanda-bhairavah. Alongsideinformation about formulae, gestures, and other aspects of ritual,panegryic interludes probably mixed with inconographic cummeditational data. This Bhairava source might have served tocodify the essential elements of the cult of Svacchanda-bhairavah,and might have been used as a guide to rituals described not onlyin the text of Svacchandatantram3 but also known throughseparate written and oral tradition. Thus, though. not exhaustive,this source likely contained more detailed instructions for specifictypes of ritual, particulary, the malevolent rites conducted in thecremation grounds.Perhaps a desire to exalt this Svacchandah form ofBhairavah formed the primary motive behind the compiling of thistext. In that case, the Svacchanda stamp of nomenclature,attributes, and the like, could have been secondarily imprinted onan "Ur-Bhairava" source. Tantric texts of this period in manytraditions, however. attest to the creation and multiplication ofdeities as the representation and personification of doctrinal andmeditational concepts .J Accordingly, the "svacchanda-" epithet,and the details of his form may simply reflect a theologicalcharacterization of Bhairavah when viewed as the supreme Sival;t,rather than the proper name of a particular minor deity whocame to be identified with Bhairavah.2 Therefore, without

1 So Marte-Therese de Mallmann asserts at the beginning of herIntroduction a Piconographie du tantrisme bouddhtoue. Bibliothequedu centre de recherches sur l'asie centrale et la haute asie 1 (Paris:Librairie Adrien-maisonneuve, 1975), p.1: H L 'on se rappellera toutd'abord que cette iconograpl1ie est TOUJOURSe reflet ou l'illustrationde notions religieuses ou de concepts phllosopl1iques. En consequence;eucun detaIl n'est gratuit: rnensuretions, couieurs; gest~attitutd~ expressions de pl1ysjonomi~ ettrtbuts, etc.3 onttous unesigni'/ica tion precise. ..

cvacchnda tantra

Documents

Transcript of cvacchnda tantra