9010-8804-1-PB

8/19/2019 9010-8804-1-PB

1/41

JIABSJournal of the International

Association of Buddhist Studies

Volume 31 Number 1–2 2008 (2010)

8/19/2019 9010-8804-1-PB

2/41

The Journal of the International Association of Buddhist Studies (ISSN0193-600XX) is the organ of theInternational Association of BuddhistStudies, Inc. As a peer-reviewed journal,it welcomes scholarly contributionspertaining to all facets of BuddhistStudies.JIABS is published twice yearly.Manuscripts should preferably be sub-mitted as e-mail attachments to:[email protected] as one single le,complete with footnotes and references,in two diff erent formats: in PDF-format,and in Rich-Text-Format (RTF) or Open-Document-Format (created e.g. by OpenOffi ce).Address books for review to:JIABS Editors, Institut für Kultur- undGeistesgeschichte Asiens, Prinz-Eugen-Strasse 8–10, A-1040 Wien, AUSTRIA

Address subscription orders and dues,changes of address, and business corre-spondence (including advertising orders)to:Dr Jérôme Ducor, IABS TreasurerDept of Oriental Languages and CulturesAnthropoleUniversity of LausanneCH-1015 Lausanne, Switzerlandemail:[email protected]: http://www.iabsinfo.netFax: +41 21 692 29 35Subscriptions to JIABS are USD 55 peryear for individuals and USD 90 per yearfor libraries and other institutions. Forinformations on membership in IABS, seeback cover.Cover: Cristina Scherrer-SchaubFont: “Gandhari Unicode” designed byAndrew Glass (http://andrewglass.org/ fonts.php)© Copyright 2010 by the InternationalAssociation of Buddhist Studies, Inc.Print: Ferdinand Berger & Söhne

EDITORIAL BOARD

KELLNER BirgitKRASSER HelmutJoint Editors

BUSWELL RobertCHEN JinhuaCOLLINS StevenCOX ColletGÓMEZ Luis O.HARRISON PaulVON HINÜBER OskarJACKSON RogerJAINI Padmanabh S.KATSURA ShōryūKUO Li-yingLOPEZ, Jr. Donald S.MACDONALD AlexanderSCHERRER-SCHAUB CristiSEYFORT RUEGG DavidSHARF RobertSTEINKELLNER ErnstTILLEMANS Tom

8/19/2019 9010-8804-1-PB

3/41

JIABSJournal of the International

Association of Buddhist Studies

Volume 31 Number 1–2 2008 (2010)

Obituaries

Jonathan A. SILK In memoriam, Erik Zürcher (13 Sept. 1928 – 7 Feb. 2008) . . . . . . 3

Articles

Diwakar ACHARYA Evidence for Mah ā yāna Buddhism and Sukh āvat ī cult in In-dia in the middle period – Early fth to late sixth century

Nepalese inscriptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .23

Early Chinese Buddhist translationsContributions to theInternational Symposium “Early Chinese

Buddhist Translations,” Vienna 18 – 21 April, 2007Guest editor: Max Deeg

Max DEEG Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .79

Max DEEGCreating religious terminology – A comparative approach toearly Chinese Buddhist translations . . . . . . . . . . . . . . . . . . . . . . . . . . . . .83

Hubert DURT Early Chinese Buddhist translations – Quotations from theearly translations in anthologies of the sixth century . . . . . . . . . . . . . 119

Toru FUNAYAMAThe work of Param ārtha: An example of Sino-Indian cross-cultural exchange . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 141

8/19/2019 9010-8804-1-PB

4/41

Contents2

Andrew GLASSGuṇabhadra, B ǎoyún, and the Sa ṃ yukt āgama . . . . . . . . . . 185

Paul HARRISON Experimental core samples of Chinese translations of two Buddhist S ūtras analysed in the light of recent Sanskrit man-uscript discoveries . . . . . . . . . . . . . . . . . . . . . . . . . 205

Elsa I. LEGITTIMO Reopening the Maitreya- les – Two almost identical early Maitreya s ūtra translations in the Chinese Canon: Wrong at-tributions and text-historical entanglements . . . . . . . . . . . . . . . . . . . . 251Jan NATTIERWho produced the Da mingdu jing (T225) ? A reas-sessment of the evidence . . . . . . . . . . . . . . . . . . . . . . 295

Jungnok PARK (†)

A new attribution of the authorship of T5 and T6 Mah ā pari-nirv āṇas ūtra . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339

Jonathan A. SILKThe Jifayue sheku tuoluoni jing – Translation, non-transla-tion, both or neither ?. . . . . . . . . . . . . . . . . . . . . . . . 369

Stefano ZACCHETTI

The nature of the Da anban shouyi jing T 602reconsidered . . . . . . . . . . . . . . . . . . . . . . . . . . . . 421

ZHU QingzhiOn some basic features of Buddhist Chinese . . . . . . . . . . . 485

Book review

Tsunehiko SUGIKIDavid B. Gray , The Cakrasamvara Tantra (The Discourse ofŚ r ī Heruka): A Study and Annotated Translation. . . . . . . . . . . . . . . . 505

• Notes on the contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 543

8/19/2019 9010-8804-1-PB

5/41

Book review

David B. Gray,The Cakrasamvara Tantra (The Discourse of Ś r ī Heruka ): A Study and Annotated Translation . The AmericanInstitute of Buddhist Studies at Columbia University in New York,

co-published with Columbia University’s Center for BuddhistStudies and Tibet House US, New York 2007. ISBN 0975373463

Introduction

The Cakrasa ṃvaratantra (CS) is a principal tantra of theCakra-saṃ vara scriptural cycle, which is one of the largest collectionsof Buddhist Yogin ī tantra literature from the early medieval SouthAsian world. TheCakrasa ṃvara tradition was imported intoneighboring areas such as Nepal, Tibet, and Bhutan, and has func-tioned as one of the most important sources in the formation ofreligio-cultural systems in these areas. Its thought and practice arealso maintained “in other regions in uenced by Tibetan Buddhism,including Mongolia, Russia, China, and elsewhere,” notes David BGray, the author of the book under review, “as Tibetan lamas havebeen living and teaching in diaspora.” (p. xv.)

Gray’s study aims at providing the rst critical translation of theCS, richly annotated and accompanied by analyses of its contentsand contexts. A critical edition of the CS is not included, whichsome may regard as a shortcoming, but Gray’s critical edition ofthe CS is forthcoming as a companion volume to his study.

1 Outline

Gray’s study consists of six parts: (1) an introduction into the studyof the CS (pp. 1–152); (2) an annotated critical translation of the CS(pp. 153–384); (3) Sanskrit-Tibetan-English and Tibetan-Sanskrit-English glossaries (pp. 385–392); (4) a Conspectus Siglorum listing

Journal of the International Association of Buddhist Studies

8/19/2019 9010-8804-1-PB

6/41

506 Tsunehiko Sugiki

the Sanskrit manuscripts (Skt mss) that Gray used for his transla-tion (pp. 405–408); (5) a bibliography (pp. 409–436); and (6) anindex of Sanskrit, Tibetan, and English terms (pp. 437–447).

Part one provides a general introduction into the study of the CS.Gray presents the textual materials which he used for his translationand explains his translation methodology. This part also includesanalyses of several important aspects of the CS and its backgroundsuch as the classi cation of tantra literature, dates when the CSand some other tantras belonging to theCakrasa ṃvara scripturalcycle were compiled, the contents of the CS, and the scriptural andideological contexts within which the CS was compiled and used. Iis regrettable that Gray does not provide a full textual and contex-tual study ofŚaiva-Buddhist interrelations, especially since he up-holds the idea that “the Buddhist Yogin ī tantras were signi cantlyin uenced byŚaiva Kāpālika practices” (p. 8, n. 19). But this omis-sion may also re ect the stance that we should avoid drawing hastyconclusions on this complex issue, for he says that “the undoubt-edly complex relationships that exist betweenŚaiva and Buddhisttantric textual traditions will only be determined conclusively onceall of the surviving texts have been critically edited and published”(p. 9, n. 19). This position may be controversial because eff orts tocreate critical editions and eff orts to determine textual relation-ships are not separate from each other – they are to be concurrentlymade and reciprocally associated. But we should not ignore thatthis rst part of Gray’s study provides much information on thecontents and contexts of the CS and theCakrasa ṃvara scripturaltradition, which is no doubt useful for anyone interested in TantricBuddhism.

Part two, the annotated critical translation, is the main part ofthe study. Gray’s translation is based on his unpublished criticaledition of the CS. This edition chie y relies on three Skt mss of theCS, on Tibetan translations of the CS, and on eleven Indian com-mentaries along with some Tibetan commentaries. It also makes useof Kalff ’s edition of selected chapters of the Abhidh ānottaratantra(Kalff 1979), and of Skt mss of this work which contains manyparallel passages and is therefore quite useful for recovering thetext of folia that are missing from the extant Skt mss of the CS. The

8/19/2019 9010-8804-1-PB

7/41

David B. Gray,The Cakrasamvara Tantra 507

text of the CS as edited by Gray is partially recorded in the annota-tion. In his footnotes, he also adduces various interpretations fromthe commentaries, which makes this book not only the rst criticaltranslation of the CS but also a useful guidebook for comparativestudies of its commentarial literature.

In short, Gray’s study is the rst full translation of the CS, itserves as a guide to its commentarial literature and provides muchtextual and contextual information on the IndianCakrasa ṃvara tradition. This makes it a ‘must-read’ for students and scholarswho research the IndianCakrasa ṃvara tradition in particular andIndian and Tibetan Tantric Buddhism in general. However, it isalso beset by problems. In the following I would like to focus onproblems in dating the tantras of theCakrasa ṃvara scriptural cy-cle and problems of Gray’s translation of the CS and the materialsused for it. I shall then turn to the analyses of the origin myths ofHeruka and hismaṇḍ ala , and of the structure and functions of theTriple Wheelmaṇḍ ala .

2 Dating the tantras of the Cakrasa ṃvara scriptural cycle

Gray notes that a precise dating of the CS and other tantras of theCakrasa ṃvara tradition is currently a diffi cult task (p. 11 and 20),but nevertheless off ers a hypothesis of his own.

2.1 Date of the CS

Gray argues that it is likely that the CS was compiled in the eighthcentury, for the following reasons: (a) The CS mentions the names ofBuddhist scriptures that can be dated in the late seventh century orthe rst half of the eighth century, such as theSarvata thā gatatattva-saṃgrahas ūtra , Guhyasam ā jatantra , Vajrabhairavatantra , Ś r ī -

pa ra mādyatantra , and theSarvabuddhasam ā yogatantra .1 (b) It isknown from Tāranātha’s History of Buddhism that Jayabhadra com-posed hisCakrasa ṃvarapañjik ā, the oldest commentary of the CS,

1 Throughout his study, Gray spells the title of this tantra with ‘samayo-ga ’. However, ‘sam ā yoga ’ (full title:Sarvabuddhasam ā yoga ḍākin ī jālasaṃvaratantra ) is more commonly used, and very likely also correct.

8/19/2019 9010-8804-1-PB

8/41


during the early- to mid-ninth century. (c) In his N āmasa ṃgī tiṭī k ā,Vilāsavajra (mid- to late-eighth century), quotes one pāda and oneword from the CS:glang chen ko rlon gos su gyon (/ zhes pa nidpal ’khor lo bde mchog gi rgyud las te ), andkeng rus (ni dpal bdemchog ’khor lo’i rgyud las so ). (pp. 11–14.)

Given that Tāranātha’s account is ambiguous and may be un-reliable, Vilāsavajra’s very short quotations appear to be the onlyevidence to support Gray’s hypothesis that the CS was already ac-tive in the eighth century. Gray correlates Vilāsavajra’s glang chenko rlon gos su gyon and keng rus respectively withhasti car ma-viruddha ṃ ca in chapter 2 of the CS (where the actual readingis hasticarm āvaruddha ṃ ca ) andkaṅk āla in chapter 48. However,the former correlation is problematic. Vilāsavajra’s work is a com-mentary on the N āmasa ṃgī ti, and the Sankrit of the pāda in ques-tion in the N āmasa ṃgī ti isgajacarmapa ṭārdradh ṛ k ,2 which cannotbe found in the CS. Vilāsavajra might have read the pāda in the CSfreely and related it to the pāda in the N āmasa ṃgī ti freely on thisbasis, but this is certainly not conclusive evidence. The Sanskritsource ofkeng rus in the N āmasa ṃgī ti is, indeed,kaṅk āla. 3 Butthis, too, is insuffi cient evidence. Clear and extensive parallel pas-sages along with a reference to the name of its source text wouldcertainly be more decisive.

There is also a problem regarding the name of the tantra towhich Vilāsavajra refers. Although Vilāsavajra calls the scripturalsource of the pāda and word in questiondPal ’khor lo bde mchoggi rgyud or dPal bde mchog ’khor lo’i rgyud , which in Sankrit isŚ r ī cakrasa ṃvaratantra , we should bear in mind the possibilitythat this tantra may have previously been named Heruk ābhidh āna rather thanCakrasa ṃvara . This is suggested by the change of itsname in its chapter 51. In chapters 1 to 50 it calls itself Heru k ā-bhi dh āna (iti ś r ī heruk ābhidh āne ….). However, in chapter 51, thatname is said to refer to the large scripture of one hundred thousandverses from which this tantra was selected, and the name of thistantra is given asCakrasa ṃvara (ś r ī cakrasa ṃvara ṃ n āma mah ā-

2 N āmasa ṃgī ti: Ādar ś ajñ ānam , 3d. 3 N āmasa ṃgī ti: Ādar ś ajñ ānam , 1c.

8/19/2019 9010-8804-1-PB

9/41


yogin ī tantrar ā ja). As I shall argue below, it is very likely that chap-ter 51, together with several verses of chapter 50, was not includedin the oldest version of this tantra: it was added to the oldest versionafter Jayabhadra, who was active after Vilāsavajra.4

Gray’s analysis of Vilāsavajra’s references is thus somewhat prob-lematic. However, we cannot deny the possibility that some form ofthe CS existed in the age of Vilāsavajra because, as Gray stresses,Vilā savajra surely mentions a tantra named ‘Cakrasa ṃvaratantra ’and distinguishes it from theSaṃvaratantra , by which he refers tothe Sarvabuddhasam ā yogatantra . It is also possible that theCa-kra saṃ varatantra mentioned by Vilāsavajra is a work diff erentfrom the CS – a work that did not survive.

One should also bear in mind the possibility that the date ofcompilation of the CS is not so long before (or might even be veryclose to) the date when Jayabhadra was active. I would here liketo call attention to four remarks about the historical stages of thecompilation of the CS and of Jayabhadra’s commentary that I madein 20015 and that are not suffi ciently taken into consideration inGray’s study. They may be of some help for future studies about thedate of compilation of the CS.

[1] There were several diff erent versions of the CS, some ofwhich Gray mentions. Gray overlooks, however, that these can beroughly divided into two: (a) a shorter version that contains chap-

ters 1 to 49 and the

rst half of chapter 50 (= 37a3 of the Vadodarams)6 of the extant CS and (b) a longer version that contains all chap-ters, from 1 to 51.

[2] The shorter version is very likely to be older than the longer,and Jayabhadra is very likely to have used one of the oldest textsthat belong to it. The text that Jayabhadra used does not know the

4 However, there is also the possibility that this tantra was calledCakrasa ṃvara from the outset because Jayabhadra, the oldest commentatoron this tantra (or at least the commentator who used the oldest version of thistantra),refers to it under that name. See also Sugiki 2001 for further discus-sion of the titles Heruk ābhidh āna andCakrasa ṃvara . 5 See Sugiki 2001. 6 See below p. 513 for the Sanskrit mss of the CS.

8/19/2019 9010-8804-1-PB

10/41


chapter divisions given in the extant CS; in fact, it gives no chapterdivisions at all. Furthermore, Jayabhadra’s commentary does notmention parts that are not contained in the shorter version. It isquite unlikely that Jayabhadra intentionally skipped commentingon these parts because they provide instructions into the systemof the internal Herukamaṇḍ ala , which is a main doctrine of theCakrasa ṃvara tradition. They also introduce other systems that arehighly Buddhist Mahāyānic and make the CS more Mahāyānic.7

[3] The shorter version (and also Jayabhadra’s commentary) isdevoid of a clear idea of internalCakrasa ṃvara holy sites corre-sponding to their external forms. This idea rst appears in the lasthalf of chapter 50 of the extant CS and is in theCakrasa ṃvaratradition generally accompanied by such terms asbāhyādhyātma- ,sab āhyādhyātma- , or the like. It became one of the principal ele-ments in the practice of ‘the creation stage’ (utpattikrama ), i.e., thevisualization of the Herukamaṇḍ ala in theCakrasa ṃvara tradi-tion after the CS. After the addition of the last half of chapter 50and of chapter 51 to the shorter version (i.e., after the compilationof a text that belongs to the longer version), commentators of theCS began to freely read this idea into some passages in chaptersthat had already been present in the shorter version, and terms likebāhyā dhyātma- , sab āhyādhyātma- , or the like then came to be in-serted into the shorter version, too.

[4] The addition of the last half of chapter 50 and of chapter51 to the shorter version can be dated between Jayabhadra andKambala because Kambala, unlike Jayabhadra, comments on thelast half of chapter 50 of the CS, although very brie y. Both Jaya-bhadra and Kambala very likely lived before the compilation of theVajra ḍākatantra , which was likely composed around or after thelate ninth century.

7 One could object that the commentary on the CS by Bhavyak ī rti, whois clearly one of the later commentators, also does not mention parts thatare not contained in the shorter version. However, this does not invalidatemy hypothesis because Bhavyak ī rti’s commentary very closely followsJayabhadra’s, as Gray also points out (p. 22).

8/19/2019 9010-8804-1-PB

11/41


2.2 Dates of the tantras of the Cakrasa ṃvara tradition that werecompiled after the CS

Gray’s approach to dating the tantras belonging to theCakrasa ṃvara tradition that were compiled after the CS can be summarized asfollows: The Abhidh ānottaratantra preservesŚaiva readings dat-ing to the ninth century, some of which are older than readingsfound in the CS. However, the compilation of the nal form ofthe Abhidh ānottaratantra cannot predate the CS because the CSmentions the Abhidh ānottaratantra under the titlecakrasa ṃvara .Furthermore, neither the Abhidh ānottaratantra nor the CS containtechnical Buddhist terminology relating to the perfection stage(niṣ pannakrama ), which became popular in and after the ninthcentury.8 By contrast, other tantras of theCakrasa ṃvara traditionsuch as theSaṃvarodayatantra , Vajra ḍākatantra , and Ḍā k ārṇa-va tantra contain Buddhist terminology relating to the perfectionstage. For these reasons, the Abhidh ānottaratantra may be datedto the eighth century, but de nitely not to before the CS. Manyof the tantras of theCakrasa ṃvara tradition appear to have beencomposed after the Abhidh ānottaratantra (p. 20.)

This analysis of the Abhidh ānottaratantra is, however, highlyproblematic. It is indeed true that the Abhidh ānottaratantra oftenpreservesŚaiva readings of the early medieval age, but the samecan also be said of other tantras of theCakrasa ṃvara tradition.Moreover, Gray’s idea that the Abhidh ānottaratantra does not con-tain technical terminology relating to the perfection stage is hardto accept. As I argued in 1999, the Abhidh ānottaratantra (like theVajra ḍākatantra , theSaṃ pu ṭ odbhavatantra , the Ḍāk ār ṇavatantra ,and other texts of theCakrasa ṃvara tradition that can be datedaround or after the late ninth century) clearly introduces the sub-tle-body system centered on the inner channels and inner circlesconnected with the doctrine of Four Blisses (catur ānanda ), evi-dently under the in uence of the Hevajra subtle-body system (fromaround the ninth century). The Abhidh ānottaratantra also gives

8 By “technical Buddhist terminology relating to the perfection stage,”Gray appears to refer to the psychosomatic subtle-body system centered onthe inner channels (nāḍī ), inner circles (cakra ), and the like.

8/19/2019 9010-8804-1-PB

12/41


instructions on psychosomatic meditation based on the subtle-bodysystem, in which the inner re or light and the inner ambrosia arevisualized to move inside and outside of the practitioner’s body.Furthermore, the Abhidh ānottaratantra contains instructions ofpsychosomatic meditation that are closely related to Lūy ī pāda’s

Mah ā yoga system and Kṛṣṇācārya’s Olicatu ṣṭ aya or Ālica tuṣṭ aya system, which were regarded as instructions to the perfection stagein Lūy ī pāda’s and Kṛṣṇācārya’s schools, respectively.9 It shouldalso be noted that the Abhidh ānottaratantra gives instructions onthe internal Herukamaṇḍ ala. One of these can be regarded as a de-veloped version of instructions given in Lūy ī pāda’sCakra saṃ var ā-bhisamaya (presumably last half of the ninth century), theSaṃ-

pu ṭ od bhavatantra , the Vajra ḍākatantra (both probably late ninthto tenth century), and several other ritual or meditational texts (i.e.,vidhi or sādhana texts) belonging to theCakrasa ṃvara tradition.(The nal section of this paper contains a discussion on the internalHerukamaṇḍ ala taught in these scriptures.)10

It is therefore unlikely that the date of the Abhidh ānottaratantra is as early as Gray believes, and it is impossible to maintain hisclear dividing line between the date and contents of the Abhidh ā-not ta ratantra and those of theSaṃ pu ṭ odbhavatantra , theVajra ḍā-ka tantra , and other tantras of theCakrasa ṃvara tradition. Gray’sintention, although not clearly stated, may be to say that the old-est parts of the Abhidh ānottaratantra were compiled in the eighthcentury, and that these refer to the passages that have parallelsin the CS, which does not contain technical Buddhist terminol-

9 For details on the teachings of the Abhidh ānottaratantra in relation toKṛṣṇācārya’s and Lūy ī pāda’s systems, see Sugiki 1999 and 2007. The con-tents of Lūy ī pāda’s Mah ā yoga system were already analyzed by MunenobuSakurai in an earlier paper, although he did not mention the textual rela-tionship between Lūy ī pāda’s works that teach the Mahāyoga system andthe Abhidh ānottaratantra (Sakurai 1997). Draft editions of two passagesthat explain the subtle body system and psychosomatic meditation based onit from the Abhidh ānottaratantra are provided in Sugiki 2007. Since thisbook may be diffi cult to access from outside Japan, these passages are pre-sented in an appendix to the present paper. 10 For details on the historical development of the internal Herukamaṇḍ ala , see Sugiki 2003b and 2007.

8/19/2019 9010-8804-1-PB

13/41


ogy concerning the perfection stage. Even if that is so, a similarproblem occurs as in his suggestion to date the CS in the eighthcentury. There also remains the question whether all parallelpassages of the CS found in the other tantras of the same tradi-tion (such as theSaṃ pu ṭ odbhavatantra , the Vajra ḍākatantra andothers) can be determined as being later than those found in the

Abhi dh ā nottaratantra . (As I mentioned above, readings that canbe considered to be old and early-medievalŚaivic are also foundin those tantras.) There is currently simply no conclusive evidencethat proves the Abhidh ānottaratantra existed in the eighth century.

As in the case of the CS, there appear to have been several stag-es in the compilation of the other tantras of theCakrasa ṃvara tra-dition. There also appear to have been complex mutual referencesto texts between the compilers of those tantras. Finally, in order tocarefully develop a plausible hypothesis on the dates of the tantrasbelonging to theCakrasa ṃvara traditions, we must also take therelationship betweenŚaiva and Buddhist tantras into consideration – and this, as mentioned above, is a point that Gray unfortunatelyneglects.

3 Gray’s translation of the CS and the materials used for it

3.1 Problems in the selection of materials

A Sanskrit edition of the CS, together with Bhavabhaṭṭa’s commen-tary, was published by Pandey in 2002 (henceforth CS-P). Althoughthis edition should be respected as a pioneering achievement, it isvery problematic, as many scholars in this eld have pointed out;Gray also accurately shows problems in Pandey’s edition. For hisown critical edition of the CS (that awaits publication), Gray usedthree Skt mss of the CS, which Pandey also used: an old palm-leaf ms owned by the Oriental Institute in Vadodara (accession no.13290), and two recent copies of it. Correctly recognizing that thelatter are copies of the palm-leaf ms., Gray uses the palm-leaf msas the main basis for his edition and translation.

In addition to these manuscripts, Gray also made use of othertexts and supporting materials: Tibetan translations of the CS, Skt

8/19/2019 9010-8804-1-PB

14/41


mss and eds, as well as Tibetan translations of eleven Indian com-mentaries, some indigenous Tibetan commentaries, as well as Sktmss and the Skt ed of the Abhidh ānottaratantra . Among the com-mentaries, he frequently favors Jayabhadra’sCakrasa ṃvarapañjik ā and Bhavabhaṭṭa’s Cakrasa ṃvaraviv ṛ ti, and also, in some pas-sages, Vajrapāṇi’s Laghutantra ṭī k ā. He also attaches much impor-tance to Kambala’s Herukas ādhananidhi and V ī ravajra’sPad ār-tha prak āś ik ā. These commentaries are favored or considered tobe important for the following reasons: (a) Skt mss or Skt eds ofJayabhadra, Bhavabhaṭṭa, and Vajrapāṇi’s commentaries are avail-able. (b) Jayabhadra’s commentary is the oldest among the surviv-ing commentaries of the CS. (c) Kambala’s commentary is alsorelatively early. (d) Many later commentators of the CS rely onJayabhadra’s or Kambala’s commentaries. (e) Bhavabhaṭṭa’s com-mentary quotes many words and phrases of the CS, although hesometimes emends these in the act of quoting. (f) Kambala’s andV ī ravajra’s commentaries give detailed explanations of rituals thatare described only brie y in the CS. And, as I mentioned earlier,the reason for using the Abhidh ānottaratantra is that it preservesold readings and contains many passages that have parallels in theCS and is hence quite useful for recovering material that is missingfrom the extant Skt mss of the CS.

These materials, however, do not suffi ce for a fully critical edi-tion and translation of the CS. Most of the supporting materials arenew paper mss or Tibetan translations. Gray did not use two olderpalm-leaf mss of Jayabhadra’s commentary, which preserve olderand better readings than the new paper mss he used.11 Neither didhe use a Skt ms of Kambala’s commentary, which is also an oldpalm-leaf ms,12 but only used a Tibetan translation of this commen-tary. Finally, Gray did not make eff ective use of Skt mss or Skt edsof Buddhist andŚaiva texts that have parallel or similar passages,

11 A draft-version of the Skt ed of whole text of Jayabhadra’s commentarybased on these two palm-leaf mss has been published in Sugiki 2001. I planto publish the nalized edition in the near future. 12 I have prepared an as yet unpublished Skt ed of whole text of Kambala’scommentary based on this palm-leaf ms.

8/19/2019 9010-8804-1-PB

15/41


such as theVajra ḍākatantra ,13 theSaṃ pu ṭ odbhavatantra , theŚaivatantras which Alexis Sanderson mentions in his series of papersthat analyze textual relationships betweenŚaiva Vidyāp īṭha tantrasand Buddhist Yogin ī tantras (such as theSiddhayoge ś var ī mata , the

Jayad rathay āmala , the Brahmay āmala , and theTantrasadbh āva),14 the ŚaivaV īṇāś ikhatantra ,15 and other related texts.16 These mate-rials are not only truly helpful in creating a critical edition andtranslation of the CS; they are actually indispensable for recoveringmissing passages of the CS that cannot be reconstructed, or onlyin an unsatisfactory manner, from the Abhidh ānottaratantra andJayabhadra’s, Bhavabhaṭṭa’s, and Vajrapāṇi’s commentaries. Let uslook at some relevant cases.

13 Draft versions of the Skt eds of chapters 1, 7, 8, 14, 18, 22, 36, 38, 42,44, 48 of the SanskritVajra ḍākatantra have been published in Sugiki 2002,Sugiki 2003a, and Sugiki 2008. I have also prepared as yet unpublished drafversions of other chapters of this tantra. 14 Sanderson 1995, Sanderson 2001.

15 The V īṇāś ikhatantra ’s instruction on thevet ālas ādhana (Skt ed,190cd–193) contains a passage that is very similar to or identical with thoseof the CS, theVajra ḍākatantra , and the Heruk ābhyudayatantra . See Sugiki2008 for details and references. I express my heartfelt thanks to the reviewerof this article that was published in the journalTantric Studies (The Centerfor Tantric Studies, University of Hamburg) who suggested that I check care-fully theV īṇāś ikhatantra ’s passage in question before submitting the nalversion of paper. I would like to add here that the versemah āś aṅkhamaya ṃ kury ād athav ā kacchapasya tu of theV īṇāś ikhatantra (Skt ed, 113cd) isalso a parallel of the CS’smah āś aṅkhamaya ṃ kury ād abhedya ṃ kaccha-

pasya tu (Skt ms, 25a3–a4). 16 For example, theCatu ṣ pīṭ hatantra , the Saṃvarodayatantra , the Ḍāk ār ṇavatantra , Lūy ī pāda’sCakrasa ṃvar ābhisamaya , and Kṛṣṇācārya’sCakrasa ṃvaras ādhana . Lūy ī pāda’s Cakrasa ṃvar ābhisamaya (= Bha ga-vad abhisamaya ) have been published in Sakurai 1998. The Skt ed of thewhole text of Kṛṣṇācārya’s Cakrasa ṃvaras ādhana has been published inSugiki 2000, which also contains a list of parallel passages found in the CSand Kṛṣṇācārya’sCakrasa ṃvaras ādhana.

8/19/2019 9010-8804-1-PB

16/41


3.2 Textual problems and problems of translation

Gray recovers pāda 3a of chapter 26 (whose folia are missing fromthe Skt mss of the CS) as ‘taṃ d ūt īṃ sarvasiddhida ṃ’ from one ofthe two paper mss of Jayabhadra’s commentary (p. 265, note 4).17 He mentions that the text in question is improperly declined, asBhavabhaṭṭa notes in his commentary (tam iti t āḥ , d ūt ī d ūtaya ḥ,CS-P: 483) (p. 265, note 4), and translates “These messengersbestow all powers” (p. 265). However, the text can be recoveredfrom the two palm-leaf mss of Jayabhadra’sCakrasa ṃvarapañjik ā and the two palm-leaf mss of theVajra ḍākatantra as ‘t āṃ d ūt īṃ sarvasiddhid āṃ ,’ ‘that female messenger bestowing all supernatu-ral eff ects [or accomplishments].’ Jayabhadra comments that ‘t āṃ d ūt īṃ sarvasiddhid āṃ’ should be read ast ā d ūtyaḥ [sarva sid dhi-d āḥ , TS ], which means ‘those female messengers bestow all super-natural eff ects [or accomplishments].’18 Though not optimal, thisis at least clearer and more natural than Gray’staṃ d ūt īṃ sarva-siddhida ṃ, and is likely to be the older version because the sourcesare older than those used by Gray. The whole verse 3 may be re-covered from the Brahmay āmala , Jayabhadra’s commentary, Kam-bala’s commentary, theVajra ḍākatantra , and Bhavabhaṭṭa’s com-mentary as follows:t āṃ d ūt īṃ sarvasiddhid āṃ dar ś an āt spar ś an āttath ā / cumban āvag ūhan ān nitya ṃ (metrically bad) yogap īṭ heviś eṣata ḥ //.19

17 Pandey’s reconstruction of the verse in question is ‘taṃ d ūt ī tusattv ārthasiddhida ṃ,’ which except for ‘tu’ is based on the text quoted inBhavabhaṭṭa’s commentary. 18 Jayabhadra’sCakrasa ṃvarapañjik ā, Skt ed, 26.1. TheVajra ḍākatantra ,my unpublished Skt ed, 34.4a. 19 t āṃ d ūt īṃ sarvasiddhid āṃ ] J; taddravya ṃ sarvad ā siddha ṃ – BY.t āṃ d ūt īṃ sarv ārthasiddhid āṃ – VḌT. taṃ d ūt ī sattv ārthasiddhidam –Bh. taṃ d ūt ī tu sattv ārthasiddhida ṃ – CS-P. :dar ś an āt spar ś an āt tath ā ]em.; dar ś an āt spar ś abhak ṣaṇāt – BY.dar ś an āt spar ś an āt – J.dar ś ana ṃ spar ś ana ṃ tath ā – VḌT, Bh, CS-P. :cumban āvag ūhan ān nitya ṃ ] J,VḌT; cumban ād g ūhan āc caiva – BY.cumban āvag ūhan ām (ityādi) – K.cumbane (-tyādi) – Bh.cumbana ṃ gūhana ṃ nityaṃ – CS-P. : yogap īṭ heviś eṣata ḥ ] J;ś ivap īṭ he vi ś eṣata ḥ – BY. yogap īṭ havi ś eṣata ḥ – K and VḌT(very likely a corruption of yogap īṭ he vi ś eṣata ḥ). yogap īṭ ham andviś eṣata

8/19/2019 9010-8804-1-PB

17/41


As for verse 2ab of chapter 27, whose folia are also lost, Gray fa-vors the reading given in one of the two paper mss of Jayabhadra’scommentary and reconstructs ‘gr āme gr āme vrajanti ca d ūtayo (r ū pa lak ṣaṇam, TS ),’ translated as “The messengers travel fromtown to town. [As for their] physical characteristic[s], …” (p. 271, n4). Pandey, on the other hand, recovers this verse as ‘gr āme gr āmevrajan tasya d ūtayo r ū palak ṣaṇam .’ Indeed, one of the two palm-leaf mss of Jayabhadra’sCakrasa ṃvarapañjik ā20 supports Gray’sreconstruction. However, Pandey’s reconstruction is better thanGray’s because it is con rmed by older sources. It is fully con rmedby the other palm-leaf ms of Jayabhadra’sCakrasa ṃvarapañjik ā,the palm-leaf ms of the Abhidh ānottaratantra , and Bhavabhaṭṭa’sCakra saṃ varaviv ṛ ti.21 The corresponding verse in the Skt ms ofthe Brahmay āmala reads gr āme gr āme vratan tasya deva t ā r ū-

pa lak ṣaṇaṃ,22 which is closer to Pandey’s reconstruction than toGray’s gr āme gr āme vrajanti ca . The text Pandey reconstructedmeans: ‘He (= the practitioner) travels (vrajan tasya : vrajan isvraja ṃ) from village to village. [In these villages, the] female mes-sengers [show their] physical characteristic[s to him].’ In this con-text, the one who travels is not a messenger but a practitioner. Thecomments on this verse by Jayabhadra, Kambala, and Bhavabhaṭṭaalso support this interpretation.

Gray translates pāda 6a of chapter 41, whose folia are also lost,as “[They are:] in Kulutā (better: Kulatā, TS)23 and [Maru], …’.24

(iti ) – Bh. yogap īṭ haṃ viś eṣata ḥ – CS-P.20 Jayabhadra’sCakrasa ṃvarapañjik ā, Skt ed 27.1 and the footnote there.

21 Jayabhadra’sCakrasa ṃvarapañjik ā, Skt ed, 27.1. Bhavabhaṭṭa’s Ca-kra saṃvaraviv ṛ ti, Skt ed, 488. 22 The Brahmay āmala , Skt ms (NGMPP A42/6). 326b3. Skt eds, San-

derson 2006: 22 (gr āme gr āme vrata ṃ tasya devat ār ū palak ṣaṇam); Hatley2007: 180 (gr āme gr āme vrata ṃ tasya devat ār ū palak ṣaṇam). 23 ‘Kulatā’ is more common in BuddhistCakrasa ṃvara scriptures thanGray’s ‘Kulutā.’ See also Bhavabhaṭṭa’s comment on this pāda , kulat ā yāmityādin ā (CS-P, 547). 24 Pandey’s reconstruction of this text is as follows:kulat ā yāṃ vivikte ca .However,vivikte is not attested in any surviving Sanskrit sources that areclosely related to this pāda .

8/19/2019 9010-8804-1-PB

18/41


He put ‘Maru’ in brackets because “the Sanskrit text here is notpreserved, and the Tibetan translations list the variantmgon pa (PM 239a, SL 130b), which is unattested elsewhere.” (p. 330, n.10.)25 However, for the word in question, we nd ‘ara ṇ ya’ in Jaya-bhadra’s commentary, which is very likely derived from the word‘ara ṇ yeś e’ that appears in a parallel passage found in theŚaivaTantra sadbh āva .26 The passage that includes pāda 6a provides thearchaic list ofCakrasa ṃvara holy sites that is derived from a cor-respondingŚaiva list, such as the one found in theTantrasadbh āva .We may thus recover pāda 6a from theTantrasadbh āva and fromJayabhadra’s commentary as ‘kulat ā yām ara ṇ ye ca ,’ and the trans-lation should be ‘[They are:] in Kulatā, Araṇya, …’

Gray translates verse 10ab of the same chapter as “The sixyogin ī s are in Kulutā (better: Kulatā, TS), and the six mothers arein the land of Maru” (p. 331–332). He appears to have followedPandey’s reconstruction ‘ṣaḍ yoginya ḥ kulat ā yāṃ marude ś e ṣaḍ mā ta r āḥ.’ However, the last pāda must be ‘marude ś e ca m ātara ḥ,’(notṣaḍ but ca ,) which can be recovered from the palm-leaf mssof theTantrasadbh āva (ara ṇ yeś e ca m ātara ḥ), Jayabhadra’s com-mentary (mātara ḥ [: no ṣaḍ ]), Kambala’s commentary (marude ś eca m ātara ḥ), theVajra ḍākatantra (marude ś e ca y ā mātar āḥ), andBhavabhaṭṭa’s commentary (mātar ā iti [: no ṣaḍ ]).27 The Tibetantranslations of the CS (mya ngam yul na ma mo rnams ) also sup-port this reconstruction, and no old sources support Pandey’s andGray’s ‘marude ś e “ ṣaḍ ” m ātar āḥ.’ The translation should there-fore be corrected to: ‘the mothers are in the land of Maru’. It is

25 Gray goes on to state that “several verses down, however,marude ś e isattested by Bhavabhaṭṭa (CS-P, 548).” However, this ‘marude ś e’ is a quota-tion not from the passage in question but from another passage of the CS

(41.10b). 26 Jayabhadra’sCakrasa ṃvarapañjik ā, Skt ed, 41.2 (ara ṇ yaṃ maru bh ū-miḥ). TheTantrasadbh āva , Skt ed, Sanderson 1995: 100, n. 20 (kulūt ā yāmara ṇ yeṣe). 27 TheTantrasadbh āva , Sanderson 1995: 100, n. 20. Jayabhadra’sCa kra-saṃvarapañjik ā, Skt ed, 41.3. Kambala’s Herukas ādhananidhi , my unpub-lished Skt ed (: Skt ms, 70a4). TheVajra ḍākatantra , Skt ed, 18.4. Bhava-bhaṭṭa’s Cakrasa ṃvaraviv ṛ ti, Skt ms, 127a4 (: Skt ed, 488).

8/19/2019 9010-8804-1-PB

19/41


very likely that Gray, as well as Pandey, misread Bhavabhaṭṭa’s‘ṣaḍ yoginya ity ādi / … vajrav ār āhī yāminyādaya ḥ ṣaṭ / mātar ā itisapta māt ṛ r ū pāḥ marude ś e /’ (Skt ms, 127a2–a4).28 But the wordṣaṭ in this passage is not the number modifyingmātar āḥ; it refersto ṣaḍ yoginya ḥ, i.e., the six yogin ī s beginning with Vajravārāh ī and Yāmin ī .29

There are also cases where Gray creates unnatural translations,some of which appear to have been caused by unnecessary or in-correct emendations of manuscript readings. Some examples fol-low. The Vadodara ms of the CS 4a5 (: Skt ed, 3.17ab) readssarv ā kiṅkar ī tasya s ādhakasya na sa ṃś aya ḥ , 30 ‘all [theḍākin ī s] are fe-male servant[s] of that adept; no doubt.’ Gray translates: “There isno doubt regarding anything done by that adept.” (p. 175.) He doesnot explain how he emended the Sanskrit text.

A further example is CS 4b1–b2, where the Skt ms reads:eṣa yogavara ḥ ś reṣṭ haḥ sarvayoge ṣu cottama ḥ / yaḥ k āṅk ṣiṣ yate

ka ś cit sa dev āsuram ānuṣān / abhibh ū ya gami ṣ yaty atra ma ṇḍ ale yo’bhi ṣiktaḥ / sarvatantroktas ādhaka ḥ / 31 ‘This is the supreme yoga, the most excellent, and it is the highestamong all yogas. Anyone who wishes [this supreme yoga] will go,conquering gods, titans, and men. [The one] who was initiated in thismaṇḍ ala is the adept of what is taught in all tantras.’

Gray translates: “This yoga is the most excellent, the highest amongall yogas, which can kill anyone, gods, titans or men. The adeptwho has been taught all tantras, and who has been initiated in the

28 Pandey’s edition of this line readsṣaḍ yoginya ity ādi / … / vajrav ār āhī yāminyādaya ḥ ṣaḍ mātar ā iti / saptam āt ṛ r ū pāḥ marude ś e / (CS-P, 548).

29 See also Jayabhadra’s comment onṣaḍ yoginya ḥ andmātara ḥ: ṣaḍ yo-ginyo vajrav āhyādica ṇḍ ik ānt āḥ // mātara ḥ k āk āsyādyāḥ // [Skt ed, 41.3] 30 Pandey’s edition readssarv āḥ kiṅkar ī s tasya s ādhakasya na sa ṃ-ś aya ḥ. 31 For yaḥ k āṅk ṣiṣ yate ka ś cit , which is supported by Bhavabhaṭṭa’s com-mentary, Kambala’s commentary reads yaḥ k āṅk ṣiṣ yati nitya ṃ (Skt ms, 11b6– b7). Foratra ma ṇḍ ale yo ’bhi ṣiktaḥ sarvatantroktas ādhaka ḥ, Kambala’scommentary readsatra ma ṇḍ al ābhiṣiktaḥ sarvatantroktas ādhana ḥ (Sktms, 11b7). These variant readings are also acceptable.

8/19/2019 9010-8804-1-PB

20/41


maṇḍala, will go forth, conquering.” (p. 176.) Again, he does notadduce his version of the text.

The Skt ms of CS 4b7–5a1 readstato jñ ātvā bh āvayen nitya ṃ siddhis tath āgatavaco yath ā , ‘therefore, should he know and al-ways visualize [themaṇḍ ala ], [there will be] accomplishment (orsupernatural eff ect), as taught by the Tathāgata.’ Gray translates,again without adducing the text: “Knowing thus, one should alwaysmeditate on the powers taught by the Tathāgata.” (p. 180.)

The Skt ms of the CS 26a3–a4 (: Skt ed, 34.7) readseṣatecakrodbh āsaṃ kury ād yath ākrama ṃ sarva sid dhi pra sā dha-kaḥ , 32 ‘he seeks the radiance of the wheel. Should he practice [thiswheel] in due succession, [he] accomplishes all supernatural eff ects(or accomplishments).’ Gray readseṣate cakrodbh āsaṃ as eṣa tecakrod bh āsaṃ and emends toeṣa tricakrodbh āsaṃ by misread-ing Bhavabhaṭṭa’s comment on the wordeṣate 33 and by favoringthe reading of one of the paper mss of Jayabhadra’s commentary,which is not attested in other materials; he then translates as “Heshould successively make the Three Wheels radiant. This is the ac-complishment of all powers.” (p. 311 and n. 15 on that page.)

4 Origin myths of Heruka and his ma ṇḍala

Heruka is the highest deity of theCakrasa ṃvara tradition. Hence,researching the origin myths of Heruka and hismaṇḍ ala has been amain concern of scholars studying this tradition. Gray unpacks thehistory of Indian versions of this myth mainly on the basis of theSarvabuddhasam ā yogatantra , theSarvatath āgatatattvasa ṃgraha ,and Indrabhūti’s commentary on the CS.

32

For cakrodbh āsaṃ, Kambala readscakranirdi ṣṭ aṃ. For the wholeline, Bhavabhaṭṭa reads eṣate cakranirdi ṣṭ aṃ sarvasiddhiprad ā yakam .(CS-P, 528. I corrected Pandey’seṣa te into eṣate .) Pandey’s edition readseṣa te cakranirdi ṣṭ aṃ sarvasiddhipras ādhakam / cakrodbh āsaṃ tath ā kury ād yath ākarm ānur ū pata ḥ // 33 Following Pandey’s edition (CS-P, 528), he reads Bhavabhaṭṭa’s eṣatemṛ gayate , an explanation of the meaning of the wordeṣate , as eṣa temṛ gayate , which makes less sense.

8/19/2019 9010-8804-1-PB

21/41


According to Gray, the myth in theSarvabuddhasam ā yogatantra described the birth of Heruka as a generation through yogic heatvia controlled breathing. Heruka burns the triple world and Hindudeities such as Rudra, Mahādeva, Viṣṇu, Brahmā, etc., reducesthem to ashes, and restores or reanimates them. Although theseHindu deities are roasted in Heruka’s process of cosmic cleansing,this does not mean that they are vili ed. They are rather portrayedas victims of a cosmic disorder in which Māras (the traditionalBuddhist villains) are active and which is the result of the inevi-table process ofkarmic conditioning. However, the version of themyth that eventually came to predominate portraysŚaiva deities asthe perpetrators of cosmic disorder. The myth in theSarva ta thā-ga ta tattvasa ṃgraha (which is a story of Vajrapāṇi’s subjugation ofMahā deva and Mahādeva’s conversion to Buddhism) and the mythfound in Indrabhūti’s commentary on the CS are examples of thisversion. The myth in Indrabhūti’s commentary is especially impor-tant because many Tibetan versions of Heruka’s origin myth con-tain the story of the origin of theCakrasa ṃvara Herukamaṇḍ ala,and Indrabhūti’s commentary on the CS is the only known Indiantext that presents a complete version of the myth. It is a likelysource of the Tibetan versions.

Gray provides a translation of the whole text of the myth as it isintroduced in Indrabhūti’s work and analyzes its content, using theTibetan versions in support for his analysis. He then argues that themyth represents the adoption of non-Buddhist elements and thatthese elements are at the same time subordinated within a Buddhiscosmic hierarchy. The myth is therefore clearly a reaction to Hindutri pur āntaka myths.

Gray’s portrayal of the history of Heruka myths is acceptable,and it is bene cial to those who are interested in this topic. Hisdiscovery that the myth in question occurs in Indrabhūti’s work isno doubt a great contribution to the study of theCakrasa ṃvara tra-dition; I myself had completely overlooked it. However, he wouldhave been able to paint a fuller picture by also considering the ver-sion of this myth that is found in Nāropāda’s ’Khor lo bde mchog

8/19/2019 9010-8804-1-PB

22/41


gi rnam par ’phrul pa (*Cakrasa ṃvaravikurva ṇa);34 Nāropāda’sversion is as likely a source of the Tibetan representations of thismyth as Indrabhūti’s. Although some descriptions diff er in thetwo versions, Gray’s analysis of Indrabhūti’s version can also beapplied to Nāropa’s: both versions represent the adoption of non-Buddhist elements and subordination of these elements within aBuddhist cosmic hierarchy. However, Nāropāda’s version shouldalso be considered because its explanations of the origination ofthe Herukamaṇḍ ala and of the subjugation of non-Buddhist di-vinities are more detailed than those given in Indrabhūti’s version.In comparison to the latter, Nāropāda’s version lends itself morenaturally to the interpretation that the myth represents the adoptionof non-Buddhist elements, and their subordination within Buddhistcosmic hierarchy, and need not be complemented with informationtaken from Tibetan versions.

Let us examine the contents of Nāropāda’s version brie y. 35 Thebeginning scene of the myth in Nāropāda’s version can be sum-marized as follows:

During the era of Kali, (1) a deity from the Thirty-three Heaven, (2)Gandharva, (3) the chief Yakṣa and (4) his attendant (g’yog ), (5) thechief Rākṣasa and (6) his attendant, (7) the chief Nāga and (8) his at-tendant, and (9) the chief Asura and (10) his attendant, transformingthemselves into twenty-four Bhairavas or ‘awful divinities’ (drag po ),each took a consort; they then captured twenty-four sites located onthe Jambū continent: (1’) four sites classi ed as pīṭ ha , (2’) four sitesclassi ed asupap īṭ ha , (3’) two sites classi ed ask ṣetra , (4’) two sitesclassi ed asupak ṣetra , (5’) two sites classi ed aschandoha , (6’) twosites classi ed asupacchandoha , (7’) two sites classi ed asmelā paka ,(8’) two sites classi ed as upamel ā paka , (9’) two sites classi ed asś ma śāna , and (10’) two sites classi ed asupa ś ma śāna , respectively.The four-bodied, four-natured, and four-faced Mahādeva, who resideson the summit of Mt. Meru with his four goddesses and his four secretgoddesses, became the lord of these Bhairavas at their request. They

34 This work is preserved in the Peking edition of Tibetan Tripiṭaka, OtaniUniversity catalogue 4628. 35 For further details, see Sugiki 2006, 2007, and 2009.

8/19/2019 9010-8804-1-PB

23/41


terrorized the people living on the Jambū continent and brought thisworld to a state of degeneration.

The cosmic disorder was primarily a result of the inevitable proc-ess ofkarmic conditioning (i.e., the arrival of the Kali era). In thisera of cosmic disorder,Śaiva divinities in Bhairava forms changedthe Jambū continent – with the twenty-four sites in its center – intoa Śaivamaṇḍ ala , took control of it, and caused it to be in an un-wholesome situation. The myth subsequently explains how theHerukamaṇḍ ala originated and howŚaiva divinities were subju-gated through enjoyment (longs spyod pa ), dissolution (thim pa ),and control (dbang du byas pa ), which are only brie y mentionedin Indrabhūti’s version:

Unhappy about this unwholesome situation, the Samyaksaṃbuddhacame down from the Akaniṣṭha Heaven to the summit of Mt. Meruin order to subjugate these awful divinities. The Samyaksaṃbuddhamanifested himself as an experiential-body divinity (longs sku ) withone face, two arms, a white complexion, and the nature of Vajradhara,and he took Samantabhadr ī (kun tu bzang mo ) as his consort. He thentransformed himself into the divinity named Heruka, who had a darkcomplexion, four faces, and twelve arms, and who took Vajravārāh ī as his consort. Subsequently, Heruka and Vajravārāh ī created twen-ty-four pairs of male and female heroic divinities who came to bethose of the triple wheels (i.e., the origination of theCakrasa ṃvara Herukamaṇḍ ala ). Each stage in the entire process of the manifesta-

tion of the Herukamaṇḍ ala as described above re

ected a particularcharacteristic of each of ve Tathāgatas (i.e., Vairocana, Amitābha,Ratnasaṃbhava, Amoghasiddhi, and Akṣobhya).36

These Buddhist divinities conquered Mahādeva and his retainers,and, subjugating them, (1) made them objects of enjoyment throughsexual assemblage and by making ornaments of their bones (= enjoy-ment), (2) eff ected the disintegration and incorporation of their con-sciousnesses (= dissolution), and (3) took control over their bodies,words, and minds (= control). In these steps of the process, the maleand femaleŚaiva divinities were subjugated along the paths of angerand sexual passion respectively. Assimilating the essence of theŚaiva

36 The text does not expound the particular characteristic of each of the ve Tathāgatas. They commonly symbolize the ve kinds of gnosis and the ve aggregates.

8/19/2019 9010-8804-1-PB

24/41


divinities by the incorporation of their consciousnesses and the orna-mentation of their bones, the Buddhist divinities then took over thetwenty-four sites on the Jambū continent. Heruka then created four

female divinities as gate-keepers, and created four other female di-vinities in addition (i.e., the eightḍākin ī s of thesamayacakra portionof theCakrasa ṃvara Herukamaṇḍ ala ). These eight female divini-ties conquered and subjugated Kinnaras of both sexes found at siteslocated in eight directions surrounding the above twenty-four sitesformerly controlled by theŚaiva divinities.Although the twenty-four pairs of Buddhist divinities established

themselves at the twenty-four sites, they had not yet attained Buddhistenlightenment. Hence, they ascended Mt. Meru. Asked by them togive instructions on Buddhist truth, and receiving their various off er-ings and hymns, the Saṃyaksaṃbuddha at the summit of Mt. Meruproduced the various tantras of theCakrasa ṃvara tradition.

As described in this myth, the summit of Mt. Meru and the Jambū continent were taken over by Buddhist divinities, and the goodBuddhistdharma (i.e., tantras of theCakrasa ṃvara tradition)was brought to this world. The Samyaksaṃbuddha’s act of cos-mic cleansing was completed. It should be noted that the Buddhistdivinities did not sweep away theŚaiva elements in this process,but instead assimilated theŚaiva essence into their cosmos. Thisis demonstrated by Buddhist divinities’ taking over of theŚaivamaṇḍ ala consisting of Mt. Meru and the twenty-four sites, whichresulted in its change into Buddhist Herukamaṇḍ ala , and by theenjoyment, dissolution, and control process, through which thebodies, words, and minds of theŚaiva divinities became constitu-ents of their Buddhist counterparts. These processes therefore rep-resent the adoption of non-Buddhist elements and their subordina-tion within a Buddhist cosmic hierarchy.

5 The structure and functions of the Triple Wheel ma ṇḍala

Tantras belonging to theCakrasa ṃvara tradition describe manyvarieties ofmaṇḍ ala s. Among them, the most popular and widelyused for practice is the Herukamaṇḍ ala consisting of ve con-

8/19/2019 9010-8804-1-PB

25/41


centric wheels, i.e., the Great Bliss Wheel (mah āsukhacakra ),37 theMind Wheel (cittacakra ), the Speech wheel (vākcakra ), the BodyWheel (k ā yacakra ), and the Pledge Wheel (samayacakra ). TheGreat Bliss Wheel, on which Heruka, his consort Vajravārāh ī , fourḍākin ī s, and four skull-bowls are depicted, is located at the centerof thismaṇḍ ala .38 The Great Bliss Wheel is surrounded by threeconcentric wheels, the Mind, Speech, and Body Wheels. Thesethree are collectively called ‘the Triple Wheel’ (tricakra ), and twen-ty-four holy sites and twenty-four coupled deities (i.e., twenty-fourpairs ofḍākin ī and vī ra ) assigned to these holy sites are depictedon them (i.e., eight holy sites with eight couples on each wheel☓ 3= twenty-four holy sites with twenty-four couples.) They are sur-rounded by the Pledge Wheel, on which eightḍākin ī s reside.39 ThisHerukamaṇḍ ala can be roughly described as having two forms,external and internal, and the deities and holy sites that constitutethismaṇḍ ala symbolize traditional Mahāyānic or Indian conceptssuch as the triple world (sky, earth, underground), the three bodiesof the Buddha (trik ā ya), the ve elements ( pañcabh ūta), the tenspiritual levels (da ś abh ūmi), the ten perfections (da ś ap āramit ā),the ten kinds of gnosis (da ś ajñ āna ), the eight vows (aṣṭ asamaya ),the thirty-seven conditions that contribute to awakening (sapta-triṃ ś ad bodhip āk ṣikadharma ), and so forth.40 The CS per se doesnot introduce the fully developed form of the Herukamaṇḍ ala, butexplains its prototypical form,41 as Gray duly notes (p. 55, 58).

37 Gray names this circle ‘gnosis wheel’ ( jñ ānacakra ) (p. 55), but it is morecommonly called ‘great bliss wheel’ (mah āsukhacakra ) in the BuddhistCakrasa ṃvara tradition. 38 Heruka and Vajravārāh ī are situated at the center of the Great BlissWheel. They are surrounded by fourḍākin ī s (i.e.,Ḍākin ī , Lāmā, Khaṇḍarohā,and Rūpiṇī ) and four skull bowls in the cardinal directions and quarters, re-spectively. The four skull bowls are not explicitly mentioned in the CS. 39 The four gate-keeperḍākin ī s Kākāsyā, Ulūkāsyā, Śvānāsyā, andŚū ka-rāsyā reside in the four directions and four otherḍākin ī s, Yamadāḍh ī (Ya-madāḍh ī is more common than Gray’s Yamadāh ī , p. 55 n. 169), Yamadūt ī ,Yamadaṃṣriṇī , and Yamamathan ī , are in the four quarters. 40 For details of the structure and symbolism of this Herukamaṇḍ ala , seealso Sugiki 2003b, 2007, and 2009. 41 For details of the prototypical form of the Heruka maṇḍala in the CS,

8/19/2019 9010-8804-1-PB

26/41


After describing the structure of the Herukamaṇḍ ala as above,Gray focuses his analysis on the structure and function of theTriple-wheel part of the Herukamaṇḍ ala and its doctrinal contexts.Since the CS does not explain every detail in full, Gray further re-lies on other sources, in particular on the Abhidh ānottaratantra ,the Yogin ī saṃcāratantra , the Saṃvarodayatantra , Umāpatideva’sVajra vār āhī sādhana , Lūy ī pāda’s Bhagavadabhisamaya (= Cakra-saṃ var ā bhisamaya ), Atiśa’s Abhisamayavibha ṅga , Abhayā ka-ragupta’s Āmnā yamañjar ī , and Bu-ston’sbDe mchog nyung ngurgyud kyi spyi rnam don gsal .

I will now examine Gray’s portrayal of the Triple Wheel, i.e., thetwenty-fourCakrasa ṃvara holy sites beginning with Pull ī ramalayaand ending with Kulatā, and the coupled deities assigned to thetwenty-four sites. The examination will focus on two points: (1) themapping of the twenty-fourCakrasa ṃvara holy sites and (2) thedevelopment of systems of the twenty-four internal holy sites.

5.1 The mapping of the twenty-four holy sites

Gray explains the geographical locations of the twenty-fourCakra-saṃvara holy sites on the Indian continent on the basis of Bu-ston’sbDe mchog nyung ngu rgyud kyi spyi rnam don gsal (noteson pp. 329–333), and their remapping over Kathmandu Valley andTibetan and Mongolian areas on the basis of Abhayākaragupta’s Āmnā ya mañjar ī and some earlier studies on the topic (pp. 70– 71). The twenty-fourCakrasa ṃvara sites, which originally re-ferred to the twenty-four sites on the Indian subcontinent, wereremapped over areas outside India – such as Kathmandu Valley,Tibet, and Mongolia – during the process of transmission of theCakrasa ṃvara tradition from India to those outside areas. This in-terpretive exibility was an essential factor in the transformationthat the tradition had to undergo as it crossed regional boundaries.Abhayākaragupta’s de nition of the nature of theCakrasa ṃvara holy sites – any sites, including Tibet and China, where living hu-

see also Sugiki 2003b, 2007, and 2009.

8/19/2019 9010-8804-1-PB

27/41


man femaleḍākin ī s resided could be regarded asCakrasa ṃvara holy sites – functioned to legitimate their remapping.

Gray’s account manages to capture an important aspect of theexpansion of theCakrasa ṃvara tradition, but it deserves to be sup-plemented by a consideration of how Indian texts of theCakra-saṃ vara tradition prior to Abhayākaragupta discuss the mappingof theCakrasa ṃvara holy sites in India.42 While the instructionsgiven in most of these texts are fragmentary, Nāropāda’s Yul nyibcu bshi’i rgyu mtshan bstan pa 43 gives detailed instructions and isvery likely the most important Indian source for Tibetan versionslike Bu-ston’sbDe mchog nyung ngu rgyud kyi spyi rnam don gsal ,which Gray used.

Nāropāda identi es geographical locations of sites that are giv-en unnatural or obscure names by theCakrasa ṃvara scriptures:Himālaya is Mt. Kailāsa, Pretapur ī (also called Pretādhivāsin ī ) re-fers to the valleys located on the border between India and Tibet,Gṛhadevatā (which, as Sanderson argued, was originally a nameof the deity of the site Saurāṣra in theŚaiva Tantrasadbh āva) isLi yul, which may refer to Khotan. Suvarṇadv ī pa is located of thecoast of west India,44 but some say that it is in east China, andNagara refers to Laṅkāpura, the land ofr āk ṣasa , but some say thatit is an area around a monastery standing on the border of Kaśm ī raand northwest India.

However, other Indian sources give diff erent information onthe geographical locations and features of theCakrasa ṃvara sites, which indicates that there were diff erent maps of them. Forexample, Arbuda has been identi ed with Mt. Abu in modernRajasthan since it was mentioned in the Mah ābh ārata , but it isidenti ed with Takṣaśilā by Nāropāda. There are three diff erentdescriptions of the geographical location of Nagara according to

42 For details of the following analyses, see Sugiki 2006, 2007, and 2009. 43 This text is preserved in Peking edition of Tibetan Tripiṭaka, OtaniUniversity catalogue 4628 (the same catalogue number as Nāropāda’s’Khorlo bde mchog gi rnam par ’phrul pa mentioned above). 44 Generally, Suvarṇadv ī pa refers to the island in the ocean off the southtip of India, often Sri Lanka.

8/19/2019 9010-8804-1-PB

28/41


Nāropāda as mentioned in the previous paragraph; but accordingto theYogin ī jālatantra (and the two commentaries on the Heva jra-tantra by Kāṅhapāda and Ratnākaraśānti), Nagara refers to Pāṭa-liputra (east India). Finally, Nāropāda describes many of the twen-ty-four sites as sites whose center is formed by sacral stones such asstoneliṅga s of various shapes and stonedharmodaya s. (The stoneliṅgas conform in shape to the body parts which theCakra saṃ vara scriptures equate with external holy sites.) These stoneliṅga s anddharmodaya s are very likely to be a Buddhist recasting ofŚaivaś ivaliṅga s and yoni s. But many other authors, including Abhayā ka-ragupta, regard goddesses or living humanḍākin ī s as sacral cent-ers of the holy sites.

While the Indian compilers of texts belonging to theCakra saṃ-vara tradition attempted to pinpoint a speci c geographical loca-tion and to de ne a speci c feature for each individual site, they didnot always reach a consensus. It is therefore likely that the locationsand features of these sites were exible rather than xed. This sug-gests that the list of names of the twenty-fourCakrasa ṃvara holysites in India was rather idealized, serving as a symbolic frame-work along which individual sites were arranged, to a certain ex-tent, according to the respective compiler’s preference. This is alsosupported by other facts. First, as Sanderson pointed out, the list oftwenty-fourCakrasa ṃvara sites was produced in the process of theBuddhist redaction of theŚaiva list of holy sites. Second, althoughthe CS provides a list of the sites in question, it does not per segive any clear descriptions of their actual geographical locations,or, for that matter, of the practice of actual pilgrimage to them.In terms of practice the CS rather focuses on the visualization orcontemplation of the holy sites in the form of amaṇḍ ala . Onlylater scriptures, such as Nāropāda’s work, consider them in termsof geographical locations.

Attention should also be paid to the change of descriptionsfrom the CS to theSaṃvarodayatantra with regard to the travelof theCakrasa ṃvara practitioner. In the CS, the places where thepractitioner travels in search forḍākin ī s are described as ‘villages’

8/19/2019 9010-8804-1-PB

29/41


(gr āma),45 and these are not yet de ned as the twenty-fourCakra-saṃ vara sites. TheSaṃvarodayatantra, on the other hand, de nesthese locations to be the twenty-four sites.46 This change of descrip-tion could result from an attempt to interpret these holy sites as asymbolic framework, mapped to sets of villages or towns in areasthat were actually controlled byCakrasa ṃvara Buddhists, or atleast accessible to them. Gray argues that IndianCakrasa ṃvara Buddhists emphasized the internal practice of the twenty-four sites(i.e., meditational practice of the bodymaṇḍ ala , in which all thesites are visualized in one’s body), and that this may have re ectedthe political reality that Buddhists did not have control over many,or any, of them (pp. 68–70). But his explanation covers only half ofthe history of the theology of these holy sites because it ignores thatIndianCakrasa ṃvara Buddhists eagerly attempted to map them tothe human body and to map and remap them over the Indian conti-nent already before Abhayākaragupta.

Interpretive exibility regarding the mapping of the twenty-fourCakrasa ṃvara holy sites was already, and often, the hermeneuticstance of IndianCakrasa ṃvara Buddhists prior to Abhayā ka-ragupta; it is not exclusively linked to the tradition’s subsequenttransmission to areas such as Kathmandu Valley, Tibet, andMongolia. Abhayākaragupta’s statement that any sites where livingḍā kinī s reside can be regarded asCakrasa ṃvara holy sites shouldbe understood in this hermeneutic context of IndianCakrasa ṃvara Buddhism, as well as in the context of the tradition’s transmissionfrom India to its outlying areas. Finally, I would like to make asmall suggestion concerning Abhayākaragupta’s mention of Tibetand China. Gray states that “the mention of Tibet and China issurely not accidental, as these were major destinations for its (=the Cakrasa ṃvara tradition’s, TS) transmission, of which eruditeIndian Buddhists such as Abhayākaragupta were certainly aware(p. 70).” This may be correct, but it is also possible that Abhayā-karagupta merely followed Nāropāda, who had mentioned Tibetan

45 See again the passagegr āme gr āme vrajan tasya d ūtayas r ū pa la k ṣa-ṇam discussed in section 3.2 of this paper. 46 TheSaṃvarodayatantra , 9.

8/19/2019 9010-8804-1-PB

30/41


and ChineseCakrasa ṃvara sites in hisYul nyi bcu bshi’i rgyu mt-shan bstan pa , and theVajra ḍākatantra , which, together with theḌākārṇavatantra, de ned Bhoṭa (i.e., Tibet) as one of theCakra-saṃ vara holy sites.

5.2 The development of systems of twenty-four internal holy sites

The twenty-fourCakrasa ṃvara holy sites have both external andinternal forms. In the internal practice of the twenty-four holy sitesthese sites are visualized at various places in the practitioner’s ownbody. The collective body of these twenty-four holy sites is noth-ing other than the highest deity, Heruka. Hence, the practitioner,through the visualization of the internal holy sites, experiences anidenti cation with Heruka as his or her innate Buddha nature.

As Gray mentions, the meditational practices of internal holysites are often seen as the highest form of practices centered on holysites in theCakrasa ṃvara tradition (pp. 68–70). TheCakrasa ṃvara Buddhists were very eager to develop a system of internal holy sitesand created many varieties of such a system. However, Gray’s por-trayal of the system of internalCakrasa ṃvara holy sites coversonly half of the tradition (which may well have been his intention)

As I argued in 2003,47 a more comprehensive view suggests thatthe theories regarding the internal Herukamaṇḍ ala comprised ofthe twenty-four sites developed in two stages, with Gray’s portrayalbeing limited to the rst: (1) the stage of the internalization of thetwenty-four holy sites (i.e., the Triple Wheel), which symbolize theda ś abh ūmi andda ś ap āramit ā, and (2) the stage of the internaliza-tion of the entire Herukamaṇḍ ala including the twenty-four holysites (i.e., the Great Bliss, Triple, and Pledge Wheels), which sym-bolizes thesaptatri ṃś adbodhip āk ṣikadharma as well as theda ś a-bhū mi andda ś ap āramit ā.48

47 Sugiki 2003b; see also Sugiki 2007. 48 Note that in some texts, thetrik ā ya , theda ś ajñ āna , thetrayoda ś abh ūmi,and some other concepts traditionally taught in Mahāyāna Buddhism are in-ternalized along with theda ś abh ūmi, da ś ap āramit ā, andsaptatri ṃś adbod hip āk ṣikadharma .

8/19/2019 9010-8804-1-PB

31/41


The Cakrasa ṃvara literature describes a variety of forms of theinternal Herukamaṇḍ ala . They can be classi ed into ve types.The rst type appears in the last half of chapter 50 of the CS. Thesecond type is introduced in the Abhidh ānottarottaratantra , theVajra ḍākatantra , the Saṃvarodayatantra , the Saṃ pu ṭ od bha va-tantra , Lūy ī pāda’s Cakrasa ṃvar ābhisamaya , Jayabhadra’sCakra-saṃvaras ādhana , and so forth. The third type can likewise befound in the Abhidh ānottaratantra , as well as in Dh ī mat’sCakra-saṃ varodayama ṇḍ alop ā yik ā and Kumārakalahaṃsapāda’s Saṃ-vara rahasyan āmas ādhana . The fourth type appears in Atiśa’s

Abhi samayavibha ṅga , Prajñārakṣita’s Abhisamayapañjik ā, Tathā-gatavajra’s Abhisamayav ṛ tti , Abhayākaragupta’sCakra saṃ var ā-bhisamaya , and Śubhākaragupta’s Abhisamayamañjar ī , which,except for the last two, are commentaries on Lūy ī pāda’s Cakra-saṃ var ābhisamaya . The internal Herukamaṇḍ ala given in theYogin ī saṃcāratantra can also be considered as of this fourth type.The fth type is described in Dārikapāda’sCakrasa ṃvaras ādhana ,Ghaṇā pāda’sCakrasa ṃvaras ādhana andK ā yama ṇḍ al ābhisamaya ,Kṛṣṇā cārya’sCakrasa ṃvaras ādhana andVasantatilak ā, and in the

Jñ āno da yatantra . The versions of the rst and second types of theinternal Herukamaṇḍ ala emerged in the rst stage of development,whereas the third, fourth, and fth types developed in the second.

Let us see the ve types of the internal Herukamaṇḍ ala in de-tail. The following elements constitute instructions of thismaṇḍ ala : (0) Basic philosophy:

A somatic philosophy that enlightenment can be obtainedthrough one’s own body: one’s body is a means for attainingenlightenment.

(i) Internalized objects:(i-1) Twenty-four holy sites and twenty-four coupled deities(i.e., the Triple wheel), which are equivalent to theda ś abh ūmi and theda ś ap āramit ā.(i-2) Thirty-seven coupled and single deities (i.e., the wholemaṇ ḍ ala including the Triple wheel), which are equivalentto thesaptatri ṃś adbodhip āk ṣikadharma as well as theda ś a-

8/19/2019 9010-8804-1-PB

32/41


bhū mi and theda ś ap āramit ā. (ii) Body counterparts:

(ii-1) The channels (nāḍī ) together with their correspondingbody ingredients (dh ātu), and body sites (sth āna etc.) wherethe channels are seated.(ii-2) Heruka’s supernatural form: Heruka’s four faces and theobjects in Heruka’s twelve hands, and the externalV ār āhī .(ii-3 ) The f our principal circles (cakra ): the mah āsukhacakra

in the head, thesaṃbhogacakra in the throat, thedharma-cakra in the heart, and thenirm āṇacakra in the abdomen;and the eight gates of the body (i.e., the eight ori ces: rightand left ears, right and left eyes, right and left nostrils, mouth,and anus).

(iii) Methods for actual practice (i.e., meditation):

Meditational process for visualization of the internal Herukamaṇḍ ala .All ve types of the internal Herukaman ḍ ala share the somaticphilosophy (the factor (0) above) which legitimizes the internalpractice of the Herukamaṇḍ ala . But the ve types are distin-guished from each other by the elements (i), (ii) and (iii) as shownin the following table.

(i) (ii) (iii)

First type (i-1) (unclear) (unclear)Second type (i-1) (ii-1) DescribedThird type (i-2) (ii-1) and (ii-2) Described

Fourth type (i-2) (ii-1) and (ii-3) DescribedFifth type (i-2) (ii-1) Described

The rst and second types internalize the twenty-four holy sitesand the twenty-four coupled deities (i-1) and therefore can be saidto aim at the somatic application of the somatic philosophy focuss

8/19/2019 9010-8804-1-PB

33/41


ing on theda ś abh ūmi and da ś ap āramit ā. Unlike the second typethe rst type gives no explanation for (ii) and (iii) even though it ar-gues that the twenty-four holy sites should be practiced internally.Put in another way, the system of the rst type remains idealized.This idealized system, however, functions as a foundation for thesecond type. The systems of the third, fourth, and fth types are inturn based on the second type with regard to the Triple-wheel partof themaṇḍ ala . For this reason, the idealized system of the rsttype can be de ned as the prototypical form of the internal Herukamaṇḍ ala . The second type attaches the elements (ii-1) and (iii) tothis prototype; in the second type channels, body ingredients, andbody sites are equated with the twenty-fourḍākin ī s, the twenty-fourvī ra s, and the twenty-four sites on the Triple Wheel, respectively.

The third, fourth and fth types internalize the element (i-2).Their aim can be described as the somatic application of the somat-ic philosophy focussing on thesaptatri ṃś adbodhip āk ṣikadharma along with theda ś abh ūmi andda ś ap āramit ā. This shift from (i-1)to (i-2) seems to have some relation to the development of an exter-nal ve-wheeled Herukamaṇḍ ala in the scriptures of theCakra-saṃ vara tradition composed after the CS.

The third type applies the concept of (ii-1) for the internaliza-tion of the Triple Wheel, and applies (ii-2) for the internalizationof the Great Bliss Wheel and the Pledge Wheel. The fourth type,on the other hand, introduces the concept of (ii-3) for the internali-zation of the Great Bliss Wheel and the Pledge Wheel. The fthtype applies the concept of (ii-1) not only for the internalization ofthe Triple Wheel but also for the internalization of the Great BlissWheel and the Pledge Wheel.

As noted above, the historical development of internal forms ofthe Herukamaṇḍ ala began in its rst stage with the internaliza-tion of the twenty-four holy sites/theda ś abh ūmi andda ś ap āramit ā (i.e., the Triple Wheel). This was followed by the internalization ofthe wholemaṇḍ ala / saptatri ṃś adbodhip āk ṣikadharma along withthe da ś abh ūmi andda ś ap āramit ā (i.e., the Great Bliss, the Triple,and the Pledge Wheels).49 The signi cance of the instruction in the

49 However, it should be noted that the arrival of the the third, fourth, and

8/19/2019 9010-8804-1-PB

34/41


internal Triple-wheel Herukamaṇḍ ala given in the CS (i.e. the rstof the ve types), is that it provides the basic conceptual frame-work forCakrasa ṃvara systems of practice of the internal Herukamaṇ ḍ a la , around which the later and more elaborate types weredeveloped.

Conclusion

Criticizing the work of others is a relatively easy task, whereasproducing original studies is diffi cult. Although Gray’s pioneeringstudy on the CS has some problems, it is nevertheless a signi cantcontribution to the study of Indian Buddhist Yogin ī tantra litera-ture, for the reasons I mentioned in the rst part of this article: Thisis the rst full translation of the CS; with its copious annotation,Gray’s study can serve as a guidebook to the commentaries on theCS, and it provides much textual and contextual information on theIndianCakrasa ṃvara tradition in general. Together with the bookunder review, Gray’s critical edition of the Sanskrit and TibetanCS, announced as a companion volume, will hopefully further pro-mote the study of Tantric Buddhism.

Acknowledgments

I am responsible for any mistakes found in this paper. However, I

would like to express my heartfelt thanks to Mr. Joseph M. Loganfor having assisted me with the English and to Dr. Birgit Kellnerfor suggestions and editorial contributions. When I was writingthis review article, Péter-Dániel Szántó also published a short re-view [Tantric Studies 1 (2008) 215–219]. When I was nalizingthis article, Alexis Sanderson’s paper “TheŚaiva Age – The Riseand Dominance ofŚaivism during the Early Medieval Period”

fth types does not signal the extinction of the tradition of the second typeVersions of the internal Herukamaṇḍ ala that can be classi ed as belong-ing to the second type continued to be practiced even after the third, fourthand fth types appeared, presumably because the second type was taught inthe classicCakrasa ṃvara canons such as the Abhidh ānottaratantra , theVajra ḍākatantra , theSaṃ pu ṭ odbhavatantra , etc. and therefore often seenas authoritative.

8/19/2019 9010-8804-1-PB

35/41


appeared [In:Genesis and Development of Tantrism , ed. ShingoEinoo. Tokyo 2009: Institute of Oriental Culture, University ofTokyo, 41–349]. Sanderson’s paper also includes his criticism ofGray’s work. Some of the arguments that I made here overlap witharguments by Szántó and Sanderson, which, I hope, the readerswill kindly pardon.

(Research for this review article was partially supported by theMinistry of Education, Science, Sports and Culture, Grant-in-Aidfor Scienti c Research (C), 20520047, 2008.)

8/19/2019 9010-8804-1-PB

36/41


Appendix: draft editions of two passages from the Abhidh ā -not taratanta that explain the subtle body system and psycho-somatic meditation based on it

These passages were originally presented in Sugiki 2007 and arereproduced here because this paper may be diffi cult to access fromoutside Japan. These are not the only passages that explain the sys-tem in question in the Abhidh ānottaratantra, but they are suffi cientto validate my analysis in section 2.2 above.Two special conventions are used in the text:

*viraja ṃ [→ virama ṃ]: all manuscripts readviraja ṃ , but this is likelyto be a later emendation or corruption; the readingvirama ṃ is morelikely to be original.*ekaṃ [= prathama ṃ]: while the reading outside the brackets (here:ekaṃ) is irregular, or its meaning is obscure, it is nevertheless deemedto be the original reading (and, hence, acceptable in the context of the

edited work). However, it should be regarded as equivalent to the termgiven in brackets.

The Abhidh ānottaratantra . Skt mss: IASWR I-100 149b5–150a3,Matsunami 10 158a4–b2. Matsunami 12 186a2–b1.catu ḥsandhy ānuṣṭ heyaṃ1 catur ānandanandana ṃ2 / nābhih ṛ tpadmas ta-nau madhye 3 jihv āmūle ś iropari // ānanda ṃ paramam 4 caiva *viraja ṃ [→ virama ṃ]5 sahaja ṃ6 tath ā / catu ḥṣaṣṭ idalam 7 *ekaṃ [= prathama ṃ]8

1 catu ḥsandhy ā- ] IASWR;catusandhy ā Matsunami 10 and 12. 2 -nandana ṃ ] em.;nandan āṃ IASWR and Matsunami 12.nandan ā Matsu-nami 10. 3 tanau madhye ] IASWR;tan ūmadhya Matsunami 10.tanumadhye Matsu-nami 12. 4

paramam ] IASWR; param ām Matsunami 10 and 12. 5 The third of the Four Blisses (catur ānanda ) is generally namedvirama .(The Four Blisses are:ānanda , parama -ānanda , virama -ānanda , and sahaja-ānanda .) 6 sahaja ṃ ] Matsunami 10 and 12; om. IASWR. 7 catu ḥṣaṣṭ i- ] Matsunami 10 and 12;catu ṣaṣṭ i IASWR. 8 The intention of this phrase is that the rst inner circle connected withānanda, the rst Bliss, is of the shape of a lotus with sixty-four petals. Hence, I

8/19/2019 9010-8804-1-PB

37/41

8/19/2019 9010-8804-1-PB

38/41


*mantha [→ manthya ]manth ānayogena 19 jñ āna*ra ś mir [→ vahnir ]20 iha karma ṇā mārutena prerito n ābhima ṇḍ ale dh ūmā yati jvalati 21 d ī p-tibhiḥ22 / < samayacakre gat ān sugat ān *dagdh ā[→ dagdhv ā]23 ,> 24 ta -thā gat ānāṃ25 saṃbhogacakra*gat ān[→ gatam ]26 up ā yaṃ triḥ pra da-k ṣiṇī k ṛ tya, ūr ṇākoś agatena 27 marmodgh āṭ anadv āreṇa 28 niḥsṛ tya 29 ,da ś a diglokadh ātusthit ānāṃ tath āgat ānāṃ jñ ānāmṛ taṃ30 gṛ hī tvā ,ś ikhā ra ndhragatena kanakadv āreṇa 31 jālandharasa ṃ jñakena 32 pra viś -

ya, dantas ī mottaragatarandhre ṇa 33 saṃbhogacakre vi ś ramya 34 , dag-dh ānāṃ35 tath āgat ānām ānanda ṃ janayant ī , n ābhima ṇḍ ale 36 sthi r ī -

19 Both the phrasesmanthamanth āna andmanthyamanth āna can be found inBuddhist esoteric scriptures, but the latter is better.

20 Generally the psychosomatic re of gnosis is named jñ ānavahni , jñ ānāgni ,or jñ ānānala , but jñ ānara ś mi appears to be acceptable. 21 jvalati ] IASWR and Matsunami 10; jvaranti (or -ra - is cancelled?)Matsunami 12.

22 d ī ptibhi ḥ ] Matsunami 10 and 12; jva ( ve letters blurred) IASWR. 23

This line explains the process of the jñ ānara ś mi or jñ ānavahni’s upwardmovement from the navel circle to the heart circle. The meaning of this line is:‘Having burnt the Sugatas residing on thesamayacakra (i.e., thedharmacakra inthe heart), …’ Hence,dagdhv ā is better. 24 Fromsamayacakre to *dagdh ā[→ dagdhv ā] (inside the brackets) ] blurredin IASWR. 25 tath āgat ānāṃ ] Matsunami 12;t ānāṃ Matsunami 10. 26 What resides on the saṃbhogacakra (i.e., thecakra in the throat) is thesoundO Ṃ , which is here referred to withup ā yaṃ.

27 -gatena ] Matsunami 10;gate IASWR and Matsunami 12. 28 marmodgh āṭ ana- ] em.;ma ( ve or six letters blurred) IASWR.rmodgh āṭ ana Matsunami 10.mamodghe ṭ ana Matsunami 12. 29 niḥsṛ tya ] em.; blurred (nis ṛ tya or nis ṛ jya ?). IASWR.niśṛ tya Matsunami10and 12. The jñ ānara ś mi or jñ ānavahni goes out of the practitioner’s bodythrough his or hermarmodgh āṭ anadv āra . Hence,niḥsṛ tya is better thanniś ritya ,another possible emendation of niśṛ tya .

30 -mṛ taṃ ] Matsunami 10 and 12;mṛ t ā LASWR. 31 kanakadv āreṇa ] Matsunami 10 and 12; kanakakalajalena IASWR. 32 jālandhara- ] IASWR and Matsunami 12; jāraṃdhara Matsunami 10. 33 -randhre ṇa ] IASWR and Matsunami 12;caṃdreṇa (or readable as raṃ-dhre ṇa?) Matsunami 10. 34 viś ramya ] IASWR and Matsunami 10;viś am ā Matsunami 12. 35 dagdh ānāṃ ] Matsunami 10 and 12;dagdh ānā IASWR. 36 -ma ṇḍ ale ] Matsunami 10;maṇḍ ala (or readable asmaṇḍ ale ?) Matsu-

8/19/2019 9010-8804-1-PB

39/41


bhavati 37 //

Abbreviations

Skt ms(s). Sanskrit manuscript(s).Skt ed(s). Sanskrit text(s) critically edited.IASWR. Mss on micro lm copies kept at the [recently defunct] Institute for

the Advanced Studies of World Religions, Stony Brook, NY. Cataloguenumbers according to: Christopher S. George and Mānabajra Bajrācārya,

Buddhist Sanskrit Manuscripts. A Title List of the Micro lm Collection ofThe Institute for Advanced Studies of World Religions. Stony Brook 1975.Matsunami. Mss on micro lm copies kept at the library of the University of

Tokyo. Catalogue numbers according to: Seiren Matsunami, A Catalogueof the Sanskrit Manuscripts in the Tokyo University Library . Tokyo 1965:Suzuki Research Foundation.

NGMPP. Mss on micro lm copies kept at National Archives in Kathmandu.Reel numbers according to Nepal-German Manuscript Preservation

Project.

Primary sources

Abhidh ānottaratantra . Skt ms: IASWR I-100, Matsunami 10 and 12. Skt ed(selected chapters): Kalff 1979.

Cakrasa ṃvaratantra , or CS.Cakrasa ṃvaramah ā yogin ī tantrar ā ja . Skt ms:Oriental institute Vadodara, accession no 13290.

CS-P. CS edited by Janardan Shastri Pandey. See Pandey 2002. Brahmay āmala, or BY. Brahmay āmala. Skt ms : NGMPP A42/6. Skt ed (se-

lected paragraphs or chapters):Sanderson 2006 and Hatley 2007.Cakrasa ṃvarapañjik ā, or J. Cakrasa ṃvarapañjik ā, Jayabhadra’s commen-

tary on the CS. Skt ed: Sugiki 2001. Herukas ādhananidhi, or K. Herukas ādhananidhipa ñ jik ā, Kambala’s com-

mentary on the CS. Skt ed: Unpublished edition by Sugiki.Vajra ḍākatantra, or VḌT. Vajra ḍākamah ā yogin ī tantrar ā ja . Skt ed:Unpublished edition by Sugiki.Cakrasa ṃvaraviv ṛ ti, or Bh.Cakrasa ṃvaraviv ṛ ti, Bhavabhaṭṭa’s commentary

on the CS. Skt ed: Pandey 2002.

nami 12. 37 -bhavati ] Matsunami 10 and 12; blurred IASWR.

8/19/2019 9010-8804-1-PB

40/41

8/19/2019 9010-8804-1-PB

41/41


Sugiki 1999. Tsunehiko Sugiki: Kṛṣṇācārya-no shi-shidai (Kṣṇācārya’sFour-staged Meditational Process). Indo-gaku-bukky ō-gaku-kenky ū ( Jour nal of Indian and Buddhist Studies ) 47/2 (1999) 880–883.

Sugiki 2000. Tsunehiko Sugiki: Kṛṣṇācārya’s Ś r ī cakrasa ṃvaras ādhana –Critical Edition with Notes.Chisan-gakuh ō ( Journal of Chisan Studies )49 (2000) 45–62.

Sugiki 2001. Tsunehiko Sugiki:Cakrasa ṃvaratantra -no seiritsu-dankai-nitsuite – oyobi Jayabhadra-sakuŚ r ī cakrasa ṃvarapañjik ā-kōtei-Bonpon(On the Making of theCakrasa ṃvaratantra and a Critical Study of Jaya-bha dra’s Ś r ī cakrasa ṃvarapañjik ā). Chisan-gakuh ō ( Journal of ChisanStudies ) 50 (2001), 91–141.

Sugiki 2002. Tsunehiko Sugiki: A Critical Study of theVajra ḍāka mah ā-tan trar ā ja (I) – Chapters 1 and 42.Chisan-gakuh ō ( Journal of ChisanStudies ) 51 (2002) 81–115.

Sugiki 2003a. Tsunehiko Sugiki: A Critical Study of theVajra ḍāka mah ā tan-tra r ā ja (II) – Sacred Districts and Practices Concerned.Chisan-gakuh ō ( Journal of Chisan Studies ) 52 (2003) 53–106.

Sugiki 2003b. Tsunehiko Sugiki: Five Types of Internal Maṇḍ ala Describedin theCakrasa ṃvara Buddhist Literature – Somatic Representations ofOne’s Innate Sacredness.T ō yō-bunka-kenky ū-sho-kiy ō (The Memoirs ofthe Institute of Oriental Culture ) 144 (2003) 157–231.

Sugiki 2006. Tsunehiko Sugiki: Haiburid-na seichi-no tajigensei: bukkyō-saṃ varakei-niokeru ikai-no chikara. In: Ikai-no k ōsaku vol.2, ed. AyakoHosoda and Kazuko Watanabe. Tokyo 2006: Lithon, 55–85.

Sugiki 2007. Tsunehiko Sugiki:Saṃvara-kei mikky ō-no shos ō ( Aspects ofSaṃ vara Esoteric Buddhism ). Tokyo 2007: Tōshindō.

Sugiki 2008. Tsunehiko Sugiki: The Homa System of theVajra ḍākatantra : ACritical Edition and a Preliminary Analysis of its Homa System.TantricStudies 1 (2008) 131–154.

Sugiki 2009. Tsunehiko Sugiki: The Structure and Traditions of the Systemsof Holy Sites in BuddhistSaṃvara Cycle and its Related ScripturalCycles in Early Medieval South Asia: The Geography of EsotericBuddhism in the Eyes of the Compilers of the Scriptures. In:Genesis and

Development of Tantrism , ed. Shingo Einoo. Institute of Oriental CultureSpecial Series 23. Tokyo 2009: Institute of Oriental Culture, Universityof Tokyo, 515–562.

Tsunehiko SUGIKI

9010-8804-1-PB

Documents

Transcript of 9010-8804-1-PB