Buddhist Epistemology as Apologetics: Studies on the History, Self-understanding and Dogmatic Foundations of Late Indian Buddhist Philosophy ... Kultur-Und Geistesgeschichte Asiens NR. 81) 9783700175834, 3700175833

Citation preview

VINCENT ELTSCHINGER BUDDHIST EPISTEMOLOGY AS APOLOGETICS STUDIES ON THE HISTORY, SELF-UNDERSTANDING AND DOGMATIC FOUNDATIONS OF LATE INDIAN BUDDHIST PHILOSOPHY

ÖSTERREICHISCHE AKADEMIE DER WISSENSCHAFTEN PHILOSOPHISCH-HISTORISCHE KLASSE SITZUNGSBERICHTE, 851. BAND

BEITRÄGE ZUR KULTUR- UND GEISTESGESCHICHTE ASIENS NR. 81

Herausgegeben von Helmut Krasser

ÖSTERREICHISCHE AKADEMIE DER WISSENSCHAFTEN PHILOSOPHISCH-HISTORISCHE KLASSE SITZUNGSBERICHTE, 851. BAND

Buddhist Epistemology as Apologetics Studies on the History, Self-understanding and Dogmatic Foundations of Late Indian Buddhist Philosophy Vincent Eltschinger

Vorgelegt von k. M. HELMUT KRASSER in der Sitzung am 13. Dezember 2013

Diese Publikation wurde einem anonymen, internationalen Peer-Review-Verfahren unterzogen. This publication has undergone the process of anonymous, international peer review.

Die verwendete Papiersorte ist aus chlorfrei gebleichtem Zellstoff hergestellt, frei von säurebildenden Bestandteilen und alterungsbeständig.

Alle Rechte vorbehalten. ISBN 978-3-7001-7583-4 Copyright © 2014 by Österreichische Akademie der Wissenschaften Wien Druck und Bindung: Prime Rate kft., Budapest Printed and bound in the EU http://hw.oeaw.ac.at/7583-4 http://verlag.oeaw.ac.at

Contents

Foreword ............................................................................................ ix Introduction: On Critical Examination and Apologetics .......................................... 1 Chapter 1: Apocalypticism, Heresy and Philosophy .......................................... 35 1.1. Introduction ..................................................................................... 35 1.2. Mleccha and pāṣaṇḍa/pāṣaṇḍin in Brahmanical Apocalypticism ................................................................................................... 40 1.2.1. Introduction .............................................................................. 40 1.2.2. The Doctrine of the yugas in the MDhŚ ................................... 41 1.2.3. The kaliyuga and its End in the MBh ....................................... 41 1.2.4. The kaliyuga in the YP ............................................................. 50 1.2.5. Conclusion ................................................................................ 53 1.2.6. The kaliyuga and its End in the Early Purāṇas ......................... 54 1.2.7. Kaliyuga and Heresiology in the ViP ....................................... 57 1.2.8. Kaliyuga and Heresiology in the Mīmāṃsā ............................. 66 1.2.9. Conclusion ................................................................................ 70 1.3. Indian Buddhist Apocalypticism ..................................................... 73 1.3.1. Introduction .............................................................................. 73 1.3.2. The Timetables and the Causes of Decline .............................. 73 1.3.3. Monastic Degeneration and Early Mahāyāna........................... 77 1.3.4. The Buddhist Appropriation of the kaliyuga ............................ 81

1.3.5. Narendrayaśas .......................................................................... 85 1.3.6. Conclusion ................................................................................ 89

Chapter 2: Buddhist Esoterism and Epistemology ............................................. 93 2.1. Introduction ..................................................................................... 93 2.2. Śaivism, Early Medieval India and the Rise of Buddhist Tantrism ............................................................................................ 96 2.2.1. Preliminary Remarks ................................................................ 96 2.2.2. Buddhist Tantrism as the Internalization of the Early Medieval Political Landscape ..................................................... 100 2.2.3. Buddhist Esoterism as an Adoption-cum-Adaptation of Śaiva Tantrism ............................................................................ 106 2.2.4. Summary ................................................................................ 113 2.2.5. Remarks on the Chronology of Early Buddhist Tantrism ...... 114 2.2.6. Dharmakīrti and the Early History of Tantrism...................... 120 2.2.7. Buddhist Hostility towards Śaivism in Early Buddhist Tantric Scriptures ........................................................................ 136 2.2.8. Conclusion: Towards a New Modality of Mahāyānist Self-Assertion? ............................................................................ 149 2.3. Buddhist Epistemology ................................................................. 154 2.3.1. Remarks on the Historiography of the Buddhist Epistemological School ......................................................................... 154 2.3.2. On the Buddhist Epistemologists' Self-Understanding .......... 169 2.3.3. The Decline of Abhidharma Scholasticism and Intersectarian Polemics ....................................................................... 174 2.4. Conclusion ..................................................................................... 189

Chapter 3: Turning Hermeneutics into Apologetics ......................................... 191 3.1. Introduction ................................................................................... 191 3.2. Evaluative Rationality ................................................................... 193

3.2.1. The Epistemologists’ Concept of yukti .................................. 193 3.2.2. yukti and āgama in Traditional Buddhist Scholasticism ........ 196 3.2.3. Personal or Textual Authority? .............................................. 201 3.2.4. Assessing Scriptures ............................................................... 210 3.2.5. Conclusion .............................................................................. 218 3.3. Practical Rationality ...................................................................... 219 3.3.1. The Epistemologists’ Concept of prekṣā ................................ 219 3.3.2. Religious Practice as Rational Action .................................... 221 3.3.3. Certainty and Rationality........................................................ 224 3.3.4. The Revised kāryānumāna Theorem ...................................... 227 3.3.5. The Logic of Buddhist Gradualism ........................................ 234 3.4. Conclusion ..................................................................................... 244

Chapter 4: Nescience, Epistemology and Soteriology ..................................... 247 4.1. Introduction ................................................................................... 247 4.2. Nescience as Erroneous Cognition (mithyopalabdhi) ................... 248 4.2.1. Nescience as Counter-Insight ................................................. 248 4.2.2. Nescience, saṃvṛti, and apoha ............................................... 255 4.3. Nescience as Personalistic False View (satkāyadṛṣṭi) ................... 266 4.3.1. The False View of Self as a Specification of Wrong Perception ......................................................................................... 266 4.3.2. Dependent Origination ........................................................... 278 4.3.3. Dharmakīrti’s Sources for the Equation avidyā = satkāyadṛṣṭi .......................................................................................... 293 4.4. Nescience, Inference, and the Path toward Salvation .................... 298 4.4.1. Introduction ............................................................................ 298 4.4.2. Perception and Insight ............................................................ 300 4.4.3. Erroneous Superimposition and Its Causes ............................ 306 4.4.4. Inference and Its Corrective Function .................................... 311

4.4.5. Inference and the Path ............................................................ 313 4.4.6. The cintāmayī prajñā .............................................................. 318

Abbreviations and Bibliography ..................................................... 329 Indices ............................................................................................. 371

Foreword

The present volume is a collection of four studies which, though originally published as independent essays, have been conceived as chapters of an organic book dedicated to the socio-historical context and the dogmatic foundations of early Indian Buddhist epistemology. The volume was intended as—and remains—a general introduction to this religio-philosophical current’s apologetic dimensions, properly speaking—proofs of the possibility of rebirth, insight, compassion, liberation and omniscience, i.e., a demonstration of the rationality of the Buddhist salvational path. Parts of the materials presented in Chapter 1 (“Apocalypticism, Heresy and Philosophy”) were first presented on the occasion of the international conference “World View and Theory in Indian Philosophy” (Barcelona, Casa Asia, 26–30 April 2009), and then twice in Japan (Tokyo University, 30 September 2009; Ryukoku University, 27 November 2009); the original study was published under the same title in a volume edited by Piotr Balcerowicz (World View and Theory in Indian Philosophy. Delhi 2012: Manohar [Warsaw Indological Studies Series 5], pp. 27–84). Chapters 2 and 3 go back to two papers delivered at the XIVth World Sanskrit Conference (Kyoto University, 1–5 September 2009): Whereas “Buddhist Esoterism and Epistemology” was initially published in the proceedings of the Kyoto panel edited by Eli Franco (Periodization and Historiography of Indian Philosophy. Vienna 2013: De Nobili [Publications of the De Nobili Research Library 37], pp. 171–273), “Turning Hermeneutics into Apologetics” first appeared in the volume of proceedings edited by myself and Helmut Krasser (Scriptural Authority, Reason, and Action. Vienna 2013: Austrian Academy of Sciences Press [Beiträge zur Kultur- und Geistesgeschichte Asiens 79], pp. 71–145). The research that resulted in Chapter 4 (“Nescience, Epistemology and Soteriology”) was originally presented in the framework of the XVth Congress of the International Association of Buddhist Studies (Atlanta, Emory University, 23–28 June 2008) and published in two parts in the Journal of the International Association of Buddhist Studies (vol. 32/1–2, 2009 [2010], pp. 39–83, and vol. 33/1–2, 2010 [2011], pp. 27–73). Except for “Nescience, Epistemology and Soteriology,” which retrieves its original unity and was added section 4.4.6 on the cintāmayī prajñā, the studies un-

x

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

derwent no substantial modification. Besides unifying styles, spellings and bibliographical information as well as adding all relevant cross-references, I have updated what was necessary. Thus, Chapter 1 incorporates materials drawn from and references to Giovanni Verardi’s recently published (2011) Hardships and Downfall of Buddhism in India. Chapter 2 has benefitted from Christian Wedemeyer’s Making Sense of Tantric Buddhism (2013) and Eltschinger 2012. As for Chapter 3, it now takes into consideration Richard Nance’s recent (2011) Speaking for Buddhas: Scriptural Commentary in Indian Buddhism. The studies that served as a basis for Chapters 1, 2 and 4 were funded by the Austrian Science Fund (FWF, project P21050-G15: “Tradition und Wandel in der indischen buddhistischen Logik”). Part of the research work that led to Chapter 1 was also made possible by the Numata Foundation, to whose generous support I owe an extremely fruitful stay in Kyoto (Ryukoku University, September-December 2009). I am very grateful to these institutions as well as to Ernst Steinkellner, Helmut Krasser and Shoryu Katsura. I wish to express my most sincere thanks to Diwakar Acharya, Piotr Balcerowicz, Johannes Bronkhorst, Danielle Feller, Peter Flügel, Erika Forte, Eli Franco, Gérard Fussman, Dominic Goodall, Harunaga Isaacson, Kyo Kano, Shoryu Katsura, Birgit Kellner, Deborah Klimburg-Salter, Helmut Krasser, Hiroshi Marui, Jan Nattier, Marion Rastelli, Isabelle Ratié, Alexander von Rospatt, Masamichi Sakai, Cristina Scherrer-Schaub, Peter Skilling, Ernst Steinkellner, François Voegeli, Toshihiko Watanabe, Yuko Yokochi, Chizuko Yoshimizu, and Kiyotaka Yoshimizu for their very precious help. I am also very grateful to Cynthia Peck-Kubaczek and Katharine Apostle, who improved the English of the original papers and the introduction. My deepest gratitude goes to two exceptional scholars, Alexis Sanderson and Lambert Schmithausen, for their extremely careful reading and improvement of the studies that were to become Chapters 1 and 4. Last but not least, I would like to address my most heartfelt thanks to Karin Preisendanz and Ernst Steinkellner for encouraging me not to postpone any further this publication, a habilitation thesis submitted to the University of Vienna. Leipzig April-June 2013

Introduction On Critical Examination and Apologetics

1.1. How seriously should we take Dharmakīrti’s Buddhist affiliation? Did the great sixth-century logician and epistemologist1 simply pay lip service to Buddhism as a doctrinal and salvational system, or was Buddhism an integral part of his intellectual outillage, one that shaped his ideological convictions and religio-philosophical agenda? There can be little doubt that during its first seven or eight decades (say between 1900 and 1980), the reception of Dharmakīrti’s philosophy betrays a clear assent to the first hypothesis.2 According to the Russian neo-Kantian scholar Th. Stcherbatsky (1866–1942), “Buddhist logic” was to be interpreted as an Indic Aufklärung the most prominent representatives of which—Dignāga, Dharmakīrti, Dharmottara—had emancipated themselves from any scriptural, dogmatic and metaphysical commitment. While Vasubandhu was ancient India’s Plato, Dignāga was its Aristotle, and Dharmakīrti its Kant. Briefly put, the Buddhist epistemologists were transcendentalist “free thinkers”3 engaged in the critical assessment of the nature and legitimate scope of human understanding. According to Stcherbatsky, “the system had apparently no connection with Buddhism as a religion, i.e., as the teaching of a path towards salvation. It claims to be the natural and general logic of the human understanding.”4 Stcherbatsky’s forces were declining as E. Frauwallner (1898– 1974), who did not share his Russian colleague’s philosophical presuppositions and advocated a historical-philological approach to Indian philoso1

On the chronology of Dharmakīrti, see below, p. 116, n. 80.

2

For a short sketch of the scholarly reception of Dharmakīrti’s and his successors’ philosophy, see Steinkellner 1982. T. Vetter’s remarkable Erkenntnisprobleme bei Dharmakīrti (1964), published with the benediction of Frauwallner, is a partial exception to this.

3

Stcherbatsky 1932: 13–14, as quoted in Steinkellner 1982: 5.

4

Stcherbatsky 1932: 2, as quoted in Steinkellner 1982: 4–5.

2

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

phy, published some of the richest and most innovating studies ever dedicated to Dharmakīrti.5 But in spite of his entirely different background, Frauwallner equally understood Buddhist logic and epistemology as an axiologically neutral philosophical system in which religious, dogmatic and soteriological issues played virtually no role. With Dharmakīrti, something like a hellenic and even “Āryan” (sic) period of Indian philosophy found its culmination—and its conclusion.6 Since the early seventies analytic philosophy has become the paragon of philosophical reflection. In spite of its overall aversion for the history of philosophy, the analytical approach has remained—be it in a purely ideal way—the dominant paradigm in the historiography of Buddhist epistemology besides the more philologically oriented “Viennese school” founded by Frauwallner.7 Unsurprisingly, analytically oriented scholarship—what M. Kapstein has termed the “problems and arguments approach,”8 the program of which partly overlaps with the Problemgeschichte inherited from Frauwallner—has generally disregarded the texts as organic wholes, the socio-historical contexts and the dogmatic frameworks in favor of an unhistorical, comparative and at times formal approach to logical quantification, linguistics and ontological theory. No less than deeply ingrained convictions, however, these different types of scholarship also reflect most of their promoters’ aspiration to see these Indian and Indo-Tibetan scholastic productions finally recognized, against 5

I am thinking essentially of Frauwallner’s pathbreaking “Beiträge zur Apohalehre” (1932, 1933a, 1935).

6

On Frauwallner’s contribution to indological studies as well as his historical and ideological background, see Franco/Preisendanz 2010, Stuchlik 2009, and also below, Chapter 2, §2.3.1.2.

7

In my opinion, Taber 2013 provides the best presentation and defense of this interpretive trend to date. According to Taber (2013: 126), a philosophical—but not necessarily analytic—engagement with Dharmakīrti’s writings amounts to “reflecting on the broader philosophical meaning of his ideas in light of similar or contrasting views of the things they treat—and not just in light of theories that were current in India in his days but even ones familiar to us only from Western philosophy (such as ‘nominalism’ and ‘idealism’); analyzing his arguments and evaluating their strengths and weaknesses as philosophical arguments and assessing the overall plausibility of his system (Is it internally consistent? Does it present us with a compelling picture of reality?); and arriving at some judgment about how well his theories hold up under the sorts of criticisms that were levelled against them by his contemporaries and subsequent generations of thinkers […].”

8

Kapstein 2003: 5.

INTRODUCTION

3

resistant Anglo-Saxon and continental European prejudices, as properly and genuinely philosophical.9 As a side effect of this lack of interest in religious issues, many of those interested in Buddhism as a soteriological system (or in Madhyamaka) have come to consider Buddhist logic and epistemology a serious betrayal of Buddhism, as the following judgment by E. Conze testifies: “The importance, validity and usefulness of Buddhist logic is circumscribed by its social purpose, and the works of the logicians can therefore exhibit the holy doctrine only in a distinctly truncated form.”10 1.2. Dharmakīrtian scholarship’s interest in the dogmatic and soteriological dimensions of Buddhist epistemology became stronger in the early eighties thanks to the pioneering work of E. Steinkellner, R. Hayes, T. Vetter, E. Franco and T. Tillemans. These scholars called attention to various topics such as Dharmakīrti’s Buddhology, its indebtedness to Dignāga, rebirth, apologetics, doctrinalism, and scriptural authority.11 Reading these innovative contributions, however, one gets—or should I say “I get”?—the impression that their authors regarded the religious issues at stake as exhibiting the logicians’ treatment of, and solution to, purely occasional and neatly circumscribed problems with little or no connection to the heart of the system. In other words, these pioneers remained very cautious not to interpret these and other issues as representing more than specific aspects—less open-minded scholars would say “peripheral aspects”—of an otherwise 9

Note Kapstein 2003: 5: “The dominant, dismissive prejudgment of the analytic tradition […] was received by late twentieth-century anglophone students of Buddhist thought with sufficient seriousness that, during the past few decades, many of us who work in this and related areas have in effect devoted our energies to proving Flew & wrong—to showing, that is, that classical Indian and Buddhist thinkers were concerned with wellformed arguments, and that the problems about which they argued were often closely similar to those that are taken to exemplify philosophy in our textbooks.” &Kapstein alludes here to the following statement by Antony Flew (Flew 1971: 36, as quoted in Kapstein 2003: 5): “Philosophy […] is concerned first, last and all the time with argument […] [B]ecause most of what is labelled Eastern Philosophy is not so concerned […] this textbook draws no materials from any sources east of Suez.”

10

Conze 1962: 267, as quoted in Steinkellner 1982: 4. Note also Conze 1962: 265 (as quoted in Steinkellner 1982: 4): “At variance with the spirit of Buddhism, it can indeed be tolerated only as a manifestation of ‘skill in means’. Logic was studied ‘in order to vanquish one’s adversaries in controversy’, and thereby to increase the monetary resources of the order.”

11

See, e.g., Steinkellner 1978, 1982, 1983, 1984, 1985, and 1986, 1988, Hayes 1984, Tillemans 1986, Franco 1989, Vetter 1990.

4

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

strictly philosophical enterprise aimed at uncovering the true nature and possibilities of human knowledge. And indeed, there can be no doubt that at the surface level at least these Buddhist intellectuals’ overarching concern was epistemology in the sense of a rational and polemical inquiry into the nature, the number and the operation of the means of valid cognition (pramāṇa). But there is also very little doubt that these intellectuals were Buddhist monks active in Buddhist educational and ritual centers as specialists of the “science of [justificative] reasons(/evidences)” (hetuvidyā); that the early descriptions of the hetuvidyā reflect its essential connection to, and function as, positive and negative apologetics on behalf of Buddhism, a connection that could only gain in strength in a context of exacerbated religious rivalry; that at least since the sixth century CE Buddhism was the object of orthodox Brahmanical hostility and had to struggle, mainly against Śaivism, for economic patronage and political support; that the Buddhist logicians understood their own intellectual enterprise as instrumental in putting Buddhists as well as non-Buddhists in an epistemic position to enter the Buddhist path; that their works provide systematically argued “rationalizations” and defenses of key Buddhist dogmas such as the two “truths” (on perception, concept formation, language, and error—the apoha theory), momentariness, selflessness and nescience; that several theoretical statements regarding the two pramāṇas (perception’s direct encounter with the ultimately true features of reality; inference’s corrective and “sapiential” functions) reflect a clear attempt to have them fit the needs of Buddhist soteriology; and finally that, of course, significant parts of these works are dedicated to defend, against the critiques of allodox schools such as materialism and Mīmāṃsā, the four nobles’ truths, the doctrine of skandhas, or the Buddhist position on rebirth, insight, compassion, the path, salvation, buddhahood, omniscience and the like. To sum up, I believe that the socio-historical matrix (religious pluralism, Brahmanical hostility, competition for patronage), the identity of the opponents (rival salvational systems with strong apologetical concerns expressing themselves through linguistic and epistemological theory), the doctrinal foundations and the issues at stake call for a description of Buddhist epistemology as an apologetical enterprise—whence the title of the present book, “Buddhist Epistemology as Apologetics.” 1.3. By heuristically resorting to the concept of “apologetic(s),” my aim is to provide a more accurate and flexible description of the ways in which the religious dimensions of Dharmakīrti’s philosophy can be meaningfully taken into historical and doctrinal account. My idea is not, and has never

INTRODUCTION

5

been, to deny Dignāga, Dharmakīrti and their successors the quality of philosophers—after all the greatest among the Christian apologists such as Clement of Alexandria, Origen, Augustine, Anselm of Canterbury, Abelard or Thomas Aquinas were no insignificant philosophers.12 As I understand them, those who wrote so extensively on pramāṇas were Buddhist monastics (rarely lay people) who engaged in a twofold defense of Buddhism as a religion, attacking its rivals’ systems—this is positive apologetics—and neutralizing their real or potential objections—this is negative apologetics.13 But this is also to say that even regarding allegedly non-confessional issues such as (ontology,) language, gnoseology and epistemology, the directions taken by these scholiasts’ philosophical answers were shaped by their Buddhist education and persuasion—by their Buddhist intellectual reflexes, their “habitus” so to say—, and that they deliberately organized these answers so as to avoid any possible contradiction with received doctrine and scripture (āgamavirodha, which included contradiction with Vasubandhu’s AKBh). In other words, I believe that our authors knew what and where the truth was before engaging in philosophical analysis, and that to them this scripturally and dogmatically given truth needed to be defended: “Under the direction of faith, the apologist constructs arguments that are valid before natural reason.”14 What Dharmakīrti and his followers did was to construct and to defend, on the basis of a logical organon inherited—and updated—from Dignāga and the hetuvidyā/vāda tradition, a system of the world, the mind, cognition and salvation that conformed in every single point to the doctrinal exigencies of (a minimal and broadly consensual version) of Buddhism. I fail to see how and why such a picture could threaten

12

Dulles 2005 provides an excellent and easily accessible overview of the history of Christian apologetics.

13

On these notions, see below, §3.1–2. Needless to say, the Buddhist epistemologists also defended Buddhist doctrinal “orthodoxy” (by which I mean nothing other than Vasubandhu’s Buddhism; see below, Chapter 2, §2.3.3.2) against coreligionists such as the Vaibhāṣikas and the (Vātsīputrīya/)Sāṃmitīyas, and polemicized against other Buddhist epistemologists holding allegedly erroneous views on diverse logico-epistemological issues (see below, Chapter 2, §2.3.3.4). On the hypothesis that rather than attacking and defending Buddhism directly against its critics, these intellectuals taught their hetuvidyā/vāda students how to refute them, see below, §3.3.

14

Dulles 2005: 284, describing the position of the French Dominican theologian Réginald Garrigou-Lagrange (1877–1964). On natural reason and natural theology, see below, p. 21, n. 63.

6

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

the claim that these intellectuals were genuine philosophers. For in my opinion, the apologists of late Antiquity or the twelfth- to fourteenth-century scholastic intellectuals, all of whom structured their account of Christianity so as to defend it in contexts of real or virtual religious pluralism and competition against pagans, Gnostics, Jews, Muslims or Christian heretics, are the most likely counterparts of the Buddhist epistemologists. This is not only to say that these intellectual traditions resemble each other—a fairly trivial observation that can only result in naive and superficial comparatism—, but also that the institutional and socio-historical contexts which at least partly account for them are roughly similar. 2.1. As the pious hagiographies have it, the Buddha did not obtain awakening immediately after taking up homeless life. Abandoning the Vedic anachorites he had first joined, Gautama became the disciple of the then famous teacher Arāḍa (or: Ārāḍa, Pali Āḷāra) Kālāma, whose salvational method relied on the attainment of an “incorporeal realm” (ārūpya) called the “stage of nothingness” (ākiñcanyāyatana). The future Buddha was soon to master Arāḍa’s meditation techniques and, though offered co-responsibility over the latter’s religious congregation, left him for the concurrent teacher Udraka (or: Uḍraka, Pali Uddaka) Rāmaputra, another representative of “mainstream meditation”15 who regarded the attainment of the stage of neither-ideation-nor-non-ideation (naivasañjñānāsañjñāyatana) as bringing about salvation.16 Śākyamuni was equally disappointed and reached the bank of the Nairañjanā River in order to give himself up to the

15

This stream consisted of mainly Jaina and “Hindu” ascetic tradition(s) that, contrary to (earliest, i.e., “pre-samāpatti”) Buddhism, understood meditation as a painful and “forceful effort to restrain the mind and bring it to a standstill” (Bronkhorst 1993: 22) and as a medium “by which the practitioner gradually puts an end to all ideations” (Bronkhorst 1993: 81). Indeed, as Putkasa/Pukkusa tells the Buddha in the Mahāparinirvāṇasūtra, “Ārāḍa Kālāma at one occasion did not hear the sound of five hundred—in one version fifty—carts passing by, even though awake and conscious” (Bronkhorst 1993: 79; the different versions of the episode are discussed in Bareau 1970: 282–295). As Bronkhorst has pointed out, the episode of the future Buddha’s turning his back on the two sages and their unsatisfactory teachings is likely to be interpreted as an explicit criticism of representatives of and/or meditative practices belonging to the mainstream tradition (Bronkhorst 1993: 87).

16

On these episodes, see Bareau 1963: 13–27 (add MV II.118,1–3, LV 243,15–18, and MV II.119,8–10). Bareau’s reasons for not giving any credence to this story are summarized in Bronkhorst 1993: 85–86, summarized and criticized in Zafiropulo 1993: 23–29.

INTRODUCTION

7

most severe forms of asceticism—which in turn proved to be no less unsatisfactory methods of salvation. Most canonical versions of the events end with a recension-specific, generally stereotyped formula pointing to the unsatisfactory character of the practices experienced by the Buddha. Here is the Theravāda formula concerning Āḷāra Kālāma: “But it occurred to me: ‘This Dhamma does not lead to disenchantment, to dispassion, to cessation, to peace, to direct knowledge, to enlightenment, to Nibbāna, but only to reappearance in the stage of nothingness.’ Not being satisfied with that Dhamma, disappointed with it, I left.”17 Its Sarvāstivāda counterpart runs as follows in the LV: “The bodhisattva replied [as follows to Udraka Rāmaputra]: ‘This path, Sire, does not lead to bliss, to detachment, to cessation, to appeasement, to super-knowledge, to awakening, to [being] a [true] śramaṇa and a [true] brāhmaṇa, to cessation.’ And thus, O monks, the bodhisattva […] [said to himself] ‘Enough,’ and departed[, saying] ‘Enough of [all] this for me.’”18 The Mūlasarvāstivāda Vinaya’s formula is much shorter: “Then the [following] occurred to the bodhisattva: ‘This path is not conducive to knowledge, is not conducive to vision, is not conducive to the supreme [and] perfect awakening.”19 Equally short is the Mahāsāṅghika/Lokottaravāda version: “But, O monks, the following occurred to me: ‘Ārāḍa’s doctrine does not lead him who practices it to the real exhaustion of suffering.’”20 As we can see, neither do the canonical versions of the events shed

17

MN I.165: tassa mayhaṃ bhikkhave etad ahosi / nāyaṃ dhammo nibbidāya na virāgāya na nirodhāya na upasamāya na abhiññāya na sambodhāya na nibbānāya saṃvattati yāvad eva ākiñcaññāyatanūpapattiyā ti / so kho ahaṃ bhikkhave taṃ dhammaṃ analaṅkaritvā tasmā dhammā nibbijjāpakkamiṃ /. Translation (slightly modified) Ñāṇamoli/Bodhi 2001: 258. See also MN I.166 (Uddaka Rāmaputta; see Ñāṇamoli/Bodhi 2001: 259).

18

LV 285,11–15: bodhisattva āha – naiṣa mārṣa mārgo nirvṛtaye na virāgāya na nirodhāya nopasamāya nābhijñāyai na sambodhaye na śrāmaṇāya na brāhmaṇāya na nirvāṇāya saṃvartate / iti hi bhikṣavo bodhisattvo […] yāvad alam iti kṛtvā prakrāmad alaṃ mamāneneti /.

19

SBhV I.97,30–98,2: atha bodhisattvasyaitad abhavat – ayaṃ mārgo nālaṃ jñānāya nālaṃ darśanāya nālam anuttarāyai samyaksambodhaye /. See also SBhV I.98,29–32 (Udraka) and SBhV I.107,1–3 (duṣkaracaryā[vasthā]).

20

MV II.119,4–5: tasya me bhikṣavaḥ etad abhūṣi / nāyam ārāḍasya dharmo niryāti tatkarasya samyagduḥkhakṣayāye /. See also MV II.120,15–16 (Udraka, with tasya rāmasya instead of ārāḍasya).

8

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

any theoretical light on the practices involved nor do the various dissatisfaction formulas provide any argument against the teachings ipso facto abandoned by Gautama. In the mind of the Buddha and/or the compilers of the various canonical recensions, the failure of these practices to bring about detachment (virāga, etc.), cessation (upaśama, nirodha, nirvāṇa, duḥkhakṣaya, etc.), knowledge (jñāna, abhijñā, etc.), intuition (darśana) and awakening (sambodha, sambodhi) is strong enough an argument to disqualify them as proper soterial paths (mārga). Consider now Aśvaghoṣa’s brief outline of the same biographical sequence in the SNa: “Then quitting the majestic and secure city of Kapilavāstu, whose population was devoted to him and which was thronged with masses of horses, elephants and chariots, he started resolutely for the forest to practise austerities. But finding that the sages were practising austerities according to varying scriptures and under varying rules and were still made wretched by desire for sensory objects, he concluded that there was no certainty in asceticism and turned away. Then, with his mind fixed on the ultimate truth, he sat at the feet of Arāḍa who preached emancipation and of Uḍraka who held the doctrine of quietude, but left them, deciding in his discrimination of paths that theirs were not the right paths. After considering which of the various sacred traditions in the world was the highest, and failing to obtain exact knowledge from others, he entered after all on austerities of extreme difficulty. Then seeing this to be a false path, he gave up that extended course of austerity too and, realizing that the sphere of trance was the highest, he ate choice food to prepare his mind for the understanding of immortality.”21 Aśvaghoṣa’s narrative displays innovating features. The future Buddha is depicted as a path expert (mārgakovida) in search of certainty (niścaya) and looking for decisive conclusions as to the (un)reliability of the respective salvational methods. No less importantly, Śākyamuni critically examines (viCAR) “which of the various sacred 21

SNa 3.1–5: tapase tataḥ kapilavāstu hayagajarathaughasaṅkulam / śrīmad abhayam anuraktajanaṃ sa vihāya niścitamanā vanaṃ yayau // vividhāgamāṃs tapasi tāṃś ca vividhaniyamāśrayān munīn / prekṣya sa viṣayatṛṣākṛpaṇān anavasthitaṃ tapa iti nyavartata // atha mokṣavādinam arāḍam upaśamamatiṃ tathoḍrakam / tattvakṛtamatir upāsya jahāv ayam apy amārga iti mārgakovidaḥ // sa vicārayan jagati kiṃ tu paramam iti taṃ tam āgamam / niścayam anadhigataḥ parataḥ paramaṃ cacāra tapa eva duṣkaram // atha naiṣa mārga iti vīkṣya tad api vipulaṃ jahau tapaḥ / dhyānaviṣayam avagamya paraṃ bubhuje varānnam amṛtatvabuddhaye //. Translation Johnston 1928: (II.)15.

INTRODUCTION

9

traditions in the world [is] the highest.” In other words, entertaining soteriological expectations in a context of religio-philosophical pluralism makes a critical appraisal of the competing salvational methods necessary: There can be neither choice nor success in soteriological matters without a thorough prior evaluation of the different systems (dharma) at hand. Aśvaghoṣa lays even more emphasis on critical examination (parīkṣā, vicāra) in his BC. Consider, e.g., the monk poet’s treatment of Gautama’s encounter with Vedic asceticism in BC 7. Here, the future Buddha’s dissatisfaction (na santutoṣa BC 7.19) with the religious aims and methods of his fellow forest dwellers is based on a detailed critical examination (parīkṣamāṇa BC 7.34) and results in a thoroughly argued (bahuyuktiyukta BC 7.32) critique of tapas.22 The BC’s treatment of Gautama’s rejection of Arāḍa’s dharma (which Aśvaghoṣa fancies, most certainly due to the great success of Sāṅkhya in his time, to be a kind of proto-Sāṅkhya doctrinally close to that of the CS23) provides a paradigmatic instance of parīkṣā—in spite of the fact that the expression itself does not occur in this context. After spelling out Arāḍa’s theoretical commitments and salvational method (a substantialist and abridged version of the Buddhist dhyānas and samāpattis) in great detail,24 Aśvaghoṣa presents the future Buddha’s very systematic arguments against his early teacher’s system.25 The passage ends with the following statement: “Thus he was not satisfied on learning the doctrine of Arāḍa, and, discerning that it was incomplete, he turned away from there.”26 As we can see, Aśvaghoṣa 1. makes a Sāṅkhya philosopher out of the philosophically uncommitted canonical Arāḍa, 2. has the bodhisattva himself refute 22

On Gautama’s arguments against asceticism, see BC 7.20–31 and Johnston 1984: (II.)96–97.

23

On the Sāṅkhya known to Aśvaghoṣa, see Sastri 1952, Rao 1962, Rao 1964, Johnston 1974: passim, Hulin 1978: 136, Larson 1979: 104–108, Kent 1982, Johnston 1984: (II.)lvi–lxii, and Maas forthcoming.

24

In BC 12.14–42, Arāḍa teaches his theoretical tenets (siddhānta, BC 12.16b), the conclusions (niścaya) of his own śāstra (on the arising and the end of saṃsāra, BC 12.15d) to the bodhisattva. Aśvaghoṣa also presents Arāḍa’s doctrine as a darśana (BC 12.13c) and a śāstra (BC 12.10b, 15d). In BC 12.45–67, the sage expounds the practical aspects of his doctrine of salvation, its soteric means (abhyupāya, BC 12.43c; upāya BC 12.66a) consisting in brahmacarya (BC 12.42c, 44a).

25

See BC 12.69–82.

26

BC 12.83: iti dharmam arāḍasya viditvā na tutoṣa saḥ / akṛtsnam iti vijñāya tataḥ pratijagāma ha //. Translation Johnston 1984: (II.)181.

10

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

the latter’s doctrine 3. on the basis of philosophical arguments instead of practical experimentation. It is because first- to second-century Sāṅkhya philosophy—an āgama or śāstra (“treatise”) among several others in competition—fails to stand critical examination that it does not qualify as a soteriologically relevant system, that it is a non-path (amārga SNa 3.3) and is accordingly to be rejected as unsatisfactory.27 Though much shorter, Aśvaghoṣa’s depiction of Gautama’s dissatisfaction with Udraka’s teachings displays very similar features and explicitly alludes to vicāra: “But, having grasped [Arāḍa]’s speech and [thoroughly] examined it, the [bodhisattva] replied [as follows].”28 As for Śākyamuni’s abandonment of severe asceticism, it conforms to a similar pattern: the sage is shown reflecting upon (anuCINT BC 12.102) the value of austerities, providing a row of reasoned arguments against them,29 and finally concluding (niścaya BC 12.107) that the salvational means can only be based on one’s intake of food (āhāramūlo ’yam upāya iti BC 12.107). This provides the rational justification for the bodhisattva uttering the now well-known formula: “This is not the way of life for passionlessness, for awakening, for liberation.”30 Aśvaghoṣa’s representation of the (future) Buddha as subordinating rational choice to a critical examination of the competing salvational systems can be seen at work on several other occasions in the BC 31—and 27

The reason why the bodhisattva condemns Arāḍa’s and Rudraka’s doctrines is that they involve a substantialist account of, and a belief in the self. Note BC 12.84cd: ātmagrāhāc ca tasyāpi jagṛhe na sa darśanam //. “And [it is] because [it entailed] the belief in a self that the [bodhisattva] did not accept [Udraka]’s [philosophical] perspective either.”

28

BC 12.68ab and d: iti tasya sa tad vākyaṃ gṛhītvā tu vicārya ca / […] pratyuttaram uvāca ca //.

29

See BC 12.103–106.

30

BC 12.101ab: nāyaṃ dharmo virāgāya na bodhāya na muktaye /. Translation (slightly modified) Johnston 1984: (II.)184.

31

See, e.g., the critiques of fire sacrifice (agni) in BC 16.55–61 (see Johnston 1984: [III.]20 and Eltschinger forthcoming a, §3.3); of the self in BC 16.80–89 (Johnston 1984: [III.]22–23 and Eltschinger 2013: 176–185); of īśvara in BC 18.18–29 (Johnston 1984: [III.]32–33); of nature in BC 18.30–41 (Johnston 1984: [III.]33–35); of time in BC 18.42 (Johnston 1984: [III.]35); of substance (dravya) in BC 18.13–46 (Johnston 1984: [III.]35–36); of puruṣa in BC 18.47–51 (Johnston 1984: [III.]36); of fortuitousness (ahetukatva) in BC 18.52–56 (Johnston 1984: [III.]36–37). See also the discourses on kings and politics in BC 20.15–52 (Johnston 1984: [III.]50–55) and on women in BC 22.21–35 (Johnston 1984: [III.]65–66).

INTRODUCTION

11

points to Aśvaghoṣa as one of the most significant early Buddhist philosophers, as was already suggested by Yijing.32 This picture, however, is also part of the poet’s wider concern with the (future) Buddha as challenging wrong systems and converting their representatives. This can be seen, e.g., in the Buddha’s dialogue with the Sāṅkhya-oriented “philosopher”33 (parīkṣaka BC 26.7) Subhadra who, freshly converted by the Buddha, claims that “previously he had held birth to be by nature, [and] now he saw that there was no salvation in that [doctrine],”34 or that “previously he had held with respect to that which is manifested (vyakta) that the self is other than the body and is not subject to change[, and that] now that he had listened to the sage’s words he knew the world to be without self and not to be the effect of self.”35 Aśvaghoṣa’s representation of the bodhisattva’s entire career is encapsulated in BC 25.9, according to which the Buddha, “after refuting the allodox erroneous paths, proceeded on such a path that

32

Together with Nāgārjuna and (Ārya)deva, Aśvaghoṣa is listed among the Buddhist doctors “of an early age” by Yijing. See Takakusu 1966: 181 and Krasser 2012a: 583– 585.

33

BC 26.7ab (P107a8–b1): / kho bo lta bu’i yoṅs su rtog pa po rnams las // thar pa’i lam gźan lhag par thob ces grag /. “It is said that you have gained a path of salvation other than that of philosophers (parīkṣaka) like myself.” Translation (slightly modified) Johnston 1984: (III.)92.

34

BC 26.15ab (P107b8–108a1): / sṅon du des ni raṅ bźin las ni skyes sñam ste // des kyaṅ der ni thar pa mthoṅ ba ma yin źiṅ /. Translation (slightly modified) Johnston 1984: (III.)93.

35

BC 26.17 (P108a3–4): / gsal ba ’di yin rnam par ’gyur ba med pa’i bdag // lus las gźan ni de yis sṅon du rtogs pa ste // thub pa’i gsuṅ ni yoṅs su thos& nas ’jig rten ni // bdag med pa daṅ bdag tu bya ba med par śes /. &thos em.: thob P. Translation (slightly modified) Johnston 1984: (III.)93. This can also be seen, although in an indirect way, in Aśvajit’s conversion of Upatiṣya (described in BC 17.4 [P75b7] as ser skya’i rigs kyi dge sloṅ slob ma du ma can, i.e., “a mendicant of Kapila’s sect, who had many pupils” [translation Johnston 1984: (III.)24]) in, e.g., BC 17.10 (P76a5–6): / rgyu med pa daṅ bya ba med daṅ dbaṅ phyug ste // źiṅ śes pa ni ṅes par de yis sṅon rig ciṅ // rgyu las ’di rnams rab tu ’jug par thos nas ni // bdag med gyur pa mthoṅ nas dam pa’i bden pa gzigs /. “Previously he had held the theory that the field-knower (kṣetrajña) is uncaused, inactive, and the originator (īśvara); on hearing that all these things take place in dependence on causes, he perceived that there is no self and saw the supreme truth.” Translation Johnston 1984: (III.)25.

12

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

he [could] teach the right path (*sanmārga?).”36 In short, “in Kāśī he turned the wheel of the Law and by being [so] judicious brought content to the world; he caused those who were to be converted to practise the way of the Law, and brought bliss to us for our good. Others he caused to see the real truth that they had not yet seen, and he united the followers of the Law with the virtues. By refuting the other systems and by argument he caused men to understand the meaning which is hard to grasp. By teaching everything to be impermanent and without self and by denying the presence of the slightest happiness in the spheres of existence, he raised aloft the banner of his fame and overturned the lofty pillars of pride.”37 Proclaiming the final truth of impermanence, selflessness and painfulness does not go without critically examining and overcoming (*niGRAH?) competing religio-philosophical claims. 2.2. The Savitarkasavicārādibhūmi of the YBh (itself a section of the YBhŚ) provides us, in its Paravāda (“allodoxy”) section,38 with a fascinat36

BC 25.9ac (P103a6–7): / lam log mu stegs rnams pham nas // ’di ’dra’i lam la gśegs nas su // gaṅ gis dam pa’i lam bstan te /. Translation Johnston 1984: (III.)83. (The senāpati Siṃha is speaking.)

37

BC 27.30–32 (P115b7–116a2): / kā śi nas su chos kyi ’khor lo bskor ba ste // blo daṅ ldan pa ñid kyis ’jig rten tshim mdzad ciṅ // ’dul bya rnams ni chos kyi cho gar spyad mdzad la // bdag cag rnams la phan pa’i ched du dge ba mdzad // gźan rnams ma mthoṅ yi de ñid mthoṅ mdzad ciṅ // yon tan rnams kyis chos byed rnams la sbyar ba ste // gźan gyi lugs rnams tshar bcad nas ni bskul gyur la // rtogs par dga’ ba’i don ni rnam par śes par mdzad // thams cad mi rtag pa daṅ bdag med par gsuṅs śiṅ // srid pa rnams su bde ba cuṅ zad ma gsuṅs la // sñan pa’i ba dan gyen du rnam par bsgreṅ ba ste // ṅa rgyal ka ba mthon po rnam par bskyel bar mdzad /. Translation (slightly modified) Johnston 1984: (III.)108.

38

YBh 118,1–160,9, YBhMS 33b5–44a5, YBhT D60b2–81a1/P71a3–93b4. YBh 118,5, 118,6, 160,8. To the best of my knowledge, the YBh provides neither an analysis nor an explanation of the compound paravāda. I take para to be used here in the same sense as its Pali equivalent in the BJSū (DN I.3 and passim; see Eltschinger forthcoming c, §1.3 and n. 36), i.e., to refer to a representative (śramaṇa or brāhmaṇa) of any non-Buddhist group and/or doctrine. In this sense, para is very close in meaning to (anya)tīrthya/tīrthika, the “allodox” (rather than heterodox/“heretic”) teacher, but insists on its function as an at least virtual opponent. The author(s) and/or compiler(s) of the YBh was/were probably aware of the problem raised by the presence, among the sixteen allodoxies, of the obviously Buddhist (ihadhārmika, Tib. chos ’di pa) sarvāstivāda. See Eltschinger forthcoming c, §2.2.3 and also §2.2.13 of a Buddhist (Mahāyānist) version of “nihilism.” On issues of orthodoxy, heterodoxy/heresy and allodoxy within Buddhism, see below, Chapter 1, p. 36, n. 3, and Eltschinger/Ratié 2013: 65, n. 57.

INTRODUCTION

13

ing outline of early Buddhist philosophy in concreto and another early testimony of the use of parīkṣā in a scholastic context. This section is remarkable in several respects. First, contrary to the BJSū and the ŚPhSū, to which it is confessedly indebted,39 it does not limit itself to listing and condemning sixteen false views qua false views in a doxographical manner, but provides sophisticated arguments against each of them. Second, contrary to the overwhelming majority of prior and contemporary Buddhist polemical works, which focus on coreligionists’ views (the pudgala, the existence of the three times, etc.) in an Abhidharma-like manner (notably by resorting to the yukti-āgama methodology),40 this section targets, by means of reason(ing) (yukti) alone, the most prominent representatives of third- to fourth-century Indian philosophy: early Sāṅkhya (satkāryavāda), (Vaiśeṣika?) atomism, Brahmanism (self, creator God, ritual violence, etc.), the Buddhist Sarvāstivāda, Jainism as well as several allodox views already mentioned in the BJSū (eternalism, annihilationism, etc.) and the ŚPhSū (Jainism, Buddhist as well as non-Buddhist “nihilism,” etc.). Third, the Paravāda section addresses non-Buddhist practices and institutions (ritual, purity, caste classes, etc.) in addition to purely theoretical tenets, and thus it echoes Aśvaghoṣa’s way of submitting both philosophical doctrines (early Sāṅkhya, self, creator God, etc.) and religious practices (asceticism, Vedic ritual) to sustained critical examination. Here are the sixteen allodoxies: “These allodoxies [amount to] sixteen, i.e.: (1) the doctrine [according to which] the effect [pre]exists in [its] 39

The Paravāda section confessedly draws upon the BJSū of the Dīrghāgama in its doxographic account of at least five allodoxies: eternalism, “extensionism,” “eel-wriggling,” fatalism and annihilationism. The expression yathāsūtram occurs whenever the YBh description is indebted to this Sūtra (see YBh 138,4, 148,3, 150,1, 150,7, 151,3). The YBh’s indebtedness to the ŚPhSū can be observed in (parts of) its account of eternalism (see Eltschinger forthcoming c, §2.2.5), Jainism (Eltschinger forthcoming c, §2.2.6), annihilationism (Eltschinger forthcoming c, §2.2.12) and “nihilism” (Eltschinger forthcoming c, §2.2.13). Important Sanskrit sources are the PrV and the SBhV of the Mūlasarvāstivāda Vinaya (see Eltschinger forthcoming c, nn. 76, 85, 86, 89, 120, 134, 145).

40

See below, Chapter 3, §3.2.2. The present author subscribes entirely to the following remark by Albrecht Wezler (Wezler 1985: 14): “For this literature [i.e., early Buddhist literature, VE] is characterized by the fact that most, if not all, debates carried on in the texts start from and centre around internal Buddhist or even Hīnayānistic differences of views and Abhidharma points of controversy. In Buddhist literature it is only gradually that heterodox doctrines are taken notice of […].”

14

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

cause, (2) the doctrine of manifestation, (3) the doctrine [according to which] the past and the future exist as [real] substances, (4) the doctrine of the self, (5) the doctrine of eternal[ity], (6) the doctrine [according to which present suffering] has past deeds as its [sole] cause, (7) the doctrine [according to which entities] such as God are agents, (8) the doctrine [according to which ritual] violence is a [religious] duty(/is righteous), (9) the doctrine of finite[ness]-and-infinite[ness], (10) the doctrine of “eel-wriggling,” (11) the doctrine [according to which things are] without a cause, (12) the doctrine of annihilation, (13) the doctrine of the [universal] deniers, (14) the doctrine [according to which the Brahmins are] the best [caste class], (15) the doctrine of purification, and (16) the doctrine of festivals and auspicious things.”41 The aim of the Paravāda section of the YBh is to criticize these sixteen allodoxies as instances of incorrect reflection (ayoniśomanaskāra) or as false views originating from nescience.42 These allo41

YBh 118,8–12 (YBhMS 33b6–34a1, YBhT D60b3–5/P71a5–8): ṣoḍaśa ime paravādāḥ / tadyathā / hetuphalasadvādaḥ / abhivyaktivādaḥ / atītānāgatadravya&sadvādaḥ / ātmavādaḥ / śāśvatavādaḥ / pūrvakṛtahetuvādaḥ' / īśvarādikartṛvādaḥ( / hiṃsā)dharmavādaḥ / antānantikavādaḥ / amarāvikṣepikavādaḥ* / ahetuvādaḥ / ucchedavādaḥ / nāstikavādaḥ+ / agravādaḥ / śuddhivādaḥ / kautukamaṅgalavādaś ca /. &°gatadravya° em. YBh: °gataḥ dravya° YBhMS. '°hetuvādaḥ YBhMS: °hetusadvādaḥ YBh. (°kartṛvādaḥ em. YBh: °kartṛkavādaḥ YBhMS (this reading is equally good; but YBhMS 40a4 reads kartṛvādaḥ). )hiṃsā° YBhMS: vihiṃsā° YBh. *°vikṣepikavādaḥ YBhMS: °vikṣepavādaḥ YBh. +nāstikavādaḥ YBhMS, YBhT (med par smra ba daṅ): nāstikavādaḥ om. YBh (does YBhMS 34a1, in spite of the akṣara’s dissimilarity with other °kya°, read nāstikya°? Whatever the case may be, YBhMS 42a4 clearly reads nāstika°). The Paravāda section of the YBh has not yet received the scholarly attention it undoubtedly deserves. Of the sixteen allodoxies successively dealt with in this passage, only two (hetuphalasadvāda and abhivyaktivāda) have been (studied, edited and) translated in their entirety so far: see Furusaka 2001 (in Japanese) and Mikogami 1969; on these two sections, see also Wezler 1985. As for the ātmavāda, the īśvarādikartṛvāda, the śāśvatavāda and the nāstikavāda sections, they have been treated, but only in part: see Shukla 1967, Hayashima 1991 (in Japanese) and Eltschinger/Ratié 2013:79–82, n. 111 (ātmavāda), Chemparathy 1968–1969: 86–89 and 94–96 (īśvarādikartṛvāda), Mikogami 1967 (in Japanese) (śāśvatavāda), Schmithausen 2000: 254–259 (nāstikavāda).

42

The concluding stanza of Dharmakīrti’s PV 1/PVSV, viz. PV 1.340, reveals very similar ideas: vedaprāmāṇyaṃ kasyacit kartṛvādaḥ snāne dharmecchā jātivādāvalepaḥ / santāpārambhaḥ pāpahānāya ceti dhvastaprajñāne pañca liṅgāni jāḍye //. “[Believing in the] authority of the Veda, claiming something [permanent, God or the self,] to be an agent, seeking merit in ablutions, taking pride in one’s caste, and undertaking penance to remove sin, these are the five signs of complete stupidity devoid of any discrimination.” Translation (slightly modified) Eltschinger/Krasser/Taber 2012: 77–78. The echo is all

INTRODUCTION

15

doxies are initiated and/or defended by (groups of) ascetics and Brahmins (and twice at least by coreligionists) who are presented as reasoners (tārkika) and investigators (mīmāṃsaka).43 They are attempting, according to the BoBh, to make their points on the basis of scripture and reason(ing), the more striking that, with the exception of vedaprāmāṇya, all of Dharmakīrti’s “signs” already occur in the Paravāda section of the YBh: kasyacit kartṛvādaḥ (note the similarity in wording) ≈ īśvarādikartṛvādaḥ and ātmavādaḥ (see Eltschinger forthcoming c, §§2.2.4 and 2.2.7; on the two traditional interpretations of Dharmakīrti’s kasyacit kartṛvādaḥ, see Eltschinger/Krasser/Taber 2012: 77, n. 171), snāne dharmecchā ≈ śuddhivādaḥ (see Eltschinger forthcoming c, §2.2.15), jātivādāvalepa ≈ agravādaḥ (see Eltschinger forthcoming c, §2.2.14), santāpārambhaḥ pāpahānāya ≈ pūrvakṛtahetuvādaḥ (see Eltschinger forthcoming c, §2.2.6). 43

YBh 119,7–9 (YBhMS 34a2–3 [nearly illegible], YBhT D60b7–61a1/P71b3–4): yuktiḥ katamā / yathāpi tat sa& eva śramaṇo vā brāhmaṇo vā tārkiko bhavati mīmāṃsakas tarkaparyāpannāyāṃ bhūmau sthitaḥ svayamprātibhānikyāṃ pārthagjanikyāṃ mīmāṃsānucaritāyām / tasyaivaṃ bhavati /. &yathāpi tat sa YBhMS (cf. also Wezler 1985: 5) yathā sa eva YBh (YBhT simply reads: ’di ltar [*yathā]). “What does reason(ing) consist of? For instance [here in the world, (a certain/)]this ascetic or (a certain/)this Brahmin is a reasoner, an investigator who remains on a level that belongs to ratiocination, which is based on one’s own wit, [which is] ordinary and pervaded with [philosophical] investigation. The following occurs to him.” The origin of this formula can be sought in the BJSū (DN I.16, 21, 23, 29: idha bhikkhave ekacco samaṇo vā brāhmaṇo vā takkī hoti vīmaṅsī. so takkapariyāhataṃ vīmaṅsānucaritaṃ sayaṃpaṭibhānaṃ evam āha […]. “In this case, brethren, some recluse or Brahman is addicted to logic and reasoning. He gives utterance to the following conclusion of his own, beaten out by his argumentations and based on his own sophistry […].” Translation Rhys Davids 1899: 28–29). The BoBh provides the duly expanded formula, which characterizes the intellectual inclinations and practices of those, most certainly Buddhists but also nonBuddhists, who construe a purely rationally based level of reality. BoBhD 25,15– 19/BoBhW 37,22–38,1: yuktiprasiddhaṃ tattvaṃ katamat / satāṃ yuktārthapaṇḍitānāṃ vicakṣaṇānāṃ tārkikāṇāṃ mīmāṃsakānāṃ tarkaparyāpannāyāṃ bhūmau sthitānāṃ svayamprātibhānikyāṃ pārthagjanikyāṃ mīmāṃsānucaritāyāṃ pratyakṣam anumānam āptāgamaṃ pramāṇaṃ niśritya suvicitaniścitajñānagocaro jñeyaṃ vastūpapattisādhanayuktyā prasādhitaṃ vyavasthāpitam idam ucyate yuktiprasiddhaṃ tattvam /. “What does reality established by means of reason(ing) consist in? [It is the level of reality that belongs] to [those] wise [persons] who are experts in reasoned matters, sagacious, reasoners [and] investigators [who] remain on a level that belongs to ratiocination, which is based on one’s own wit, ordinary and pervaded with [philosophical] investigation; [this kind of reality consists in] something cognizable that is the object of a well-examined and ascertained cognition, [something] that is demonstrated and determined by [that type of] reason(ing) which proves by means of arguments on the basis of a means of valid cognition, [viz.] perception(/the perceptible), inference [and] trustworthy scripture. This is what [we] call ‘reality established by means of reason(ing).’”

16

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

that is, by depending on the (three) means of valid cognition or pramāṇas (a term that, to the best of my knowledge, never occurs in this technical sense in the entire Paravāda section). This section’s critical stance and methodology are best reflected in its concluding statement: “Thus [it appears that] these sixteen allodoxies are entirely unreasonable once they are evaluated by means of a(/the) twofold (dvayābhinirhāra) [type of] reason(ing) consisting in a [critical] examination.”44 The author(s) and/or compiler(s) of the Paravāda section thus regarded it and its/their polemical endeavor as a parīkṣā.45 2.3. The Sanskrit term parīkṣā provides an important element in the YBhŚ’s (somewhat cryptic) definition of the hetuvidyā, the “science of [justificative] reasons(/evidences).”46 The fact is of primary importance granting the impact of the so-called Hetuvidyā section of the YBhŚ (HV) on later Buddhist philosophers, especially those who are labeled “epistemologists.”47 The HV testifies to an early connection between critical 44

YBh 160,8–9 (YBhMS 44a5, YBhT D80b7–81a1/P93b3–4): itīme ṣoḍaśa paravādā dvayābhinirhārayā& parīkṣāyuktyo'paparīkṣya sarvathā na yujyante /. &dvayābhinirhārayā YBhMS, YBhT (mṅon par bsgrub pa rnam pa gñis): abhinirhārayā YBh. 'parīkṣāyuktyo° em. YBhT (brtag pa’i rigs pa): parīkṣyā yuktyo° YBh, YBhMS.

45

A difficult point remains: What does dvayābhinirhāra (obviously a bahuvrīhi compound, loosely translated above as “twofold”; on abhinirhāra, see BHSD 52b–53a s.v., Schmithausen 1969: 168 [nn. 187 and 191] and Deleanu 2006: II.477 [n. 31]) mean in the present context? Or: what does “two(fold)” refer to in the (very literal) expression “the accomplishment/realization of which is two(fold)”? I am inclined to understand “two(fold)” as pointing to āgama and yukti, and to interpret the whole expression as: “the accomplishment/realization of which is twofold [i.e., bears on the scriptures and reason(ing)s relied upon by each opponent],” and less literally as: “that is to be carried out in a twofold way [by submitting each opponent’s scripture and reason(ing) to the test].” One might, of course, think of this expression as pointing to an examination carried out by means of both scripture and reason(ing)—in line with the argumentative method that is most typical of Buddhist scholasticism. However, (1) the Buddhists were perfectly well aware of the fact that quoting Buddhist scriptures was useless against outsiders; (2) to the best of my knowledge, the Paravāda section does not criticize any allodoxy on the basis of āgama.

46

On the hetuvidyā, see below, p. 166, n. 229, and p. 167, n. 232.

47

See, e.g., Yaita 1999. On this section, see also below, Chapter 2, §2.3.1.3. On its nature and aims, see Eltschinger 2012: 452–473, and note Kang’s insightful remark (Kang 2003: 30): “In der HV ist die Betrachtungsweise letztendlich diejenige eines Bodhisattva, der durch angemessene Benutzung der sprachlichen Mittel die buddhistische Lehre zu verkünden, aber auch zu rechtfertigen anstrebt. Dennoch sind in der Erklärung der

INTRODUCTION

17

examination as a philosophical and (negative) apologetical concern and the hetuvidyā as a normative discourse about dialectics and logico-epistemological theory. Here is the HV definition of the hetuvidyā: “What does the science of [justificative] reasons(/evidences) consist of? It is something that exists for the sake of a [critical] examination. But what does this [something], i.e., dialectics, consist of?”48 Whereas the text remains entirely silent on the definition and the scope of the parīkṣā, it explains “something” (vastu) as “dialectics” (vāda) and thus suggests the latter to be aimed at, and instrumental in, critical examination. In other words, whatever critical examination may consist of and be concerned with, it has hetuvidyā ≈ vāda for its means. Now according to the HV (1,2–4), the hetuvidyā ≈ vāda consists of all the eristic-dialectical items subsumed under the following seven headings: speech (vāda, HV 1,5–4,1), target group of a debate (vādādhikaraṇa, HV 4,2–5),49 basis of a debate (vādādhiṣṭhāna, HV 4,6– 17,1), ornament of a debate (vādālaṅkāra, HV 17,2–20,16), defeat in a debate (vādanigraha, HV 20,17–25,8), getting out of a debate (vādaniḥsaraṇa, HV 25,9–27,7), and the properties held to be variously useful in a debate (vāde bahukarā dharmāḥ, HV 27,8–16).50 This is tantamount to claiming that the logical and epistemological items instantiating the category “basis of a debate,” and especially its subdivision “proof” (sādhana),51 i.e., “the thesis, the reason, the example, similarity, dissimilarity,

einzelnen Unterarten von vāda die eigentlich dialektischen Aspekte nicht vorrangig: die Darstellung orientiert sich zunächst an möglichen Gesprächssituationen für den Bodhisattva […].” 48

For the Sanskrit text of HV 1,1–2, see below, p. 161, n. 214.

49

On the meaning of adhikaraṇa as “target group,” see Kang’s useful discussion in Kang 2003: 154–175 [= Appendix iii].

50

According to HV 27,8–16, these properties are three in number: knowledge of one’s own as well as others’ systems (svaparasamayajñatā), which allows one to debate (kathāṃ KṚ) on every topic (vastu), self-confidence (vaiśāradya, see Kang 2003: 148– 153), which allows one to debate before every audience (parṣad), and eloquence (pratibhāna), which enables one to answer (uttara) any kind of utterance.

51

According to the HV (4,6–7), the basis of a debate is tenfold, i.e., a twofold probandum (sādhya) and an eightfold probans (sādhana). The thing (artha) to be proven may consist of either a nature (svabhāva) or a property (viśeṣa) such as permanent (nitya) or impermanent (anitya), pure (anāsrava) or impure (sāsrava), conditioned (saṃskṛta) or unconditioned (asaṃskṛta).

18

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

perception(/the perceptible), inference, and authoritative scripture,”52 are regarded as the bases or instruments of critical examination. As we can see, those items that were to become key-components of logical theory (pratijñā, hetu, udāharaṇa) and those that would come to be known as the (four, three or two) “means of valid cognition”53 (pramāṇa) now play a significant role in the parīkṣā. The pramāṇas, in particular, are responsible for providing material evidence in favour of a thesis concerning the point to be proven, as the HV makes very clear:54 “What does the reason(/evidence) consist of? It is the justification that, based on an example and drawn from similarity and dissimilarity, or perception, or inference, or scripture, is aimed at establishing the point [at stake] in the thesis.”55 Although the HV’s understanding of the nature, the purpose and the target(s) of the parīkṣā remains unclear (I would be very reluctant to interpret them in the light of the Paravāda section in spite of their belonging to the same corpus), one thing is certain: The HV provides us with the earliest known Buddhist instance of a connection between parīkṣā, hetuvidyā and the pramāṇas. 2.4. At this point, let us pause for a while and attempt a provisional definition of parīkṣā in the early Indian Buddhist context. A critical examination consists in the evaluation, by means of reason(ing) (identified or not to the pramāṇas) and/or scripture (provided the opponent belongs to the same 52

HV 4,15–16: pratijñā hetur udāharaṇaṃ sārūpyaṃ vairūpyaṃ pratyakṣam anumānam āptāgamaś ca /. See also below, p. 159, n. 205.

53

See below, p. 159–159, nn. 205–206.

54

HV 4,17–5,2: pratijñā katamā / dvividhaṃ sādhyam artham ārabhya yo 'nyonyaṃ svapakṣaparigrahaḥ śāstraparigrahato vā svapratibhānato vā parāvajñāto vā parānuśravato vā tattvābhisandhānato vā svapakṣāvasthānato vā parapakṣadūṣaṇato vā parābhibhavato vā parānukampanato vā /. “What does the thesis consist of? It consists in each [debater]’s adoption of his own proposition regarding the point to be proven, which is twofold [i.e., a nature or a property. The adoption of one’s own proposition may proceed] either by adopting a[n already available] treatise, or through one’s own wit, or due to contempt for an opponent, or due to hearsay about an opponent ?, or by interest in the truth, or by sticking to? one’s own proposition, or due to [one’s wish to] refute an opponent’s proposition, or due to [one’s wish to] overcome an opponent, or by sympathy for an opponent.”

55

For the Sanskrit text of HV 5,3–5, see below, p. 159, n. 205. See also CS 3.8.33, where the reason (hetu) is defined as the cause of apprehension (upalabdhikāraṇa), which comprises perception (pratyakṣa), inference (anumāna), (scriptural) tradition (aitihya) und analogy (aupamya). See Preisendanz 1994: 531.

INTRODUCTION

19

confessional denomination56), of an opponents’ theoretical and/or practical tenets in order to assess their rationality and, further, the soteriological relevance of the system as a whole (note that a parīkṣā does not necessarily entail, and in actual fact rarely entails, a systematic exposition of the examinator’s own tenets [svamata]). As the so-called dialectical tradition developed and underwent its “epistemological turn” in the hands of Dignāga,57 critical examination remained true to its nature, purpose and targets, but its association with the theory and practice of pramāṇas grew ever more intimate—so much so that both became nearly indistinguishable. As we shall see in Chapter 2, the epistemological turn of the hetuvidyā coincided with a shift in the polemical targets and audience of the Buddhist dialecticians, a shift that is mirrored in the titles of several works authored by Dignāga, viz. the Nyāyaparīkṣā (“[Critical] examination of the Nyāya [system]”), Vaiśeṣikaparīkṣā (“[Critical] examination of the Vaiśeṣika [system]”), and the Sāṅkhyaparīkṣā (“[Critical] examination of the Sāṅkhya [system]”),58 which may well have consisted in monographs targeting, on the likely model of works such as Vasubandhu’s Paramārthasaptatikā,59 the (epistemo)logical theories (and religio-philosophical doctrines?) of these non-Buddhist schools. 2.5. There is at least one hint to the fact that Dignāga’s philosophy amounted to a parīkṣā in the mind of Dharmakīrti. In this oft-quoted passage, Dharmakīrti outlines, allegedly in the footsteps of Dignāga, his posi-

56

Christian apologists such as Alan of Lille (d. 1202), Thomas Aquinas (1225–1274) or Philippe de Mornay (1549–1623) also showed a clear consciousness of the fact that resorting to Christian scriptural authorities was of no argumentative value against the Muslims and the pagans (unlike Jews and heretical Christians); see Dulles 2005: 110, 115 and 157 respectively. In particular, Philippe de Mornay “[…] calls attention to the necessity of arguing from principles that are accepted by the adversary as well. In the case of pagans one may appeal to self-evident principles and to demonstrable philosophical truths; in the case of Jews, to the Old Testament” (Dulles 2005: 157). This aspect of apologetical methods has also been emphasized by Griffiths (1991: 15): “Apologetics also usually uses only methods of argumentation and criteria of knowledge acceptable to the adversary. This is to rule out, among other things, appeals to sources of authority not recognized by one side in the debate.” Griffiths (1991: 82) refers to these scriptural sources as “community-specific self-guaranteeing authority sources.”

57

See below, Chapter 2, §2.3.1.

58

See below, p. 157, nn. 195 and 196.

59

See below, p. 157, n. 196, and p. 175, n. 251.

20

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

tion concerning the appraisal of scriptural authority (āgamaprāmāṇya): “The person [who wishes to engage in religious practice] cannot live without resorting to scriptural authority[, and this for two reasons: first,] because [it is only in scripture that this person] learns the great benefits and evils [that are to be expected from] engaging in and refraining from certain [actions/intentions] whose results [remain entirely] imperceptible [to him/her; and second,] because [this person] does not see [anything] contradictory to the existence of these [desirable or undesirable results]. Thus if [this person] is [necessarily] to act [on a scriptural basis], it is better that (s)he act in this way [i.e., after evaluating scripture, and] this is the reason why [Dignāga recommends that scriptural] authority [be decided] through [critical] examination […] The [treatise]’s reliability consists in the fact that neither perception nor the two kinds of inferences invalidates the empirical or transempirical things [that are] their [respective] objects. [A treatise’s] not being invalidated by perception consists [first] in the fact that the things it holds to be perceptible are indeed such [i.e., perceptible], as [the five skandhas, i.e., colours] such as blue, [affective sensations such as] pleasure and pain, [ideation consisting in one’s] grasping the characteristics [of things, conditioning factors] such as desire, and cognitions[, which are all perceived by sensory perception and self-awareness. Second, a treatise’s not being invalidated by perception consists] in the fact that the [things] it does not hold to be such [i.e., perceptible,] are [indeed] imperceptible, as [pseudo-constituents] such as pleasure, which [the Sāṅkhya erroneously takes to] combine in the form of sounds, etc., and [categories] such as substances, motions, universals and connections[, which the Vaiśeṣika erroneously takes to be perceptible]. Similarly, [a treatise’s not being invalidated by inference] consists [first] in the fact that the [things] it holds to be the objects of an inference that does not depend on scripture are really such [i.e., inferable], as the four nobles’ truths, [and second] in the fact that the [things it holds to be] non-inferable are really such [i.e., noninferable], like the self, [God,] etc. [And this type of invalidation is] also [relevant] concerning an inference that depends on scripture[, which consists in identifying internal contradictions within a treatise]: For example, once it is admitted that demerit has the nature of [defilements] such as desire and the [corporeal and verbal acts] that originate from them, one does not prescribe [things] such as ablutions and fire oblation in order to remove it [i.e., de-

INTRODUCTION

21

merit, because they cannot annihilate its cause].”60 This has generally been interpreted, and rightly so, as reflecting (the first part of) Dharmakīrti’s doctrine of scriptural authority, what Tibetan scholastics called the “threefold analysis”61—something that could be, and as a matter of fact has often been, considered a mere appendix to the master’s logic and epistemology.62 In my opinion, however, this important passage can and indeed should also be interpreted as reflecting Dharmakīrti’s understanding of his own philosophical enterprise. For if what is at stake here is a given scripture’s or treatise’s (Sāṅkhya, Vaiśeṣika, Buddhism) reliability in transempirical, suprarational matters, the method prescribed in order to assess this reliability is nothing but philosophy itself in the form of the critical examination, carried out by means of the pramāṇas, of the empirically and rationally accessible tenets of the contemporary salvational systems,63 and this, as in

60

For the Sanskrit text, philological and bibliographical remarks, see below, p. 221, n. 80, and p. 212, n. 52.

61

As already pointed out by Tillemans (1993: 9–24) and Keira (2006), Dharmakīrti, contrary to his Tibetan followers, does not resort to expressions such *parīkṣātraya to refer to this doctrine. The closest known equivalent to the Tibetan expression dpyad pa gsum is Kamalaśīla’s tshul gsum gyi brtag pa (see MĀ D148b4–5, quoted and translated in Keira 2006: 190[/625], n. 15, and 182[/633]). As Verhagen (2008: 244–247) rightly points out, Dharmakīrti’s doctrine is likely to have been inspired by Vasubandhu’s threefold exegetical method as it appears, e.g., in VY 173,18–177,18 (see Verhagen 2008: 255–258).

62

See below, Chapter 3, §3.2.4, and p. 212, n. 52.

63

This echoes medieval researches and controversies concerning the relationship between faith and reason. Anselm of Canterbury (1033–1109) quite radically advocated the “intrinsic demonstrability of Christian faith” (Dulles 2005: 104). According to him, “faith is objectively rational” (Dulles 2005: 100) because “God […] is supreme truth and hence eminently intelligible; and all that God does is conformed to reason” (Dulles 2005: 100). As Dulles has it, for Anselm, “to understand is nothing other than to grasp the objective reasons, the rationes, that underlie and illumine the data of faith” (Dulles 2005: 100). It is in this sense that Anselm of Canterbury has been regarded by some as a rationalist because “the deepest mysteries of faith are in his system accessible to unaided reason” (Dulles 2005: 104). In a somewhat comparable way, Peter Abelard (1079–1142) “maintained that human reason, making use of objectively accessible evidences, could achieve some kind of inchoative faith, paving the way for the supernatural act of faith elicited under the influence of grace and charity” (Dulles 2005: 107). Dispensing with reason would result in various dangers and absurdities; moreover, “[a]t the very least, […] reason is needed to select what authority one is going to follow” (Dulles 2005: 108)—a conviction that all the early medieval Indian philosophers would at least rhetorically

22

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

endorse (see below, pp. 195, n. 17). Thomas Aquinas (1225–1274) came up with a more balanced position of compromise distinguishing between “two sets of truths—those accessible to reason alone (praeambula fidei) and those inaccessible without divine revelation accepted by a supernatural act of faith” (Dulles 2005: 120). The first set of truths includes the existence of a creator God, the end of the world, properties of God like “eternity, immutability, omnipotence, and omniscience” (Dulles 2005: 120; see also Michon 2000: 9–10), and is liable to “stringent proofs from intrinsic reasons” (Dulles 2005: 127; these rationes internae are otherwise described as “necessitating” or “demonstrative internal arguments” [Dulles 2005: 127]). The second type of truths— “strictly supernatural truths” (Dulles 2005: 127) or “truths that are beyond the investigative power of human reason” (Dulles 2005: 116; rationes convenientiae, rationes probabiles)—includes the Trinity, the Incarnation, the sacraments, the resurrection of the body, the last judgment, and the consummation of the world; here, “the reasonableness of faith turn[s] chiefly on extrinsic evidences” (Dulles 2005: 127) of a purely probable character, such as scriptural authority, miracles, the prediction of future contingent events; these “extrinsic signs of credibility, while they do not directly prove the truth of any individual articles of belief, guarantee the contents of the faith in general” (Dulles 2005: 118) and give “at least a well-founded probability” (Dulles 2005: 131). This provides the basis of the systematic distinction between “intrinsicist” and “extrinsicist apologetics.” Thomas Aquinas’ distinction is reminiscent of Dharmakīrti’s differentiation between yuktiviṣaya and ayuktiviṣaya, i.e., between states of affairs (rather than truths) that are or are not within the scope of reason(ing) (see below, Chapter 3, §§3.2.1.3 and 3.2.3.4, and p. 206, n. 41). Whereas propositions bearing on the first type of “objects” can be verified/falsified, and this in terms of correspondence, those bearing on the second type of “objects” are neither verifiable nor falsifiable, but liable to be evaluated as regards their internal coherence/consistence. It is to be noted that Dharmakīrti, although his attitude cannot be adequately described as “rationalistic,” leaves very little room for purely scriptural “objects,” most of which (the efficiency of mantras, cosmology, the details of karmic retribution, etc.) have no real significance for soteriological understanding and practice. According to Dharmakīrti, the properly salvific contents of the Buddhist faith can be demonstrated by unaided reason. As for the relevance of the first type of truths for apologetics, Dharmakīrti’s position is certainly close to Thomas’ (Michon 2000: 10): “Toutes ces conclusions sont, d’après Thomas, établies par la raison, et seulement confirmées par la révélation biblique. Voilà qui lui permet implicitement de dire que la seule raison naturelle parvient à réfuter non seulement les erreurs des philosophes, mais aussi celles des religions non chrétiennes, le judaïsme et l’islam.” Note also Griffiths 1991: 63: “Many theological thinkers regard positive apologetics as a branch of natural theology. This seems quite proper; if we think of natural theology as the branch of theological thought that operates independently of revelation by the application of unaided human reason to the solution of difficult theological and other intellectual problems, then positive apologetics quite clearly belongs to it. Positive apologetics […] is a discourse that uses only methods of argumentation and criteria of knowledge that are acceptable to the adversary. This limitation was designed to rule out appeals to sources of revelation not accepted by those beyond the bounds of the

INTRODUCTION

23

Aśvaghoṣa’s BC, with a view to appraising their soteriological potential. In Dharmakīrti’s opinion indeed, a śāstra is primarily if not solely expected to propose a human goal (puruṣārtha), together with a practicable expedient (anuguṇopāya), in a coherent way (sambaddha).64 If it is so, philosophy qua critical examination, far from exhausting itself in a self-contained autonomous inquiry into the nature of cognition, language, logic, or being, serves the purpose of rationally evaluating, on the empirical basis of such an inquiry, the competing (and concurrent) treatises as to the probability of their being soteriologically relevant. Dharmakīrti’s critical examination targets Sāṅkhya cosmology (the doctrine of the three guṇas) and Vaiśeṣika ontology (the doctrine of the “categories”), the Brahmanical notions of the self (ātman) and a creator God (īśvara) as well as practices such as ablutions ([tīrtha]snāna), fire oblation (agnihotra), fasting (upavāsa) and various other religious observances (vrata).65 And as we shall see in Chapter 3, the only known occurrence of this method in Dharmakīrti’s writings, in and ad PV 1.332–334, targets the Veda.66 But his critical examination also bears on the Buddhist four nobles’ truths (caturāryasatya, āryasatyacatuṣṭaya) and the five constituents (skandha). Thus on the one hand, Dharmakīrti searches for (and of course finds) faults in the allodox teachings (those alluded to here are meant merely as examples); on the other, he identifies truths in the most important Buddhist doctrines. Now this evaluation strategy—and for this very reason philosophy qua critical examination—amounts to one of the two modalities traditionally ascribed to apologetics, viz. positive apologetics. 3.1. What does apologetics consist of? The word, which derives from the Greek expressions ἀπολογέομαι (“to defend”) and ἀπολογία (“defence”), could be defined in minimal terms as “the defence of a religion against actual or perceived opponents,”67 or the “effort to secure faith by rational

religious community engaging in positive apologetics; its inclusion makes it very clear that positive apologetics is a branch of natural theology in the usual sense of that term.” 64

See PV 1.214, PVSV 108,9–15, Dunne 2004: 361–362, Yaita 2005b: 444, Eltschinger 2007a: 102–104 and 220–221.

65

See Eltschinger 2007a: 108–109.

66

See below, Chapter 3, §3.2.4.2, and Eltschinger/Krasser/Taber 2012: 65–71.

67

Edwards et al. 1999: 1.

24

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

proofs,”68 on the model of the early Christian apologists whose aim it was to “give arguments for the truth of Christianity,”69 to “giv[e] reasons in support of Christianity.”70 According to Griffiths, apologetics concerns “the ways in which religious intellectuals typically think about the claims to truth espoused by their own communities, and […] the ways in which they typically think about the claims to truth expressed by other communities.”71 Apologetics thus basically reflects and responds to a context of religious pluralism, of competing religious claims to truth, and is “occasioned by a particular encounter, a particular realization on the part of the representative intellectuals of some religious community that the doctrine-expressing sentences72 to which they assent are challenged […] by those of some other religious community.”73 And indeed, “since religious communities characteristically assert the ordered sets of sentences that express their doctrines because they take them to be true, it is part of their epistemic duty to consider whether a challenging sentence or set of such makes it improper to continue asserting what the community asserts. This will usually initially involve engagement in some form of negative apologetics—an attempt to see whether the competing assertion fails in its claims—but it may often (and should) pass from there into positive apologetics—the attempt to show not only that the attack fails, but that the doctrines of the community under 68

Dulles 2005: 308.

69

Dulles 2005: 60.

70

Dulles 2005: 61.

71

Griffiths 1991: 1. Griffiths often resorts to the expression “representative intellectuals of religious communities” (Griffiths 1991: 3, et passim; for a definition, see Griffiths 1991: 6–7).

72

By “doctrine-expressing sentences,” Griffiths (1991: 9) means “sentences in some natural language (whatever the language of the community happens to be), which are taken by the community either to make or to entail claims about the nature of things, or claims about the value of certain courses of action; these sentences must also be regarded by the community as of some significance for its religious life, and for the salvation of its members.” According to Griffiths 1991: 20, “[r]epresentative intellectuals of religious communities have typically tended to think of the sentences they construct as being capable of comprehension outside their own community; as extending their claims beyond the bounds of their own community; and as being, simply, both true and important.” As noted by the same author (ibid.), “[i]t seems clear enough that Buddhists have typically regarded the four noble truths as possessing universal truth, and as being comprehensible to, applicable to, and desirable for non-Buddhists.”

73

Griffiths 1991: 78.

INTRODUCTION

25

attack are cognitively superior to those of its challenger.”74 As this quotation suggests, apologetics is traditionally viewed to proceed according to two distinct modalities, viz. positive and negative.75 According to Griffiths again, “[p]ositive apologetics […] is a discourse designed to show that the ordered set of doctrine-expressing sentences constituting a particular religious community’s doctrines is cognitively superior […] to that constituting another religious community’s doctrines.”76 In other words, “[t]hrough positive apologetics, [a given community] tries to show that its doctrines are superior, cognitively, epistemically, or ethically, to those of competing religious communities,”77 and this is most certainly what Dharmakīrti’s parīkṣā, as those of his predecessors, is concerned to demonstrate. The examples adduced by Dharmakīrti himself—the five constituents and the four nobles’ truths—are illuminating. For claiming the perceptibility of the five skandhas amounts to vindicating that they are ultimately real, provide a better description of the world than the concurrent but incompatible truthclaims, and hence that the Buddhist analysis of being is true (in traditional dogmatics, the five constituents provide an exhaustive account of conditioned being). In much the same way, the four nobles’ truths entail (and actually are) descriptive claims about the nature of things in addition to outlining the salvational path—right morality, right psychic concentration, and right insight. That the truths do indeed describe the basic structure of reality is made clear by the Abhidharmic doctrine (to which Dharmakīrti explicitly assents) according to which each and every existent conforms to, or “possesses,” the sixteen aspects of the four truths, hence is impermanent/momentary, selfless, painful, empty, etc.78 In other words, traditional Buddhist doctrine describes reality in a much more satisfactory way than Sāṅkhya or Vaiśeṣika, whose tenets can be demonstrated as theoretically erroneous and practically deceptive. And the same holds true for the Buddhist rejection of a self or a creator God as well as of Buddhist soteriology which, instead of enjoining ablutions or fire oblations in order to get rid of 74

Griffiths 1991: 15–16.

75

See Griffiths 1991: 14–15. Note also Griffiths 1991: 14: “Where negative apologetics is defensive, positive apologetics is offensive; where negative apologetics mans the barricades, positive apologetics takes the battle to the enemy’s camp.”

76

Griffiths 1991: 14.

77

Griffiths 1991: 60.

78

See Eltschinger forthcoming c, §2, and below p. 276, n. 100.

26

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

the defilements, prescribes the mental cultivation of (the intuition of) selflessness, because contrary to these physical practices, it counteracts the (purely psychological) cause of the defilements, i.e., the false view of a self or personalistic belief. 3.2. Now, is there anything like negative apologetics in Dharmakīrti’s writings? Recall that “in this mode, apologetical discourse is designed to show that a given critique of (any one of) the central truth-claims expressed by an ordered set of doctrine-expressing sentences fails, or that a critique of the set as a whole (as to, say, its consistence or coherence) does the same.”79 Negative apologetics, which is often resorted to in order to demonstrate the lack of incompatibility between sentences, “consists in a series of defensive manoeuvres through which a given community defends its doctrines from external attack.”80 Needless to say, negative apologetics permeates Dharmakīrti’s works, but finds its paradigmatic expression in Dharmakīrti’s second—or rather, alternative—evaluative strategy as it is spelled out in PV 1.217 and PVSV 109,15–19: “Or, because the true nature of what is to be eliminated and what is to be realized, together with [their respective] means, is ascertained (prasiddhi) [by means of a pramāṇa based on the force of something real], the [treatise’s] principal point is reliable. Therefore [even when it bears] on another[, transempirical matter, the cognition derived from this treatise is] an inference. The reliability [of this treatise] consists in the fact that what is to be eliminated, what is to be realized, as well as their [respective] means as [they are] taught in it are nonbelying [when they are submitted to inferential analysis], as the four nobles’ truths [when they are analyzed] according to the method that will be presented [in the second chapter]. And since all that which serves a human goal and is [therefore] worthy of endeavor is reliable, to admit that it is such [i.e., reliable,] on another[, transempirical] matter also, [can]not be deceiving[, and this for two reasons: first,] because there [can] be no damage [in accepting it,] and [second,] because the speaker[, after having said what is true concerning the principal point,] has no [personal] interest in purposelessly speaking untruth [concerning this third kind of objects].”81 As we can see, Dharmakīrti again insists on the four nobles’ truths in the 79

Griffiths 1991: 14.

80

Griffiths 1991: 60.

81

For the Sanskrit text, philological and doctrinal remarks as well as a bibliography, see below, p. 214, n. 61, and Chapter 3, §3.2.4.3.

INTRODUCTION

27

way in which, he says, they are going to be dealt with in PV 2. Dharmakīrti refers here to PV 2.145–279,82 which consists of a rational defense of the four nobles’ truths against (a variety of) materialism (which vindicates the impossibility of rebirth, hence of the path),83 against Mīmāṃsā (which claims that human beings cannot go beyond their “natural” determinations, hence are incapable of bringing mental properties such as compassion and insight to their highest and salvific degree),84 against Śaivism (which advocates Tantric initiation as a self-sufficient and measurable salvational means),85 against Sāṅkhya and Nyāya (according to which a real, enduring, substantial self obtains liberation),86 etc. This is of course just an example, for Dharmakīrti’s defense of selflessness against Sāṅkhya, Nyāya and Vaiśeṣika,87 or his heated—and repeated—defense of momentariness against the attacks of the Nyāya and the Mīmāṃsā,88 are best interpreted as negative apologetics on behalf of the pillars of Buddhist dogmatics. This is in fact hardly surprising given earlier or contemporary statements concerning the purpose of the hetuvidyā and/or dialectics.89 The MSA(Bh) characterizes the hetuvidyā as an essentially polemical science (as is also the śabdavidyā) aimed at defeating others (anya) defined either as opponent debaters (paravādin) or as unbelievers (anadhimukta). This statement makes any pronouncement as to the type of apologetics involved difficult. That the hetuvidyā entails both positive and negative attitudes seems to be reflected in the BoBh’s first statement on this science, according to which the hetuśāstra reveals the advantage resulting from criticizing (upālambhakathā) others and escaping (vipramokṣa) the critique of one’s opponents. Sthiramati makes it very clear while commenting on the MSA(Bh) statement in the light of the BoBh (and most probably Dignāga): “[Opponents] such as the outsiders state that the means of valid cognition,

82

For an overview, see Vetter 1990: 52–168.

83

See below, p. 173, n. 243.

84

See below, p. 173, n. 244, and Eltschinger 2005b: 182–184 and 2007a: 232.

85

See below, Chapter 2, §.2.2.6.1–2.

86

See Eltschinger/Ratié 2013: 187–283.

87

See Eltschinger/Ratié 2013: 117–173.

88

See, e.g., Steinkellner 1968–1969, Yoshimizu 1999, Yoshimizu 2011 and Sakai 2011– 2012.

89

See below, Chapter 2, §2.3.1.4.

28

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

direct perception and inference are [valid] reasons establishing the self, permanence, etc.; by means of the [hetuvidyā, the bodhisattva] knows that [these outsiders] make mistakes that do not fit the means of valid cognition. By [himself resorting to] the means of valid cognition, direct perception and inference, he defeats rival debaters such as the outsiders who criticize the Law of the Buddha, and criticizes the outsiders’ works that teach [the existence of] the self.”90 The dual apologetical agenda of the hetuvidyā can hardly be stated more clearly. Finally, protecting the saddharma is one of the purposes of dialectics according to the *UH.91 3.3. This leaves many questions open, among them those of the addressees and purposes of early epistemological works92—outsiders or insiders? Purely polemical or (at least also) edificatory and catechetical? The question has been much debated in the fields of early Christian apologetics and Buddhist epistemology. As Jacobsen has it, “[i]n recent scholarship on apologetics, defence is still seen as the main characteristic of apologetics, but it is often stressed that the defence against attacks from outsiders is directed inwards as a way to strengthen members of the attacked group as much as it is directed outwards against the critics.”93 In other words, “many of these apologetic texts which are characterised by their defensive elements at the same time have a very offensive and constructive character. It 90

For the Tibetan text, see below, p. 169, n. 237.

91

Note that besides protecting Buddhism and proselytizing the non-Buddhists (a purpose also hinted at in the BoBh’s second statement on the hetuvidyā, see below, pp. 167–168, nn. 232–233), the *UH also insists on the compassion at work in positive apologetics (see below, Chapter 2, §2.3.1.4), a dimension that is not altogether absent in the Buddhist epistemologists’ understanding of their own endeavor (see below, Chapter 2, §2.3.2). Concerning upāyakauśalya and karuṇā, note Griffiths 1991: 62: “Engagement in positive apologetics will […] be regarded as desirable only if it also seems that it might help those at whom it is aimed to approach more closely to that condition of dispassionate compassion toward which Buddhist soteriology ultimately points all human beings. In order to be helped in this way by having a positive apologetic aimed at one, one would presumably have to be obsessively entangled in false intellectual views. Since, according to standard issue Buddhist diagnoses of the human condition, that is the condition in which most of us are mired, it is perhaps not surprising that Buddhist intellectuals have, traditionally, engaged in a wide variety of intellectual practices that fit very appropriately under the heading of positive apologetics.”

92

There can be no question of repeating here what I shall say below (Chapter 2, §2.3.2) on the Buddhist epistemologists’ self-understanding.

93

Jacobsen 2009: 7.

INTRODUCTION

29

is obvious that Jews, Christians and adherents of other religious beliefs in antiquity and late antiquity lived in a context where they had to define and defend their beliefs and ways of life in relation to many kinds of criticism and attacks. It is also clear that most religious or cultural groupings at that time had no clear-cut definition of their beliefs and ways of life. Their religious and cultural identities were under construction or reconstruction. Apologetics can best be seen as part of these processes of self-definition and definition of others. When a religious community defends and presents its own ideas against criticism from the outside, this group is, at the same time, embarking on the process of self-definition and self-construction.”94 In other words, could Buddhist epistemological literature have at least also addressed coreligionists in need of arguments either in order to dispel their personal doubts or to face real opponents? It might be with similar ideas in mind that M. Kapstein and S. McClintock have recently appealed to the work of P. Hadot, an eminent French specialist of ancient Greek and Latin philosophy, in order to provide an overall interpretation of Buddhist epistemology (and especially Śāntarakṣita’s and Kamalaśīla’s TS[P]) as Buddhist philosophia.95 In uncovering the practical, institutional and quasireligious dimensions of ancient Western philosophical traditions, Hadot has paved the way for a renewed interest in philosophy as a spiritual phenomenon. As stressed in Hadot's many works, ancient and early medieval philosophy should not be viewed as a discourse committed to the systematic presentation of doctrines and arguments, but rather as a way of life entailing 94

Jacobsen 2009: 12. Note also Jacobsen 2009: 13–14: “With few exceptions Jewish and Christian apologies are explicitly addressed to readers outside the group to which the author of the apology belongs. One could thus assume that apologetic writings or utterances are directly addressed to those who have raised charges against the group represented by the apologist. Thus it is in some cases […] One could also very well imagine that the writings also had a function within the Christian or Jewish communities, because they could provide Jews and Christians with arguments which they could use when forced to explain and defend themselves […] In such cases [as Origen’s Contra Celsum, VE], the function of apologetics is to strengthen the accused group itself.”

95

See Kapstein 2003: 3–22, McClintock 2010: 14–22, and Eltschinger 2008c: 514–522. For a detailed presentation of Hadot’s views on philosophia, see Eltschinger 2008c: 487–514; for a (partly critical) discussion of the usefulness of the model provided by Hadot’s work, see Eltschinger 2008c: 522–539; for a critical answer to my discussion, see McClintock 2010: 18–22. Although I believe that my good friend Sara McClintock distorted parts of my position (while exhibiting flaws that I am entirely ready to acknowledge), I consider the debate to be closed by now.

30

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

spiritual exercises (discursive as well as non-discursive) that mobilizes the philosopher's whole personality and aims at his transformation. According to Kapstein, Hadot’s conception of philosophy indeed “offers a useful model for our thinking on Buddhism in its relation to philosophy. It is a model that depends upon a view of philosophy that emphasizes the ‘techniques of the self’—systematic spiritual exercises and ascesis, and paideia, here primarily the formation of the person as a moral agent refined through philosophical education.”96 It is to be noted that both Kapstein and McClintock assent to the apologetic dimension of Buddhist epistemological literature. As Kapstein has it, “much of Śāntarakṣita’s text is devoted to the critical scrutiny of doctrines propounded by rival schools […] Now, how does the study of this and its refutation properly belong to the program of self-culture through spiritual exercise that we have been discussing? For on the surface, at least, it appears that Śāntarakṣita’s purpose is primarily to dismiss the teachings of Buddhism’s opponents and to defend the faith; philosophical argument, so it seems, is deployed here primarily in the service of doctrinal apologetics. It is undeniable that this is an entirely legitimate way of reading Śāntarakṣita’s text, and that it succeeds very well in capturing one of the dimensions to which it is directed. Nevertheless, even here I believe that Śāntarakṣita’s dialectical course suggests an alternative reading. As Kamalaśīla insists, the doctrines Śāntarakṣita is concerned to refute all involve ‘mistaken views of the self’ (vitathātmadṛṣṭi). As such, the importance of critical reflection upon them lies precisely in the fact that they are not just others’ views of themselves, but that, potentially at least, they are views that any of us may harbor, whether explicitly or not, with respect to ourselves. Śāntarakṣita’s critical journey through the byways of Indian philosophy is therefore no mere exercise in doxography; rather, it is a therapy whereby one must challenge one’s own self-understandings so as to disclose and finally uproot the misunderstandings that are concealed therein.”97 Interpreted in the light of Hadot’s ideas, Buddhist epistemology, besides developing negative and positive apologetical arguments against its (mainly) non-Buddhist opponents, has the therapeutical98 and transforma-

96

Kapstein 2003: 8–9.

97

Kapstein 2003: 15.

98

Note McClintock 2010: 22 : “Thus, while the application of logic and reasoning does serve the purpose of refuting the grossly inaccurate views of opponents, it also has a

INTRODUCTION

31

tive function of supplying spiritual exercises designed to sharpen the Buddhist practitioners’ ability to identify, fight and neutralize misconceptions by way of arguments and rational inquiry, and hence to strengthen their conviction. As McClintock nicely puts it, “these Buddhists […] recommend and engage in specific exercises of reason—including, and especially, the act of rational inquiry itself—as a central element in a way of life in which one seeks to transform oneself through developing and perfecting wisdom in an effort to remove suffering, that unfortunate state of ignorance and ‘unhappy disquiet.’”99 The epistemologists’ concern to argue for the possibility (sambhava) of the Buddhist salvational system is foreshadowed in several early Yogācāra statements to the effect that the striving bodhisattva should be made aware of the attainability (śakyaprāptitva) or realizability of the Buddhist path, an awareness believed to strengthen his faith and determination.100 In much the same way, the idea according to which epistemological literature directly addresses its non-Buddhist critics has been challenged by H. Krasser, according to whom most of these works, in addition to being written down by students “who were being trained to debate,”101 was destined, first and foremost, to enable students or graduates to meet objections from without and debate “on the marketplace.”102 4. As its subtitle indicates, this book does not deal primarily with the contents and arguments of the above-mentioned apologetical program—with Dharmakīrti’s “religious philosophy,” as I have termed it elsewhere—but with this program’s historical and doctrinal foundations: a socio-historical context of Brahmanical hostility toward pāṣaṇḍas, new patterns of Buddhist self-diction, reinvented models of theoretical and apologetical rationality, and the dogmatic infrastructure underlying Buddhist epistemology. therapeutic function of helping Buddhist practitioners both to remove gross misconceptions and to begin to correct subtle ones.” 99

McClintock 2010: 16.

100

See below, Chapter 3, §3.3.5.2.

101

Krasser 2011: 60.

102

For a well-known Western parallel, note Dulles 2005: 102: “Another benefit that Anselm [of Canterbury, VE] has in mind for his readers is the properly apologetic one, that ‘as far as possible they may always be ready to convince anyone who demands of them a reason of that hope which is in us.’” Note also Dulles 2005: 102: “The aim of theological reasoning, therefore, is partly to equip believers to deal with nonbelievers. […] The common ground between them and believers is not faith but reason.”

32

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

The first chapter, “Apocalypticism, Heresy and Philosophy,” argues that the rise of Buddhist epistemology (or equivalently, the “epistemological turn” of the Buddhist dialectical tradition) was due, not so much to a few individuals’ belief-neutral search for philosophical truth but to a socio-historical matrix of growing Brahmanical hostility toward the non-Vedic, mainly Buddhist, Jaina, Sāṅkhya, Pāśupata and Pāñcarātrika religio-philosophical and/or socio-religious denominations. This can be shown, or at least strongly suggested, by pointing to the discontinuities observable in the evolution of Brahmanical apocalyptic discourse, especially in the changing perception of the evils of the kaliyuga in Epic and Purāṇic literature. These representations at least partly shaped Kumārila’s fierce anti-Buddhist rhetoric, a fact that points to interesting connections between Purāṇic ideas and orthodox Brahmanical philosophy, between “mentalities” (a term that remains to be conceptualized in the indological context) and elite literary productions. As Griffiths rightly points out, “[a]rguments for the superiority of one religious community’s doctrine-expressing sentences over those of another rarely (perhaps never) take place in an institutional vacuum. That is to say, such arguments have, historically, rarely been separated from situations in which there are more than purely intellectual or even purely soteriological issues at stake.”103 This book’s initial chapter is meant to demonstrate just this and hopefully does so by balancing the “continuist,” “incrementalist” and “internal” assumptions underlying the dominant approach known as Problemgeschichte with a “discontinuist” and “external” historiography of religio-philosophical ideas.104 Chapter 2 (“Buddhist Esoterism and Epistemology”) is an attempt to show, on this very basis, that Buddhist “Tantrism” and Buddhist “logic,” two roughly contemporary phenomena regarded here as the main literary outcomes of this difficult period, share interesting features in terms of polemical targets and self-understanding. And indeed, these two traditions’ (rather sparse) self-depictions tend to consider vanquishing the misleading and harmful representatives of the non-Buddhist environment (orthodox Brahmanical 103

Griffiths 1991: 62. Note also Griffiths 1991: 77: “There is always a political context for an apologetical engagement. Just as intellects are never disembodied (at least in this world), so religious communities and their representative intellectuals are never without connections to the systems that structure and apply political, military, and economic power.”

104

On the distinction between “internal” and “external” approaches as applied to philosophical texts, see Taber’s useful overview in Taber 2013: 144–147.

INTRODUCTION

33

schools—Mīmāṃsā, Nyāya, and Vyākaraṇa—, but also Sāṅkhya, on the philosophical level, and Śaivite deities—mainly Śiva/Maheśvara and Umā— on the symbolic, ritual and soteriological level) a necessary (but by no means sufficient) condition of salvation. In my opinion, the Buddhist elites’ sustained polemical and apologetical concern with the non-Buddhists accounts for the rise of Buddhist epistemology and esoterism as well as for the concomitant erosion of Abhidharmic creativity (or equivalently, for the Abhidharma’s rather sudden transformation into AKBh exegesis and commentary). Notwithstanding a few exceptions, Buddhist Abhidharma (but the remark is true of most Buddhist śāstra works during the “Middle period of Indian Buddhism”105) had addressed and targeted fellow Buddhists on a heresiological basis. From the end of the fifth century on, the socio-historical matrix has changed to such an extent that intra-Buddhist polemics becomes less relevant (at least in the form it had had theretofore) and partly recedes into the background in favor of inter- or cross-confessional controversies: The Buddhists now mostly have to meet objections from without, and this becomes the main concern of the best minds of the time. Chapter 3 (“Turning Hermeneutics into Apologetics”) suggests that departing from Abhidharma and addressing new targets resulted in the abandonment of scholastic, confession-specific terminology (the “dharma-logy”) and methods (the yukti-āgama proof strategy) as well as the development of new models of theoretical and apologetical rationality: first, the construction of a clear-cut concept of reason(ing) as opposed to scripture, something made necessary by most protagonists’ new insistence on reason and rejection of their adversaries as following blind faith;106 second, the gradual constitution of a concept of practical rationality that served the apologetic purpose of defending the very possibility, or rationality, of the Buddhist path. Finally, Chapter 4 (“Nescience, Epistemology and Soteriology”) looks into the dogmatic foundations and religious background of early Buddhist epistemology—or, to put it in a more satisfactory way, examines the extent to which Buddhist epistemology can be said to be Buddhist at all as regards its deeper doctrinal structure. This fourth chapter is an attempt to interpret the foundations of Buddhist epistemology—the apoha theory, the pramā105

See below, p. 78, n. 179.

106

One of the cultured Hellenists’ main arguments against early Christianity was “that the Christians, despising reason, rely upon blind faith” (Dulles 2005: 71, dealing with Theodoret of Cyrrhus [c. 393–457]).

34

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

ṇas, the doctrine of inference, etc.—as a rationalization and an apologetically updated version of Buddhist dogmas on the structure of ultimate and conventional realities, on the cognitive bases of error and its elimination, and on the cintāmayī prajñā as a salvific means of a (mainly) inferential order.

Chapter 1 Apocalypticism, Heresy and Philosophy

1.1. INTRODUCTION The greater part of this chapter deals with the evolution of the Brahmanical and Buddhist apocalyptic prophecies found in texts that date from the second to the sixth century CE. As I shall try to demonstrate in the first part of this chapter, the Brahmanical prophecies provide an interesting and somewhat corroborative echo to Harry Falk’s suggestive assumptions regarding the “tidal waves in Indian history”:1 while the Mārkaṇḍeya section of the MBh as well as the YP, both likely to have been composed or 1

Harry Falk (2006) has attracted the attention of scholars to the oscillating, “tidal”-like changes in social order, economy, technology, political self-representation and religiosity that took place during the four phases of Indian political history between 320 BCE and the fifth century CE (Mauryas; Śuṅgas, Kaṇvas, etc.; Śakas, Pahlavas and Kuṣāṇas; Guptas). According to Falk (2006: 146), the “character of society changes twice […] from an extroverted attitude, with thriving foreign contacts on an economic as well as scientific level to an introverted one, where traditional, or ‘brahmanical’ values are preferred to inspiration gained from outside.” Whereas the Mauryas and the foreign dynasties ruling between 50 BCE and 250 CE show a “positive attitude toward innovation,” “those who defeated or succeeded these innovative dynasties reverted to introspective self-containment” (Falk 2006: 146). While he admits that this dichotomy is a “gross simplification” liable to occasional exceptions, Falk concludes as follows (2006: 164): “All cases have yielded similar results: the foreign or foreign-inspired dynasties display an extroverted attitude; they have contacts to foreign rulers, they deal in large scale with foreign economies; they present themselves as individuals; men look at their male forefathers, women enjoy roles in public life. The kings want people to count their reign in dynastic years and they are polytheistic, with remarkable tolerance. The indigenous dynasties, however, are the very opposite. Their main attitude is introverted; they are modest in their titles and stress their descent from pure brahmin mothers, when at the same time women withdraw from public activities. The economy is rather local with no or little need for high-value currencies; their religion is basically henotheistic; their feeling for time is Vedic with no need for a starting point.” For similar observations, see Verardi 2011: 69–107.

36

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

at least updated during the first two and a half centuries CE, consider foreign, mleccha rule2 as the hallmark of the kaliyuga and/or of its final period (yugānta), Gupta Purāṇas such as the VāP, the BḍP and the ViP hold the increase in “heresy” and “heretics” (pāṣaṇḍa, pāṣaṇḍin)3 to be the most

2

On mlecchas, see Thapar 1971 and Parasher 1991 (especially pp. 222–261 on mlecchas as foreigners and foreign rulers whose behaviour opposes the varṇāśramadharma, as well as on the great flexibility apparent in this designation).

3

For a conceptual discussion on the applicability of “heresy” to things Indian, especially to the pāṣaṇḍas/pāṣaṇḍins, nagnas and nāstikas, see Doniger O’Flaherty 1971: 272–287 and 1983: 109–117; see also Verardi 2011: 107–109. As is well known, the word pāṣaṇḍa seems to occur first in the inscriptions of Aśoka: in Rock Edict 12 (see Bloch 1950: 121–124), where sarvapāṣaṇḍa covers both pravrajitas (wandering ascetics) and gṛhasthas (householders), in Rock Edict 7 (see Bloch 1950: 110–111), where sarvapāṣaṇḍas are said to wish “la maîtrise des sens et la pureté de l’âme” (Bloch 1950: 110, sayamaṃ ca bhāvasuddhiṃ ca [Girnar]), and in Pillar Edict 6 (see Bloch 1950: 166– 168), where it enters an interesting but problematic relationship with sarvanikāya (“tous les groupes,” Bloch 1950: 168). Contrary to later usage, pāṣaṇḍa seems to refer here to “sectarians” (Norman 1990: 147) or “members of a religious sect” (Bailey 1952: 427) in a neutral way—hence perhaps its Greek translation (in the rendering of the twelfth Rock Edict) as διατριβή, “école philosophique,” “école(s) (de pensée)” (Benveniste 1964: 139 and passim; Schlumberger/Benveniste 1967–1968: 194 and 199: “In the Asoka inscriptions it designates any kind of faith, including the king’s faith”). The etymology is controverted: according to Bailey (1952: 427–428), pāṣaṇḍa comes from old Iranian *frašanta (< fraš), “asker,” “questioner” (see also Norman 1990: 147 and Mayrhofer 1956–1972: 266); according to Mayrhofer (1956–1972: 265–266), the expression is to be connected to parṣad (pariṣád) and its derivatives pāriṣada and pārṣada (“Teilnehmer an einer Versammlung,” Mayrhofer ibid.). In the following, “heretic”/“heresy” is used, by convention more than by conviction, to translate pāṣaṇḍin/pāṣaṇḍa, the terms used most often by Brahmanical sources to designate non- or anti-Vedic (ascetic) groups. I render the functionally equivalent Buddhist term tīrthika as “outsider” or sometimes simply as “non-Buddhist,” not as “heretic,” since “heresy” usually designates, within one and the same denomination, an ethical and/or doctrinal and/or practical option (hence often a “sub-denomination”) considered deviant by self-appointed representatives of orthodoxy/orthopraxy. The only term I might otherwise conceive of as an equivalent of tīrthika is “gentile,” which came to designate non-Jews for Jews and pagans for Christians. However, at least originally, this word contains a strong ethnic or national component that is entirely missing in the Buddhist notion of tīrthika. I also occasionally resort to the notions of “heterodox(y)” and “allodox(y)” (as a possible translation of Skt. paravāda; see Eltschinger forthcoming b, n. 14) in order to distinguish, from a strictly Buddhist point of view, dissenting coreligionists (e.g., Vātsīputrīyas for Sarvāstivādins) and non-Buddhists respectively. On this distinction, see Scherrer-Schaub 1991: xli, n. 63.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

37

distinctive feature of the kaliyuga. My aim here is to show that this new apocalyptic concern with Buddhism and Jainism can be seen as one part of the ideological and rhetorical background against which the Brahmanical philosopher Kumārila Bhaṭṭa (sixth century4) turned the ritualistic Mīmāṃsā into the most uncompromising anti-Buddhist philosophical system ever created in ancient India. The second part of this chapter deals with the Indian Buddhist counterparts to the Brahmanical views on the degeneration of dharma. At the present stage of my research and reflections, the Indian Buddhist prophecies regarding the decline and final demise of the saddharma do not seem to have had any discernable impact on the development of Buddhist philosophy during the sixth century. Nevertheless, they provide indicative evidence to the effect that by the turn of the sixth century, certain Indian Buddhists started paying attention to new threats and, in reaction, modified their apocalyptic schemes by taking up the hitherto purely Brahmanical5 pattern of four ages (yuga). On the one hand, Brahmanical apocalyptic rhetoric testifies to an increasing concern with Buddhist and other anti-Vedic denominations as well as hostility towards the same; on the other hand, the Buddhist prophecies simultaneously bear witness to a growing awareness of external causes of decline, such as political (Mihirakula) and socioreligious (Śaiva threats; Pūraṇa Kāśyapa paradigm) hostility. Considering that there is surprisingly little evidence for philosophical polemics between Buddhists and non-Buddhists before the (end of the) fifth century,6 one might see these roughly contemporaneous apocalyptic prophecies as providing interesting materials with which to formulate the social and ideological background of this sudden transposition of mutual rivalry to the philosophical level. But before I turn to these materials, let me add a few remarks on my use of the terms “apocalypticism” and “apocalyptic.” Since Norman Cohn’s influential book The Pursuit of the Millennium,7 social scientists and theologians have paid increasing attention to messianism, millenarianism and apocalypticism. According to the American historian Bernard McGinn, apocalypticism does not consist primarily in an

4

On the chronology of Kumārila and Dharmakīrti, see below, p. 116, n. 80.

5

But see below, p. 82, n. 196, Chapter 2, §2.3.1.3, and Eltschinger 2012: 439–452.

6

See below, p. 71, n. 152.

7

Cohn 1970.

38

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

interest in the “End” as such, but in a concern with the structure and meaning of history, in the “ability to locate current events within a scheme of universal meaning, to transcendentalize the present by viewing it from the aspect of the End.”8 In other words, apocalypticism is less concerned with the eschaton (already described in generally available prophecies) than with the social and political present,9 which is seen as in a crisis and needing a meaning that is related to the End. As such, apocalypticism is a “way in which contemporary political and social events were given religious validation by incorporation into a transcendent scheme of meaning.”10 Of particular relevance is the apocalyptic rhetoric’s regular tendency to proceed ex post facto in a vaticinium ex eventu, i.e., as “history disguised as prophecy.”11 As McGinn has it, [o]ne of the characteristics of apocalyptic eschatology is its drive to find meaning in current events by seeing them in light of the scenario of the End. Such a posteriori, or after-the-fact, uses of apocalypticism are often reactions to major historical changes […] that do not fit into the received view of providential history. By making a place for such events in the story of the end, the final point that gives all history meaning, apocalyptic eschatology incorporates the unex12 pected into the divinely foreordained and gives it permanent significance.

Developing a “typology of the negative and positive functions of apocalyptic beliefs,”13 McGinn distinguishes between a positive and a negative political use of apocalypticism, these in turn being divided into a priori and a

8

McGinn 1998: 32.

9

Note also McGinn 1998: 31: “[T]heologians and students of the history of religions should […] be prepared to admit how inextricably apocalypticism is intermingled with the ‘political’ sphere at all times.”

10

McGinn 1998: 31. Note that contrary to Cohn, McGinn does not hold apocalypticism to originate in oppressed, poor and uneducated social milieux. Rather, “[t]he social type to which most medieval apocalyptic propagandists conform was […] the well-educated and well-situated clerical intelligentsia” (McGinn 1998: 31–32).

11

McGinn 1998: 7.

12

McGinn 2000: 88. See also McGinn 2000: 120–121.

13

McGinn 1998: 30.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

39

posteriori positive uses on the one hand, and active and passive negative uses on the other hand.14 The present state of our knowledge forbids any clear-cut statement concerning the political purpose(s) and/or use(s) of ancient Indian apocalyptic prophecies. While it seems reasonable to view persons with political power as one intended (but not necessarily reached) audience of at least some Brahmanical texts, nothing similar can be said concerning Buddhist predictions, which are likely to have served the purely intramonastic purpose of clerical repristination.15 Similarly, our sources provide no hint of active negative apocalypticism. True, most documents point to a negative form of apocalypticism in that they unambiguously condemn what they regard as the dramatic signs of the End (e.g., mleccha rule, adharma and heresy in Brahmanical sources; moral and disciplinary laxity, lack of knowledge and spiritual attainments in the Buddhist sources) and lay at least reformist demands for their eradication. What can be said, however, is that most of these documents testify to an a posteriori apocalyptic reading of current states of affairs and often reflect conscious attempts at updating preexisting end-of-time scenarios and symbols.

14

In positive political uses, the “rhetoric of apocalypticism [is] used in support of the political and social order” (McGinn 1998: 33); in its a priori form, it makes use of a preexisting apocalyptic scenario in order to provide present events with a meaning and to prompt people to make decisions and to act; in its a posteriori form, it extends the scenario “to include transcendentalized versions of recent events, thus giving final validation to the present by making a place for it at the End” (McGinn 1998: 33; a posteriori positive apocalypticism proved especially productive in connection with three decisive historical events: the conversion of the Roman Empire, the encounter with Islam, and the “Great Reform” of the eleventh and twelfth centuries). In negative political uses, the “rhetoric [is] directed against the present evil order” (McGinn 1998: 32); whereas its active form comes “very close to revolution” (McGinn 1988: 133; as in the case of the Hussite Taborites in fourteenth century Bohemia), its passive form is by far the most frequent.

15

As Alexis Sanderson has kindly informed me (personal communication, May 2010), the MMK might provide an exception to this. In verses 53.816–818 and in 427,6ff., “the end-time is used to explain […] why the state must now rely on the services of the Mantranaya for the maintenance of sovereignty and to defend itself against the enemies of the faith.” The expression used in this connection is paścime kāle paścime samaye śāsanavipralope vartamāne.

40

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

1.2. MLECCHA AND PĀṢAṆḌA/PĀṢAṆḌIN IN BRAHMANICAL APOCALYPTICISM 1.2.1. Introduction Apart from the respective names of the four ages, which are borrowed from the Vedic game of dice and its four throws (from the winning throw, kṛta, to the losing throw, kali), nothing is known about the prehistory of the doctrine of the four yugas.16 A fairly developed form of the theory of the four ages is found in its two earliest extant sources, the MDhŚ and the MBh.17 Unfortunately, both the dates of these two texts and their mutual relations belong to one of the most intricate problems to be encountered in Indology.18

16

Secondary literature about the four yugas includes Huntington 1964, Church 1971, Dwivedi 1977, Yadava 1978–1979, Sharma 1982, von Stietencron 1986, Kane 1993: 885–968, Biardeau 1994, Koskikallio 1994, Michaels 1998: 330–335, Sharma 2001: 45– 76 (≈ Sharma 1982), González-Reimann 2002, Verardi 2011: 139–147 (and passim).

17

On the Babylonian origin of the kalpa theory and of the duration assigned to the kaliyuga (432,000 years), see Pingree 1963: 238–239. Here the author simply asserts that “the first text to describe the kalpa precisely […] was a pre-second-century work which was the common source of a passage occurring in the twelfth book of the MBh and in the first book of the MDhŚ.”

18

On this problem, see Bronkhorst 2012. According to Olivelle, the MDhŚ is to be dated to the second-third centuries CE on account of its mentioning the minting of gold currency, and “the author(s) of the epic knew of and drew upon material from the MDhŚ” (Olivelle 2005: 23). But as pointed out by Bronkhorst, this statement is problematic since leading scholars in the field of Epics (Biardeau, Hiltebeitel, Sutton, and Fitzgerald as far as the “main MBh” is concerned) are inclined to date the MBh to the period between the third and first century BCE. A solution to this seems to be to consider the problem, as Olivelle himself seems to do, in the light of Fitzgerald’s opinion that “the main MBh […] was fashioned and promulgated sometime in the second or first centuries BCE” (Fitzgerald 2006: 269), but that it “grew in the four or five hundred years between the fixing of the main MBh and its recognition as a śatasāhasrikī saṃhitā during the Gupta era” (Fitzgerald 2006: 270). In other words, the MDhŚ might be “a text which was composed, in its earlier form, at a time when the ‘main MBh’ existed already, but the archetype underlying its Critical Edition not yet. In a certain way the MDhŚ is therefore contemporaneous with the MBh, in the sense specified” (Bronkhorst 2012: 146).

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

41

1.2.2. The Doctrine of the yugas in the MDhŚ The MDhŚ contains only a very short account of the four yugas. It is to be emphasized that MDhŚ 1.81–86 is of no apocalyptic purport, but rather consists in a cosmological description of the gradual degeneration of dharma and humankind through the ages. According to the MDhŚ, dharma is complete, “four-footed” (catuṣpāt), in the kṛtayuga, but has one less “foot,” or portion (pāda), in every subsequent yuga due to the gradual appearance of thievery (caurikā), lying (anṛta) and deceit (māyā). As for the religious duties (dharmāḥ) of human beings, they are different in each age, “in proportion to the decrease of each age” (yugahrāsānurūpatas).19 During the kṛtayuga, humans are free of sickness (aroga), achieve all their goals (sarvasiddhārtha) and have a lifespan of four hundred years (caturvarṣaśatāyus), but their life-spans (āyus), expectations (āśis) and powers (prabhāva) decrease from age to age (anuyuga). And whereas austerities (tapas) are paramount (para) in the kṛtayuga, knowledge (jñāna) predominates in the tretāyuga, sacrifice (yajña) in the dvāparayuga, and giving (dāna) is the only thing left in the kaliyuga. As one might expect given the non-apocalyptic nature of the text, MDhŚ 1.81–86 does not allude to any historical event or figure as characteristic of any of the four ages. To be more precise, the MDhŚ, though it is aware of mlecchas (without any clear hint at mleccha rulers on āryāvarta, however),20 heresies (pāṣaṇḍa) and heretics (pāṣaṇḍin), does not associate any of them with the kaliyuga, as the texts to be considered below variously do. 1.2.3. The kaliyuga and its End in the MBh 1.2.3.1. A few preliminary remarks are necessary before turning to the MBh’s depictions of the kaliyuga. Recent studies by Sutton, Biardeau, Fitzgerald and Hiltebeitel have shed new light on the old issue of Buddhism in/and the Sanskrit Epics. All four agree that “Aśoka is the most highimpact historical figure on the MBh,” although they “differ over how and where to trace that impact.”21 For instance, Biardeau poses, “as [her] 19

Translation Doniger/Smith 1991: [12/]79. On MDhŚ 1.85 and dharmāḥ, see Lingat 1961, who discusses the problem (raised already in Medhātithi’s Manubhāṣya [69,3–6]) of an eternal dharma being subject to change according to time/yugas.

20

See Parasher 1991: 246–247.

21

Hiltebeitel 2005: 114.

42

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

fundamental hypothesis, a causal rapport between the conversion of Aśoka and the composition of the MBh,” the latter making up a “brahmanical manifesto.”22 And indeed, whatever their views on the way to trace the figure of Aśoka (and/or other kings) in the MBh and the chronology of the Epic, these scholars seem to agree that the MBh is part of a new Brahmanical self-assertive strategy made necessary, inter alia, by the growing impact of Buddhism and its religio-political power.23 And in the MBh, this new self-assertion and its demands with regard to the “dharma king” are cast into an “apocalyptic […] narrative,” i.e., into a “narrative of apocalyptic purge and its demand for kṣatriya kingship.”24 My point in sketching current research on the MBh is the following: One can view the 22

Biardeau 2002: II.749 as translated in Hiltebeitel 2005: 109.

23

Note, e.g., Hiltebeitel 2005: 116: “The MBh [is] designed to sustain a subtle, patient, and emerging political theology that propounds a new bhakti framework in which royal patronage and Brahman prestige are provided with new places and meanings […] The MBh […] often reflects greater antipathy toward Buddhism than Biardeau indicates.” Or, in Fitzgerald’s (2006: 276) words: “If we step back and look at the whole of the MBh in the historical context which it most likely deliberately addressed—namely, postNandan, post-Mauryan, particularly post-Aśokan, north India sometime after 250 BCE, we can gain, I believe, some general insight into the development of the MBh’s arguments on varṇadharma and brahmins. Aśoka had preempted the brahmin monopoly on the teaching of dharma (he launched a ‘Dharma-campaign’ in his various edicts, presuming to teach ‘Dharma’ on his own authority); and not only did he do an endrun around the brahmin magisterium, he, like his Mauryan predecessors, patronized the nāstika (‘heathen’) elites of Jains, Buddhists, and others and became a lay Buddhist himself. And within this relatively cosmopolitan empire Aśoka forbade brahmins to offer animal sacrifices and criticized various of their festivals. The Nandas, the Mauryas—Aśoka in particular—and their client kings were certainly kings fostering the kind of saṅkara the MBh finds abhorrent, the kind of saṅkara the MBh wishes to guarantee can never arise. The newly chartered kingship of the MBh is designed expressly to apply violence (the king’s daṇḍa) within society to prevent saṅkara from arising, to maintain varṇadharma. According to the MBh this is the main purpose of the king and his instruments of force […] These tales make unambiguous the epic’s understanding of the basic political reality that there can be no varṇadharma, no hierarchy, without the king’s violence.” See also Fitzgerald 2004: 120.

24

Fitzgerald 2004: 122. As Biardeau (2002: II.747 as translated in Hiltebeitel 2005: 109) points out, the MBh “is a matter of an apocalyptic crisis of the world that leads a temporal cycle […] from its genesis to its end, while passing through a long war-sacrifice, or, better still, a long sacrificial session of apocalyptic magnitude—since it brings about the birth and disappearance of all the sovereigns of the Earth, all the while assuring the continuity of the future existence of the Earth purged for a while of the wicked.”

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

43

MBh per se as an apocalyptic narrative developed against the background of a “long history of Brahman dissatisfactions […] with heterodoxies and heterodox rule”25 and see its scattered explicit depictions of the kaliyuga, mainly in the form of apocalyptic prophecies, as not necessarily mirroring this original concern. In other words, the MBh, as a compilation, may well be cryptically concerned with Buddhism (inter alia), and yet the passages of relevance to us need not, as originally independent materials, reflect the Epic’s overall purpose. The MBh devotes several passages of noticeable length to the four yugas. In the Epic, two kinds of texts should be distinguished: On the one hand, passages (such as 3.148) of a purely cosmological purport and close in intent to MDhŚ 1.81–86; on the other hand, passages (such as 3.186.24– 55 and 3.188, in the so-called Mārkaṇḍeya section) which present themselves as apocalyptic prophecies (generally in the future tense) predicting the pitiable condition of society, kingship, morality and behaviour at the collapse of the kaliyuga. Whereas the first kind of sources remain entirely silent about heresy and the like, the second type provide, though surprisingly rarely, a few hints that heretics are heralding the last times. 1.2.3.2. As a purely cosmological account of the yugas, MBh 3.148.5–3726 is well in tune with the spirit (but certainly not with the doctrine) of the MDhŚ. It provides Hanumat’s answer to the following request by Bhīma: Enumerate the yugas and [describe] the proper conduct in each yuga. Tell me about the modes of Dharma, desire (kāma), and profit (artha); about size, courage, 27 and life and death.

According to Hanumat, there was only one Veda (with one set of mantras and injunctions, but without sāman, yajus and ṛc sounds!) and accordingly one dharma (consisting in renunciation—sannyāsa —, i.e., the practice of ātmayoga, one’s “union with the [supreme] self”) in the kṛtayuga.28 The dharma of the stages of life (āśrama, leading to the supreme goal) and 25

Hiltebeitel 2005: 122.

26

See Biardeau’s short summary in Biardeau 2002: I.538–539 and González-Reimann’s translation in González-Reimann 2002: 215–221. My account of MBh 3.148.5–37 is largely indebted to González-Reimann’s translation.

27

MBh 3.148.9: yugasaṅkhyāṃ samācakṣva ācāraṃ ca yuge yuge / dharmakāmārthabhāvāṃś ca varṣma vīryaṃ bhavābhavau //. Translation González-Reimann 2002: 215.

28

MBh 3.148.13, 19, 21.

44

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

caste-classes (varṇa, including svadharma, “own [socioreligious] duty”) was respected.29 There were neither supernatural beings (gods, etc.) nor traiguṇya (“three qualities,” i.e., goodness, passion, and darkness), trade, human labour, tasks, diseases, suffering or bad dispositions (calumny, arrogance, slander, laziness, hatred, fear, envy, quarrelling, or selfishness).30 In the tretāyuga, dharma is diminished by one quarter.31 This age is characterized by the appearance of ritual (sattra, kriyā, yajña) and by people being truthful and devoted to charity.32 Though this is not stated explicitly, we ought to understand that this second age is dominated by the sattva(guṇa). In the dvāparayuga, dharma is reduced by a half, while the hitherto unique Veda is divided into four parts (ritual, too, becomes differentiated).33 In many respects, the dvāparayuga heralds the kaliyuga: people become passionate (rājasa) and decline in their regard for truth despite their still being devoted to austerities and giving;34 diseases and calamities gradually appear35 so that, “greatly afflicted by them, some people practice austerities [while o]thers perform sacrifices, wishing to obtain earthly goods or heaven.”36 In the kaliyuga, the dark (tāmasa) age, only a fourth part of dharma remains.37 Vedic practices decline along with the performance of sacrifices.38 Lazy and impassioned, people are beset with natural disasters, diseases,39 calamities, and mental as well as physical suffering.40 And 29

MBh 3.148.17, 20.

30

MBh 3.148.10, 12–15, 22.

31

MBh 3.148.23.

32

MBh 3.148.22–25.

33

MBh 3.148.26, 28.

34

MBh 3.148.28–29.

35

MBh 3.148.30.

36

According to MBh 3.148.31 (yair ardyamānāḥ subhṛśaṃ tapas tapyanti mānavāḥ / kāmakāmāḥ svargakāmā yajñāṃs tanvanti cāpare //) and parts of González-Reimann’s translation (2002: 217).

37

MBh 3.148.32–33.

38

MBh 3.148.33.

39

Note González-Reimann 2002: 168–169: “Even medical treatises blamed the existence of sickness on the kaliyuga. In a chapter devoted to the spread of epidemics, the CS (3d– 5th centuries) […] explains the origins of disease as the consequence of the gradually declining energy and well-being of people brought about by the passage of the yugas. As rains, crops, and the quality of diet and life-style diminish, so does the body’s resistance

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

45

[w]hen the world is in a state of decay, the conditions that allow for its advancement are destroyed. The religious practices (dharmas) performed at the end of the yuga produce opposite results.41

1.2.3.3. As apocalyptic prophecies focusing on the kaliyuga and especially its end (yugānta), our second and third MBh passages yield a decidedly different testimony. Both have been invoked as bearing explicit witness to Buddhism and heresy as a prominent feature of the Epic’s description of the last times.42 But are heresy and/or Buddhism so prominent in these prophecies? The relevant passage of MBh 3.186 covers about 30 verses, with only a single short hint of pāṣaṇḍas. As for MBh 3.188, it comprises 93 verses, of which vv. 14–93 form the prophecy proper. The verses supposedly concerned with Buddhism are only two, representing approximately 䱵40 of the relevant text, and it is possible that they are a later addition.43 Much more striking is this second passage’s (MBh 3.188) strong, almost refrain-like emphasis on mlecchas or, better, the barbarization of the world. Indeed, vv. 29, 37, and 45 begin with the statement: mlecchabhūtaṃ jagat sarvaṃ bhaviṣyati (“the entire world will become barbarized”); as for verse 70, it declares that the earth will be overthrown with mlecchas (mahī mlecchasamākīrṇā bhaviṣyati).44 And though MBh 3.186.24–55 lacks such a refrain-like insistence, it is even more precise regarding the mlecchas:

to disease, and this coincides with the shorter life-spans caused by the advance of the yugas [CS 3.3.24 (= González-Reimann 2002: 189, n. 22)]. At the conclusion of this cosmic etiology of disease the treatise quotes the following: In each yuga a fourth of dharma is sequentially removed, as are the qualities of living beings. In this way, the world is [gradually] destroyed [CS 3.3.25: yuge yuge dharmapādaḥ krameṇānena hīyate / guṇapādaś ca bhūtānām evaṃ lokaḥ pralīyate // (= González-Reimann 2002: 189, n. 24)].” 40

MBh 3.148.34.

41

MBh 3.148.36: loke kṣīṇe kṣayaṃ yānti bhāvā lokapravartakāḥ / yugakṣayakṛtā dharmāḥ prārthanāni vikurvate //. Translation González-Reimann 2002: 218.

42

See Biardeau’s summary of and commentary on MBh 3.188 in Biardeau 2002: I.588– 612; see also Biardeau 2002: II.759–762 and Hiltebeitel 2005: 124–129.

43

See below, §1.2.3.4.

44

I can hardly resist the impression that the kernel of this text originally consisted in a depiction of the dark ages centred on the omnipresence of mleccha(-rule) that was built upon this refrain (occurring every eight verses?).

46

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

[At the collapse of the yuga,] O prince among humans, [there will be] on earth numerous mleccha kings, ill-governing, evil and entirely dedicated to lying. [These] sovereigns, O [you] best among humans, [will be] Āndhras, Śakas, Pulindas, Ya45 vanas, Kāmbojas, Aurṇikas, Śūdras and Ābhīras.

I shall turn back to this strong concern with mlecchas later on. 1.2.3.4. In addition to statements concerning the general decline of dharma and the paucity of righteous people (dharmiṣṭha),46 MBh 3.186.24–55 gives considerable emphasis to lying and falsity, hypocrisy (especially religious hypocrisy and imposture), trickery (especially in business and trade matters, which it does not see positively), false accusation and adultery: these are unmistakable signs of the yugānta.47 Religious duties (svādhyāya, yajña, japa, vrata, piṇḍa) sink into oblivion;48 capital sins (pānapa, gurutalpaga, brahmavadhyā, various types of thievery), immorality, ill-behaviour, greed, and violations of spoken etiquette corrupt social order;49 transgressions regarding licit food, intercaste relationships and marriage bonds tend to increase dramatically.50 Last but not least, the burden of taxes (karabhāra), the decline of human faculties (āyus, bala, tejas, parākrama, deha, sāra, satya) and various kinds of natural and physiological disturbances are considered as making life extremely precarious and dangerous.51 pāṣaṇḍas are but one element of this eloquent description: “[These formerly righteous persons are] confused by numerous pāṣaṇḍas and praise the food [begged] from others?.”52 If generous, one might also interpret our text’s allusion to the regions (diś) that are covered with the signs (dhvaja)

45

MBh 3.186.29–30: bahavo mleccharājānaḥ pṛthivyāṃ manujādhipa / mithyānuśāsinaḥ pāpā mṛṣāvādaparāyaṇāḥ // āndhrāḥ śakāḥ pulindāś ca yavanāś ca narādhipāḥ / kāmbojā aurṇikāḥ śūdrās tathābhīrā narottama //.

46

MBh 3.186.45, 47–48.

47

MBh 3.186.25, 33, 38, 40–42, 46, 49, 55.

48

MBh 3.186.25, 27–28.

49

MBh 3.186.33, 35, 38, 40, 42, 50.

50

MBh 3.186.26–28, 31, 42, 55.

51

MBh 3.186.32, 35, 37, 40, 44, 48, 51–54.

52

MBh 3.186.43ab: bahupāṣaṇḍasaṅkīrṇāḥ parānnaguṇavādinaḥ /.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

47

of erroneous dharma(s) as a hint of unspecified heterodoxies.53 That’s all, and it is not much. 1.2.3.5. In MBh 3.188, the longer of the two passages, the ṛṣi Mārkaṇḍeya’s teaching on the evil times (kaluṣaṃ kālam) to come starts with a fourverse statement about the now familiar reduction of dharma by one quarter in every yuga, and the parallel decline in human beings’ lifespan (āyus), vigour (vīrya), intelligence (buddhi), force (bala) and glory (tejas).54 Among the characteristic features of the kaliyuga that recur in this prophecy, one might mention the following: macro- and microcosmic disturbances growing into cosmic disasters,55 shortening of life-expectancy, lessening of strength and firmness,56 restless wandering and living in towns and cities,57 lack of truthfulness,58 and the abandoning of behavioural observances, sacrifices, religious festivals and funerary rituals.59 Also repeated in this passage is the increase of ignorance as well as its concomitant pseudo-erudition.60 Worth mentioning in this connection is the extremely peripheral place allotted to the features traditionally ascribed to the groups challenging Vedic authority and Veda-based dharma, such as being a denier (nāstika), being contemptuous (nindaka) of the Veda, or being led astray by ratiocination (hetuvāda).61 These traits, however, do not describe pāṣaṇḍas (a term which only occurs at the beginning of MBh 3.189), but human beings, especially Brahmins. Similarly, religious hypocrisy is ascribed to all four caste-classes.62 As for the topos of śūdras

53

MBh 3.186.39: lobhamohaparītāś ca mithyādharmadhvajāvṛtāḥ / bhikṣārthaṃ pṛthivīpāla cañcūryante dvijair diśaḥ //. “Twiceborn, O protector of the earth, are [then] roaming about in order to beg [alms] in the regions seized by greed and error and covered with the signs of erroneous dharma(s).” Begging for alms does not necessarily refer to extra-Vedic denominations or individuals.

54

See MBh 3.188.10–13.

55

MBh 3.188.68–69, 74–76, 79, 81.

56

MBh 3.188.15–16, 31, 47–48, 62.

57

MBh 3.188.58, 60, 83–84.

58

MBh 3.188.15, 18, 38, 76.

59

MBh 3.188.22, 26, 29, 45, 69.

60

MBh 3.188.16, 32, 36, 38, 41, 46, 54, 72–73.

61

MBh 3.188.22, 26.

62

MBh 3.188.14.

48

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

teaching Brahmins, in spite of its later occasional connection with Buddhism, it also has nothing to do with heretics.63 The bulk of this passage is concerned with immorality, whose most recurrent aspects are violence, harshness and cruelty.64 Of great frequency are also theft,65 human passions (especially greed),66 offences against marriage bonds and family relationships,67 mixing of castes and transgressions of svadharma,68 the breaking of dietary rules and licit means of livelihood,69 and, last but not least, violence to and taxes imposed on Brahmins.70 The only passage likely alluding to heresy can therefore only be interpreted as of comparatively marginal importance: The world will be totally upside down: people will abandon the gods and will offer pūjā to eḍūkas; Śūdras will refuse to serve the twiceborn at the collapse of the yuga. In the hermitages of the great Ṛṣis, in the settlements of Brahmans, at the gods’ temples (devasthāneṣu), in the Caitya sanctuaries, and in the abodes of the Nāgas, the Earth will be marked by eḍūkas and not adorned by houses of the 71 gods. At the expiration of the yuga, that will be the mark of the yuga’s end.

These verses have been interpreted (Biardeau, Hiltebeitel, etc.) as referring to (proto)stūpas (the eḍūkas) and, as such, as presenting Buddhism as a sign of the last times. But as recent studies by Flügel and Bronkhorst clearly show, stūpas were also erected by the Jains (and maybe the old Magadhan culture as a whole).72 Moreover, the meaning of eḍūka is 63

MBh 3.188.54.

64

MBh 3.188.17, 22, 28, 32, 50, 52, 55–56, 59, 71, 77–78, 84.

65

MBh 3.188.22, 30, 34, 36–37.

66

MBh 3.188.16–17, 24, 31–33, 39, 50, 53, 55, 57, 82.

67

MBh 3.188.35, 42, 44, 49, 77.

68

MBh 3.188.18–19, 41, 57, 62–63, 69.

69

MBh 3.188.21, 24–25, 27, 43, 52, 67, 71.

70

MBh 3.188.57–58, 61, 70.

71

MBh 3.188.64–66: viparītaś ca loko 'yaṃ bhaviṣyaty adharottaraḥ / eḍūkān pūjayiṣyanti varjayiṣyanti devatāḥ / śūdrāḥ paricariṣyanti na dvijān yugasaṅkṣaye // āśrameṣu maharṣīṇāṃ brāhmaṇāvasatheṣu ca / devasthāneṣu caityeṣu nāgānām ālayeṣu ca // eḍūkacihnā pṛthivī na devagṛhabhūṣitā / bhaviṣyati yuge kṣīṇe tad yugāntasya lakṣaṇam //. Translation (slightly modified) Hiltebeitel 2005: 126; see also Biardeau 2002: II.759 and Bakker 2007: 15.

72

See Flügel 2008 and Bronkhorst 2009.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

49

not at all unambiguous, and thus this term does not necessarily refer to stūpas.73 One can also note, finally, that these verses contain the only occurrence of (yugāntasya) lakṣaṇam, a wording reminiscent of later, Purāṇic formulations. Here it might be suspected of placing emphasis on an element that was lacking in a more original version of the text. 1.2.3.6. The so-called Mārkaṇḍeya section provides what might be the first (and rather tenuous) occurrence of heresy in the apocalyptic scenario. Not only this, it also ends on the famous, properly messianic prophecy of the advent of Kalki(n)74 that would eventually crystallize into Viṣṇu’s tenth and final avatāra.75 In other words, MBh 3.186 and 188 present a vivid socioreligious depiction of the kaliyuga and prophesize the imminent rise of a new kṛtayuga: As the kṛta dawns, things return to their proper course. A brāhmaṇa called Kalkin is born, and he becomes a universal ruler, a cakravartin. He destroys the foreigners, and re-establishes the rule of the twice-born by means of a Vedic horse sacrifice. Order is then restored and Dharma adhered to, the brahmanical social order

73

According to Amarasiṃha (AmK 2.2.4a2b), an eḍūka is a walled structure in which bones are deposited (eḍūkaṃ yadantarnyastakīkasam /). The best discussion of the terms eḍūka/aiḍūka to date is Bakker 2007; see also Allchin 1957. Contrary to the aiḍūka (for which the locus classicus is VDhP 3.84.1–15, quoted and discussed in Bakker 2007: 12, n. 4, and 12–13), which did not contain any physical remnants of the dead, eḍūkas can be defined as “[b]urial mounds or sepulchral monuments that actually contain the ashes and/or bones of the deceased” (Bakker 2007: 19; for a physical description, see Bakker 2007: 43). According to Bakker (2007: 40), “we may take it for certain that the author of the VDhP calqued his Hindu aiḍūka on a Buddhist example [see already Bakker 2007: 13–14, VE], but by doing this he elaborated on what must have been a monumental tradition that was common to all Indian religions, most pronounced within Buddhism, less in Jainism and inconspicuous in the Hindu mainstream.” Referring to the MBh passage under consideration, Bakker (2007: 33, n. 76) quotes Allchin 1957: 1 (“Since Lassen it has been generally accepted that this description of the Kali-yuga refers to the spread of Buddhist practices and the popular desertion of Brahmanical temples.”), and adds: “This is not to say, however, that the composer of this passage was exclusively thinking of Buddhism. He might have lashed out at all pan-Indian practices frowned upon by the orthodox that involved the erection of monuments over mortuary remains. Eḍūka thus seems to be a wider term than, for instance, stūpa, and it clearly has here, if not in all cases where it occurs, a pejorative connotation.”

74

See MBh 3.188.89–3.189.9 and Hiltebeitel’s (2005: 127) translation.

75

See below, §1.2.7.2, and p. 59, n. 116.

50

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

prevails, and rituals are again practiced. Each social class performs its appointed 76 duty and the people prosper.

Besides multiple allusions to the final destruction of all mlecchas and dasyus, this prediction contains a brief mention of pāṣaṇḍa(s): At the time when the new kṛtayuga dawns in the areas conquered by Kalki(n) and his fellow Brahmins, “hermitages [now] bereft of heretics will be firm in truth.”77 1.2.4. The kaliyuga in the YP What about the dates and provenance of these apocalyptic and messianic prophecies? Many scholars are ready to accept that major parts of the MBh’s books 3 and 12 represent later additions. As González-Reimann points out, “Mārkaṇḍeya’s entire teaching on the yugas has no connection to the Epic story.”78 Moreover, Mārkaṇḍeya himself acknowledges external sources at the time of concluding his prophecy: “I have told you everything about the past and the future as I remember it from the purāṇa recited by [the god] Vāyu, which is praised by the ṛṣis.”79 Basing himself on close similarities between Mārkaṇḍeya’s teachings and the extant VāP, González-Reimann is inclined to postulate (in addition to an older version of the VāP) (a) common source(s) for both texts. According to Mitchiner and González-Reimann, the so-called YP is a very likely candidate for being a source of both the Mārkaṇḍeya section and (an older version of) the VāP and BḍP.80 This text, strictly speaking not a Purāṇa, is nothing other than a 115-verse chapter of the Gārgīyajyotiṣa, a work that is referred to in the MBh at least twice.81 As González-Reimann observes, the YP presents itself as Śaṅkara’s (= Śiva’s) answer to Skanda’s question about the events supposed to take place at the yugānta, a question very similar to the one 76

González-Reimann 2002: 95.

77

MBh 3.189.9cd1: āśramāḥ hatapāṣaṇḍāḥ& sthitāḥ satye […] //. &hatapāṣaṇḍāḥ (S MSS; hatapākhaṇḍāḥ K4, D1 MSS): sahapāṣaṇḍāḥ Ed. Translation (modified) Hiltebeitel 2005: 127.

78

González-Reimann 2002: 95.

79

MBh 3.189.14: etat te sarvam ākhyātam atītānāgataṃ mayā / vāyuproktam anusmṛtya purāṇam ṛṣisaṃstutam //. Translation González-Reimann 2002: 96.

80

On the VāP and BḍP, see below, §1.2.6.

81

MBh 9.36.14–17 and 13.18.25–26.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

51

posed by Yudhiṣṭhira to Mārkaṇḍeya. And even more than Mārkaṇḍeya’s teachings, the YP insists on foreign, mleccha rule as the hallmark of the yugānta. After a short liminary dialogue between Skanda and Śaṅkara (vv. 1–4), the YP goes into an original depiction of the kṛtayuga (vv. 6–13),82 the tretāyuga (vv. 14–22)83 and the dvāparayuga (vv. 23–36, with an interesting allusion to the plot of the MBh). The apocalyptic prophecy regarding the kaliyuga is built on a lengthy succession of kings, wars, coups d’état and an ever-increasing state of desolation and ruin.84 According to the YP, the first king of the kaliyuga was Janamejaya Pārīkṣit, who quarrelled with the twiceborn. But the first blow to the ideal city of Puṣpapura (Pāṭaliputra?) came from a coalition between Pañcālas, Māthuras and (Indo-)Greeks (yavana), which brought political disorder in the different realms (vv. 47– 48) and was responsible for the appearance of adharma.85 Let it be noted in 82

The kṛtayuga is described as of supreme virtue (brāhmaṇyam uttamam), devoid of fear (bhaya), death (mṛtyu), thieves (taskara), greed (lobha), anger (krodha), passion (roṣa), deceit (dambha), calumny (paiśunya), meanness (kṣudratā), sexual union (maithuna), and full of (punitive) justice (daṇḍapradhānatā). Men were possessed of the strength of tapas. The devas, gandharvas, kinnaras, dānavas, yakṣas, rākṣasas and uragas were born during the kṛtayuga, which ends in a great battle (mahāyuddha)!

83

During the tretāyuga, the kṣatra was created by Svayambhū and feminine qualities (straiṇā guṇāḥ) as well as women (nārī) appeared; the four social classes were occupied with their own deeds (svakarman). The agnihotra was regularly performed and there were multitudes of śāstras, japas and mantras; the priests dwelt in forests (araṇyāyatana) and were learned in the Vedas (vedaparāyaṇa); human beings were not deceitful (aśaṭha), kṣatriyas were skilful in protecting people (prajāpālana), “judicious and steadfast guardians of the Vipras” (viprāṇāṃ viśeṣajñā rakṣitāraḥ samāhitāḥ, Mitchiner 1986: 89); vaiśyas delighted at their own deeds, and śūdras at service (śuśrūṣā). The world was devoted to dharma and absorbed in the practice of truth (satyavratasamāhita).

84

For a summary, see Parasher 1991: 239–241.

85

YP 50 and 52–55: brāhmaṇāḥ kṣatriyā vaiśyāḥ śūdrāś caiva yugakṣaye / samaveśāḥ samācārā bhaviṣyanti na saṃśayaḥ // […] cīravalkalasaṃvītā& jaṭāvalkaladhāriṇaḥ / bhikṣukā vṛṣalā loke hoṣyanti laghuvikriyāḥ // tretāgniṃ vṛṣalā loke hoṣyanti laghuvikriyāḥ / oṅkāraprathitair mantrair yugānte samupasthite // agnikārye ca japye ca agnike ca dṛḍhavratāḥ / śūdrāḥ kaliyugasyānte bhaviṣyanti na saṃśayaḥ // bhovādinas tathā śūdrā brāhmaṇāś cāryavādinaḥ / dharmabhītatamā vṛddhā janaṃ bhokṣyanti nirbhayāḥ //. &°saṃvītā em. (Alexis Sanderson, May 2010): °sañcītā Ed. “At the end of the yuga, Brahmins, kṣatriyas, vaiśyas and śūdras will be similar in dress, and of similar conduct—there is no doubt […] In the world, low-born bhikṣukas (mendicants), quick to

52

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

passing that the only occurrence of pāṣaṇḍa in the YP appears along with this description of adharma: “At that end of the yuga, men will be allied with pāṣaṇḍas: they will make friends for the motive of [gaining] women.”86 After the fall of the Yavanas and their disappearance from Madhyadeśa, seven kings will reign in Sāketa; a mighty king of the Śakas will follow their rule, but his kingdom will soon be destroyed. The YP then mentions the names of a series of rulers (the mleccha Āmrāṭa Lohitākṣa, who will destroy the four varṇas; Gopāla, a righteous king; Puṣpaka; Anaraṇya; the Brahmin Vikuyaśas; Agnimitra, who will quarrel with the Brahmins; Agniveśya, who will fight against the Śabaras). After their rule and the resultant shortage of men, women will become predominant (strīpradhāna). The king Satuvara “will afflict the earth with his rule” (hatvā daṇḍena medinīm, v. 87, Mitchiner 1986: 95), before the last rulers, the “terrible” (ghora) Śakas, “will destroy a quarter of living beings by arms” (caturbhāgaṃ ca śastreṇa nāśayiṣyanti prāṇinām, v. 89ab, Mitchiner 1986: 95) and “take a quarter of the wealth to their own city” (hariṣyanti śakāḥ kośacaturbhāgaṃ svakaṃ puram, v. 89cd, Mitchiner 1986: 95). Disasters and cataclysms herald the end of the kaliyuga. Note that the YP does not mention the advent of the kṛtayuga or Kalki(n), but explains that, as things get worse and the end of the yuga approaches, several geographical regions […] will be safe havens […] Although it does not explicitly say so, the text implies that at that time, and in those places, a new kṛtayuga will commence […] According to Mitchiner, the author of the YP was convinced that

anger, clothed in bark-cloth [and] wearing matted hair and bark-garments, will offer oblations into the three śrauta fires. In the world, once the end of the yuga has approached, low-born men—contemptible and loathsome—will offer oblations into the three śrauta fires with mantras manifested by [i.e., opening with] the sacred syllable ‘oṃ’. śūdras who practice with firm commitment the (daily? agnike to be read āhnike?) observance of the fire-sacrifice and the repetition [of mantras] will exist at the end of the yuga—there is no doubt. śūdras will also be utterers of ‘bho’, and Brahmins will be utterers of ‘ārya’: [and] the elders, most fearful of the dharma, will fearlessly exploit the people.” Translation (modified) Mitchiner 1986: 92. For v. 51, see below, n. 86. 86

YP 51ad1: pāṣaṇḍaiś ca samāyuktā narās tasmin yugakṣaye / strīnimittaṃ ca mitrāṇi kariṣyanti […] //. Translation Mitchiner 1986: 92.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

53

the kaliyuga had ended after the defeat of the Śakas and a new kṛtayuga was 87 beginning […].

This, again, accords well with Mārkaṇḍeya’s teachings, which clearly reflect the belief that a thousand-year kaliyuga would imminently come to its end. As suggested by González-Reimann and other scholars, [a]t some point in the last centuries BCE and the early centuries CE, there must have been real expectations that kali would end in the foreseeable future, but later, as social conditions did not bear out these expectations, the figures were reinter88 preted and the end of the yuga was pushed farther into the future.

Mitchiner dates the YP to the last quarter of the first century BCE,89 and remarks that the mention of the Ābhīras as mleccha rulers in the Mārkaṇḍeya section makes it unlikely that it is earlier than 250 CE, a conclusion which coincides with those of Dwivedi, Yadava, and R.S. Sharma, who, on the basis of the social, political, religious and economic conditions described, attribute the 90 Mārkaṇḍeya section on yugānta in the Epic to the third or fourth centuries CE.

Taking 250 CE as the point in time when the Mārkaṇḍeya section may have been composed or, at least, updated, we might conclude that, by this time, apocalyptic prophecies were still interpreting mleccha rule and political chaos as endtime symbols which were incomparably more significant than pāṣaṇḍa/pāṣaṇḍins. 1.2.5. Conclusion The passages examined so far yield consistent results. Purely cosmological depictions of the yugas such as MDhŚ 1.81–86 and MBh 3.148.5–37 do not allude to historical events, figures or threats when dealing with the kaliyuga. In contrast, passages dedicated to apocalyptic eschatology, i.e., prophecies such as MBh 3.186, 3.188 and the YP, do include lengthy ex 87

González-Reimann 2002: 99.

88

González-Reimann 2002: 97. According to Nattier, also the Buddhists regularly updated their reckoning of the saddharmavipralopa (see below, §1.3.2.1).

89

See also Parasher 1991: 239.

90

González-Reimann 2002: 99. See Dwivedi 1979: 290, Yadava 1978–1979: 32, Sharma 1982: 187–188, and Sharma 2001: 45–76.

54

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

post facto descriptions of their sociohistorical environment. And contrary to what would be expected from later Purāṇic accounts or interpolated passages of the MBh,91 pāṣaṇḍa and pāṣaṇḍins play virtually no role in these early prophecies. From hundreds of verses, only five can be interpreted as hinting at (signs of) allodoxies and/or allopraxies heralding the End: MBh 3.186.43 (bahupāṣaṇḍasaṅkīrṇā āśramāḥ), MBh 3.188.64 and 66 (eḍūka), MBh 3.189.9 (āśramāḥ hatapāṣaṇḍāḥ), and YP 51 (pāṣaṇḍaiś ca samāyuktā narāḥ). What does seem to matter in these apocalyptic and messianic foreseeings is the overwhelming presence of foreign, mleccha rulers alluded to in refrain-like manner and at times even mentioned by name. This, however, is hardly surprising on the hypothesis that the bulk of these texts were likely composed during the first 250 years of the first millennium CE, i.e., at a time when significant parts of northwestern and northern India were under Śaka, Pahlava and especially Kuṣāṇa rule. In other words, these predictions mirror the Brahmanical orthodoxy’s most dramatic concern of the day and are likely to be meant as incentives for the main addressee of the Epic, the dharma king. 1.2.6. The kaliyuga and its End in the Early Purāṇas The vāyuproktaṃ purāṇam referred to as an authoritative scripture at the close of the Mārkaṇḍeya section (MBh 3.189.14) underwent a process of reworking and lengthening that led to the composition, sometime during the fourth century CE (i.e., in the early Gupta era), of what Vielle has proposed to call the “‘classical’ vāyuprokta Brahmāṇḍapurāṇa.”92 As already recognized by scholars such as Kern and Kirfel, this text forms the nucleus of the extant VāP and BḍP.93 In other words, the common parts of these two originally identical Purāṇas are likely to provide us with the most ancient layer of extant Purāṇic literature. Fortunately, these Purāṇas’ accounts of the yugas, especially of the dvāpara- and kaliyuga, belong to this early layer, providing an extremely important landmark in the history of apocalyptic prophecies and hence, anxieties. VāP 58 ≈ BḍP 1.2.31 can be analyzed as follows: (1) a (gloomy, kaliyuga-like) description of the (second half of the) dvāparayuga (VāP 58.1–29 ≈ BḍP 1.2.31.1–29); (2) a 91

See below, §1.2.7.3.

92

See Vielle 2005. On MBh 3.189.14, see above, §1.2.4, and p. 50, n. 79.

93

See Vielle 2005: 537–543, and also Rocher 1986: 33, 157 and 244–245.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

55

description of the conditions and events characteristic of the kaliyuga (VāP 58.30–100 ≈ BḍP 1.2.31.30–100): (2a) a general depiction of the kaliyuga (VāP 58.30–72 ≈ BḍP 1.2.31.30–72); (2b) advent of Pramiti (VāP 58.73–88 ≈ BḍP 1.2.31.73–88); (2c) last events (VāP 58.89–100 ≈ BḍP 1.2.31.89– 100); (3) rise of a new kṛtayuga (VāP 58.101–126 ≈ BḍP 1.2.31.101– 126).94 In the following, I shall focus briefly on the first part (= 2a) of these two Purāṇas’ apocalyptic prophecy.95 Something is extremely striking about

94

Contrary to the Epic Kalki(n), Pramiti is not a brāhmaṇa (although he is surrounded by myriads of brāhmaṇas) and not directly responsible for the advent of a new kṛtayuga. After defeating thousands of mleccha kings (the famous list given in VāP 58.81–83 ≈ BḍP 1.2.31.81–83 includes, among others, draviḍas, siṃhalas, gāndhāras, pahlavas, yavanas, tuṣāras, barbaras, and cīnas) and slaughtering innumerable pāṣaṇḍas and adhārmikas, Pramiti dies. Only a few human beings survive the end of the kaliyuga, but in an increasingly miserable moral and material condition. According to VāP 58.99–101 ≈ BḍP 1.2.31.99–101, at one point, their condition is so pitiable that they become disgusted (nirveda), start pondering (vicāraṇa), attain equanimity (sāmyāvasthā), awakening (sambodha) and piousness (dharmaśīlatā), thus provoking the sudden rise of the new kṛtayuga (ahorātraṃ tadā tāsāṃ yugaṃ tu parivartate). In other words, Pramiti annihilates the moral and political conditions that were responsible for the prolongation of the kaliyuga and creates those that will eventually lead to the advent of the kṛtayuga.

95

Provided that the use of the future tense can be used as a stylistic criterion for apocalyptic discourse, only one part of this passage can properly be termed a “prophecy,” however. On phraseological and stylistic grounds, I am fairly convinced that VāP 58.30– 72 ≈ BḍP 1.2.31.30–72 is composed of at least two distinct layers. In VāP 58.30–41 ≈ BḍP 1.2.31.30–41, the future tense is missing entirely; tiṣya occurs four times instead of the more usual kali; the locatives tiṣye and kalau lack any fix syntactic position in the verses in which they occur. On the contrary, VāP 58.42–63 ≈ BḍP 1.2.31.42–63 is written in the future tense; tiṣya never occurs; the locative yugānte pary°/pratyupasthite (missing in VāP 58.30–41 ≈ BḍP 1.2.31.30–41, but strongly reminiscent of Epic formulations and thus likely to belong to the same source as that of the Mārkaṇḍeya Section, possibly the original vāyuproktaṃ [brahmāṇḍa]purāṇam) occurs five times as pāda d, here again in refrain-like manner (almost all of the locatives, i.e., yugānte, tasmin kalau yuge, yugānteṣu, yugakṣaye, occur in the verse’s last pāda). The last part of the passage, viz. VāP 58.64–72 ≈ BḍP 1.2.31.64–72, is of a more hybrid character. Note, however, that neither the future tense nor yugānte pary°/pratyupasthite occurs in VāP 58.64–70 ≈ BḍP 1.2.31.64–70, whereas both are found in the last two verses, viz. VāP 58.71–72 ≈ BḍP 1.2.31.71–72. On this passage of the VāP, see also Verardi 2011: 140.

56

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

this prophecy: whereas the mlecchas are never mentioned in this account,96 the pāṣaṇḍas are given much more importance than in the MBh and the YP, which testifies to a significant shift in emphasis having taken place. During the kaliyuga, “kings, nearly all śūdras, promote heretics(/heresies)”97 and everything is “filled with heretics of wrong morality and conduct.”98 Interestingly, our texts provide several details concerning the identity of the pāṣaṇḍas. Without a doubt, the following verse alludes to the Buddhists: And when the end of the [kali]yuga [will be] imminent, śūdras with white teeth, having subdued their senses, [with] shaved [heads] and wearing reddish garments 99 will practise religion.

And in a subsequent passage, our sources even go to the trouble of classifying the pāṣaṇḍas: And when the Kali age [will have been] reached, there will appear in like manner [people] wearing reddish garments as well as Nirgranthas and [people] wearing [human] skulls, others who [will] sell the Veda(s), others who [will] sell the sacred places(/objects), and yet other heretics who [will] challenge [the system of] 100 caste-classes and life stages.

The Buddhists, the Jains and the Kāpālikas start a long career as unmistakable signs of the (end of the) kaliyuga.

96

Note that the mlecchas occur, most probably as a literary stereotype, in the prophecy concerning the exploits of Pramiti; see above, p. 55, n. 94, and VāP 58.78, 84, 89 ≈ BḍP 1.2.31.78, 84, 89.

97

VāP 58.40ab ≈ BḍP 1.2.31.40ab: rājānaḥ śūdrabhūyiṣṭhāḥ pāṣaṇḍānāṃ& pravartakāḥ /. & BḍP reads pākhaṇḍānāṃ.

98

VāP 58.52ab ≈ BḍP 1.2.31.52cd: kuśīlacaryāpāṣaṇḍair vṛthārūpaiḥ& samāvṛtam' /. &BḍP reads vyādharūpaiḥ. 'samāvṛtam em. (BḍP): samāvṛttam VāP.

99

VāP 58.59 ≈ BḍP 1.2.31.59cd–60ab: śukladantā jitākṣāś ca muṇḍāḥ kāṣāyavāsasaḥ / śūdrā dharmaṃ cariṣyanti yugānte paryupasthite& //. &BḍP reads samupasthite.

100

VāP 58.64cd–66ab ≈ BḍP 1.2.31.65–66: kāṣāyiṇaś& ca nirgranthās tathā kāpālinaś' ca ha // vedavikrayiṇaś cānye tīrthavikrayiṇo 'pare / varṇāśramāṇāṃ ye cānye pāṣaṇḍāḥ( paripanthinaḥ // utpadyante tathā) te vai samprāpte* tu kalau yuge /. &kāṣāyiṇaś em. (BḍP): kaṣāyiṇaś VāP. 'BḍP reads kāpālikāś. (BḍP reads pākhaṇḍāḥ. )BḍP reads tadā. * samprāpte em. (BḍP): saprāpte VāP.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

57

1.2.7. Kaliyuga and Heresiology in the ViP 1.2.7.1. As we shall see in the following, the ViP strengthens the tendency observed in the VāP and the BḍP. Although Hazra and Pathak101 date it to the beginning of the fourth century (i.e., the dawn of the Gupta dynasty), scholars such as Winternitz, Renou, Hacker, Gail, Doniger O’Flaherty and González-Reimann102 are rather inclined to date the ViP to the fifth or the beginning of the sixth century, i.e., in the last part or the close of the Gupta rule over northern India. 1.2.7.2. The ViP (6.1) also contains an apocalyptic prophecy describing the miserable condition of mankind during the kaliyuga. Here, Parāśara answers Maitreya’s question about the nature of the kaliyuga (kaleḥ svarūpam) “in which the blessed four-footed dharma gets ruined.”103 This prophecy betrays a strong concern for the immorality and misbehaviour of women: self-willed, frivolous, speaking harsh and false words, women will marry wealthy men only, have numerous progeny but little fortune, stuff themselves and have unlimited power over their husbands.104 The kaliyuga will also be characterized by various disorders in the stages of life, 105 members of the different varṇas will attach themselves to prohibited activities,106 kings will fail to protect people and despoil them by levying excessive taxes.107 Starvation, mainly due to draught, will plague people and doom them to exile and the violation of dietary rules.108 Life will not exceed twenty years and young girls of the age of five, six or seven will give birth to children.109 Human understanding will decline, whereas vices, defilements and sin will increase.110 Decisions and hierarchy will depend 101

See Hazra 1937 and Pathak 1997: I.22–23.

102

See Winternitz 1981: 521, Renou 1985: 418 (§834), Hacker 1960: 67, Gail 1969: 921, Doniger O’Flaherty 1983: 108 (between 400 and 500 CE), González-Reimann 2002: 170–171 (fifth century). For other tentative dates, see Rocher 1986: 249.

103

ViP 6.1.8ab: dharmaś catuṣpād bhagavān yasmin viplavam ṛcchati //. See ViP 6.1.8–9.

104

ViP 6.1.16–18, 21, 28–31.

105

ViP 6.1.10, 14, 32–33.

106

ViP 6.1.10, 23, 36–37, 53.

107

ViP 6.1.34–35, 40.

108

ViP 6.1.24–26, 38, 52–54.

109

ViP 6.1.39–43.

110

ViP 6.1.43, 57–58.

58

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

on wealth alone.111 Rituals will be abandoned or left to humans’ fancy or arbitrariness, since the Vedas will be neglected and religious authority will amount to free thinking.112 Considering these and other features, the ViP’s prophecy may not seem to differ in any significant way from that of the MBh. But when it comes to foreign rule and heresy, the shifts are more obvious than those found in the VāP and the BḍP. First, ViP 6.1 does not allude to mlecchas at all (although ViP 4.24.98 still has Viṣṇu’s avatāra Kalkin destroy the mlecchas and dasyus and re-establish people in their svadharmas). Second, and again in stark contrast to the MBh, heresy and heretics are now seen as one of the most prominent features of the kaliyuga. Consider first the following part of the prophecy: Given to the vow of living on alms and bearing the marks of homelessness, the vilest śūdras, [duly] honoured, will betake themselves to a livelihood based on heresy. And the Vedic path having disappeared, people being beset with heresy, 113 human beings will have a short life due to the increase of adharma.

Most interesting are, however, verses 6.1.45 and 49–51, which are worth quoting in full: As soon as an increase in heresy comes to notice, O Maitreya, the high-minded should infer an increase of kali […] When there is no delight in Vedic teachings [but] fondness for heresies, the intelligent wise ones should infer an increase of kali. During the kali[yuga], O Maitreya, people sullied by heresies will fail to honour Viṣṇu, the lord of the world, the god who is the creator of everything. People struck by heresies will speak thus: “Wherefore the gods, wherefore the Brahmins (vipra), wherefore the Vedas, wherefore the purity born of the ablu114 tions?”

111

ViP 6.1.12, 16, 19–20, 22.

112

ViP 6.1.10, 11, 13–15, 27.

113

ViP 6.1.37 and 39: bhaikṣavrataparāḥ śūdrāḥ pravrajyāliṅgino 'dhamāḥ / pāṣaṇḍasaṃśrayāṃ vṛttim āśrayiṣyanti satkṛtāḥ // vedamārge pralīne ca pāṣaṇḍāḍhye tato jane / adharmavṛddhyā lokānām alpam āyur bhaviṣyati //.

114

ViP 6.1.45 and 49–51: yadā yadā hi pāṣaṇḍavṛddhir maitreya lakṣyate / tadā tadā kaler vṛddhir anumeyā mahātmabhiḥ // […] na prītir vedavādeṣu pāṣaṇḍeṣu yadā ratiḥ / kaler vṛddhis tadā prājñair anumeyā vicakṣaṇaiḥ // kalau jagatpatiṃ viṣṇuṃ sarvasraṣṭāram īśvaram / nārcayiṣyanti maitreya pāṣaṇḍopahatā janāḥ // kiṃ devaiḥ kiṃ dvijair vedaiḥ kiṃ śaucenāmbujanmanā / ity evaṃ vipra vakṣyanti pāṣaṇḍopahatā janāḥ //.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

59

As we can see, the insistence on heresy and the total absence of foreign rulers betray a significant evolution from the MBh and the YP, as well as a consolidation of the features observed in the VāP and the BḍP. 1.2.7.3. That this strong emphasis on heresy is not just peripheral or incidental to the ViP can be seen from many other passages in this early Purāṇa. Most significant in this connection is a famous myth which likely finds its first occurrence in our text. It would soon provide the ideological background for the seventh-century integration of the Buddha into the list of Viṣṇu’s avatāras. According to this myth, Viṣṇu produced a deluder or heresiarch called Māyāmoha (or, in other accounts, Māyāpuruṣa) out of his own body in order to vanquish the Asuras (or Daityas) who were invincible due to their pious Vedic observances.115 The early date of the ViP’s version of the story can be established both by the fact that the heresiarch is not yet considered as an avatāra of Viṣṇu,116 and by the fact that his bewildering preaching is not located at the beginning of the kaliyuga, as most later versions of the story have it. As we shall see, this myth sheds more light on the identity of the heretics alluded to in the prophecy discussed above.117 115

On this and parallel myths, see Choudhary 1956: 240–243, Doniger O’Flaherty 1971: 306–333 (especially pp. 308–313), Klostermaier 1989: 54–55, Dandekar 1997 (especially pp. 146–148), Pathak 1997: 24 (who claims, on rather shaky arguments, this myth to be a later addition), Saindon 2004: 24–41 (especially pp. 34–35), and Verardi 2011: 265–266, and 275. Note also González-Reimann 2002: 191, n. 32: “That heretic teachers who did not accept the existence of the ātman and criticized the Vedas and animal sacrifices were really gods who incarnated in order to confuse—and thus destroy—the demons, is an idea that had been stated earlier in the MaitrUp (7.8–10), but with no mention of the yugas.” That the MaitrUp is earlier than the ViP is far from certain.

116

The Buddha is not yet regarded as Viṣṇu’s avatāra in the ViP or the Harivaṃśa. According to Hazra 1940: 41–42, “it is highly probable that the Buddha began to be regarded as an incarnation of Viṣṇu from about 550 A.D.” (quoted in Saindon 2004: 20). According to Hazra (1940: 41, 84, 88) and Kane (1997: 721), this incarnation became popular during the seventh century, as is testified by a Pallava inscription dated to the end of the seventh century (see Kane 1997: 721, n. 1725 and Mahalingam 1988: no. 250). The doctrine of Viṣṇu’s ten avatāras (with the Buddha as the ninth, taking place at the beginning of the kaliyuga, and Kalki[n] as the tenth and final one, at the end of the kaliyuga) is well established in the Matsya°, Garuḍa° and Agnipurāṇa (see Choudhary 1956: 239, nn. 1 and 9 for references).

117

That the Jaina anekāntavāda and the Buddhist madhyamaka and vijñānavāda (see below) are meant here as heresies (pāṣaṇḍa) is made clear by ViP 3.18.23, which describes them in the following way: anyān apy anyapāṣaṇḍaprakārair bahubhir dvija / daiteyān mohayām āsa māyāmoha ’timohakṛt //. “By the numerous ways of [yet] other

60

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

Maitreya asks Parāśara about the identity, the practice/behaviour (samācāra) and the nature of the nagna, i.e., the heretic (ViP 3.17.4). Parāśara’s answer is as follows: The triple [sacred science], called the ṛc, the yajus and the sāman, is the [protecting] cover (āvṛti) of the caste-classes, O twiceborn; one who is “naked” is any sinner who through delusion casts off that garment. Since, O twiceborn, the triple [sacred science] is the garment of all the caste-classes, one is [said to be] nagna 118 when it has been abandoned, O twiceborn, no doubt [about this].

Parāśara then sets about telling the story that his grandfather Vasiṣṭha had already recited to Bhīṣma. After a grandiose hundred-year battle, the gods were defeated by the demons (asura, dait[e]ya). They went to see Viṣṇu on the shore of the ocean of milk, praised him and asked him how to vanquish the demons: We have not been able to slay those enemies who, [because they are] intent upon the duties of their own caste-classes and follow the Vedic path, are protected (āvṛta) by [their] tapas. Be pleased to give us, O soul of all, the means which 119 might enable us, O Bhagavat, to slay these demons.

As we can see, the Asuras owe their invulnerability to their strict adherence to the Veda and the rules laid down in it. As an answer, Viṣṇu produced a Māyāmoha out of his own body, handed it over to the gods and said: This Māyāmoha will bewilder all these demons. As a result [of this bewilderment], it will be possible to kill them [since they will be] excluded from the Vedic 120 path.

heresies, Māyāmoha, the absolute deluder (atimohakṛt) deluded all the other demons, O twiceborn.” 118

ViP 3.17.5–6: ṛgyajuḥsāmasañjñeyaṃ trayī varṇāvṛtir dvija / etām ujjhati yo mohāt sa nagnaḥ pātakī dvija // trayī samastavarṇānāṃ dvija saṃvaraṇaṃ yataḥ / nagno bhavaty ujjhitāyām atas tasyāṃ na saṃśayaḥ //. On this explanation of nagna, see also Verardi 2011: 265–266, 269, 285.

119

ViP 3.17.39–40: svavarṇadharmābhiratā vedamārgānusāriṇaḥ / na śakyās te 'rayo hantum asmābhis tapasāvṛtāḥ // tam upāyam aśeṣātmann asmākaṃ dātum arhasi / yena tān asurān hantuṃ bhavema bhagavan kṣamāḥ //.

120

ViP 3.17.42: māyāmoho 'yam akhilān daityāṃs tān mohayiṣyati / tato vadhyā bhaviṣyanti vedamārgabahiṣkṛtāḥ //.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

61

Māyāmoha sets out and sees the demons practising austerities on the bank of the Narmadā river. Disguised first as a Jaina ascetic, i.e., naked (digambara), bald (muṇḍa) and carrying a peacock-feather fan, Māyāmoha asks them whether they expect this-worldly (aihika) or other-worldly (pāratrya) results from their austerities, and invites them to conform to his teaching (vākya, dharma) if they really long for liberation (mukti). Māyāmoha then teaches them the anekāntavāda,121 and the demons are taken away from the Vedic path (vedamārgād apākṛtāḥ) and made to give up their svadharma. The demons spread this delusive teaching, and “within a few days, the triple [sacred science] was generally abandoned by these demons.”122 Māyāmoha then addresses other Asuras under the guise of a Buddhist monk, i.e., wearing red garments (raktāmbaradhṛt): If you long for heaven or for nirvāṇa, O Asuras, then away with the corrupted dharmas of animal slaughter, etc., [and] listen [to me]. Know that all this only consists in mental representation. Understand my speech, which has [already] been told in this way by perfect sages (samyagbudha) in the [world]. This world, being without a foundation, is addicted to the objects of a [purely] erroneous cognition; corrupted by [defilements] such as attachment, it is made to roam about 123 incessantly in the narrow corridor/defile of existence.

As the author(s) of this passage insist, the Buddhist Māyāmoha seduces the demons mainly through reason(ing) (yukti) and argumented speech (vacanaṃ yuktiyojitam), urging them to renounce the Vedas in the following way: Trustworthy words do not fall on earth from the sky, O [you my sweet] Suras; I and others like you should [only] accept words possessed with reason(ing).124

121

See ViP 3.18.9–10. Skt. anekāntavāda occurs at ViP 3.18.12a.

122

ViP 3.18.15cd: alpair ahobhiḥ santyaktā tair daityaiḥ prāyaśas trayī //.

123

ViP 3.18.17–19: svargārthaṃ yadi vo vāñchā nirvāṇārtham athāsurāḥ / tad alaṃ paśughātādiduṣṭadharmair nibodhata // vijñānamayam evaitad aśeṣam avagacchata / budhyadhvaṃ me vacaḥ samyagbudhair evam ihoditam // jagad etad anādhāraṃ bhrāntijñānārthatatparam / rāgādiduṣṭam atyarthaṃ bhrāmyate bhavasaṅkaṭe //.

124

ViP 3.18.31: na hy āptavādā nabhaso nipatanti mahīṃ surāḥ / yuktimad vacanaṃ grāhyaṃ mayānyaiś ca bhavadvidhaiḥ //. surāḥ in place of the expected (but metrically impossible) asurāḥ is a rather clumsy formulation. As Alexis Sanderson (personal communication, May 2010) suggests to me, this might also be explained as coaxing flattery.

62

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

According to him, to consider ritual violence (hiṃsā) to be virtuous does not stand to reason(ing) (na yuktisahaṃ vākyam), just as it is childish speech (arbhakodita) to consider that offerings (havis) burnt in fire yield results. As a result of his bewildering teaching and the Asuras’ spreading it further, the latter gave up their “native/innate” (nija) dharma, the dharma of the triple sacred science (trayīdharma; or: taught in the Veda and the smṛtis—vedasmṛtyudita) and all instruction based on the Vedic path (sarvāṃ trayīmārgāśritāṃ kathām). And while some of the demons started condemning (vinindā) the Vedas, others despised the gods and still others reviled the whole complex of sacrifices and ritual acts (yajñakarmakalāpa). Here is the predictable conclusion of Parāśara’s story: These demons having thus gone astray, the immortal [gods] undertook most excellent preparations and [made themselves] ready for fighting. Then, O twiceborn, a battle between gods and demons again took place, and the demons, [now] enemies 125 of the true [Vedic] path (sanmārga), were slain by the gods.

A few verses later, the author(s) draw(s) the moral of the story: This disadvantage was born of [the demons’] having conversed with heretics […] 126 Therefore, one should avoid conversation and contact with the sinful heretics.

In other words, the Māyāmoha myth serves as an aetiological foundation for the normative literature’s recurrent prohibition of any kind of contact with pāṣaṇḍins. According to the MDhŚ, one “should not give honour, even with mere words, to heretics,”127 nor should one “live in a kingdom ruled by a servant […] or overrun by gangs of heretics.”128 It is hardly surprising, then, that the ViP prohibits conversation (sambhāṣaṇa, ālāpa), 125

ViP 3.18.33–34: ittham unmārgayāteṣu teṣu daityeṣu te 'marāḥ / udyogaṃ paramaṃ kṛtvā yuddhāya samupasthitāḥ // tato daivāsuraṃ yuddhaṃ punar evābhavad dvija / hatāś ca te 'surā devaiḥ sanmārgaparipanthinaḥ //.

126

ViP 3.18.79a and 96ab: pāṣaṇḍālāpajāto ’yaṃ doṣaḥ […] tasmāt pāṣaṇḍibhiḥ pāpair ālāpasparśanaṃ tyajet /. See also ViP 3.18.98 (where the sinful heretics are termed vedavādavirodhin) and 3.18.103.

127

MDhŚ 4.30: pāṣaṇḍinaḥ […] vāṅmātreṇāpi nārcayet //. Translation Doniger/Smith 1991: 77. See also YSmṛti 1.130 and 271–272 (Joshi 1987: 311–312). YSmṛti 2.70 and NSmṛti 4.180 disqualify heretics and deniers as witnesses (Doniger O’Flaherty 1983: 121).

128

MDhŚ 4.61: na śūdrarājye nivaset […] na pāṣaṇḍigaṇākrānte […]. Translation Doniger/Smith 1991: 79.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

63

civility (anupraśnā), physical contact (sparśana), cohabitation or commensality with heretics. And whereas the mere sight of a heretic requires one to stare at the sun for purification, physical contact with him dooms one to ablutions.129 The Māyāmoha section ends with the following conclusion: One should not give honour, even with mere words, to heretics, people who persist in wrong action, people who act like cats, hypocrites, rationalists, and people who live like herons. As for contact, even at a distance, with these extremely sinful, misbehaving heretics, it should be avoided. Therefore, may one shun them. These are the naked ones whom I have presented to you, who [upon being] seen ruin funerary rites, people conversation with whom destroys the merit 130 of one day.

1.2.7.4. As the normative conclusions drawn by the author(s) already testified, the Māyāmoha myth, far from being a benign story, is expected to have a direct and compelling bearing upon those listening and adhering to the teachings of the ViP. This relevance for contemporary society did not escape the attention of those who, at an indeterminate point in time, adapted and interpolated the story into the MBh.131 In doing so, these late editors filled what is likely to have been perceived as a lack in the great Epic, viz. its paying no or only marginal (explicit) attention to heretics as heralding the kaliyuga. The Buddha, now Viṣṇu’s last avatāra before Kalki(n), is depicted as active at the beginning of the kaliyuga; the narrative framework of the devāsura struggle has been cut off in order to anchor the episode in human time: When the kaliyuga begins, I [Viṣṇu] will speak at the court of righteous kings in the language of Magadha after sitting under the rājataru tree. As Buddha, the son of Śuddhodana, wearing a reddish garment, [with my head] shaved and with white teeth, I will confuse mankind. When I become the Buddha, the śūdras will enjoy [the benefits of] the ceremonies for the ancestors; all men will shave [their heads],

129

See ViP 3.18.43–50 and 95–96.

130

ViP 3.18.100–102: pāṣaṇḍino vikarmasthān vaiḍālavratikāñ chaṭhān / haitukān bakavṛttīṃś ca vāṅmātreṇāpi nārcayet // dūratas tais tu samparkas tyajyaś cāpy atipāpibhiḥ / pāṣaṇḍibhir durācārais tasmāt tān parivrajayet // ete nagnās tavākhyātā dṛṣṭāḥ śrāddhopaghātakāḥ / yeṣāṃ sambhāṣaṇāt puṃsāṃ dinapuṇyaṃ pranaśyati //. ViP 3.18.100 = MDhŚ 4.30 (translation Doniger/Smith 1991: 77). On the cats’ and the herons’ behaviour, see MDhŚ 4.195–196.

131

MBh vol. 12.2/XVI: 2104a–2105a, Appendix I.31.

64

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

and they will dress in red. The brāhmaṇas will not study [the Vedas], and they will not keep the [ritual] fire […] When the kaliyuga ends [I will be born as] Kalkin, a yellowish-brown brāhmaṇa, son of Viṣṇuyaśas and having a Yājñavalkya chaplain. And with the help of all my brāhmaṇa followers I will completely 132 destroy the foreigners and the heretics.

The son of Śuddhodana is reported to rely on the bauddhadharma and to teach compassion (dayā) toward living beings; seeing the whole world as merely consisting of the skandhas, śūdras will make donations to śūdras, people will engage in actions that merit rebirth in hell (narakārha), the Brahmins will be devoid of purity and neglect their svadharma, sons will pay no obeisance to their fathers, women to their husbands, servants to their masters, and the mixture of caste-classes (varṇasaṅkara) will prevail.133 1.2.7.5. Here, as elsewhere, the meaning of the Māyāmoha myth (and of the concomitant adoption of the Buddha as an avatāra of Viṣṇu134) is perfectly clear: What occurred to the demons of old threatens contemporary society if it allows heretical indoctrination to undermine its Vedic foundations and the Brahmins’ symbolic and ritual power. Decades or even centuries of indigenous Gupta rule have made the mleccha threat momentarily recede into the background of apocalyptic eschatology. Instead, the most conservative milieux now perceive allodoxy, mainly Buddhism and Jainism, as the most threatening challenge to Brahmanism and accordingly update their prophecies and create new myths. Note, however, that references to mlec132

MBh vol. 12.2/XVI: 2104ab, Appendix I.31, lines 1–7, 18–21: tataḥ kaliyugasyādau bhūtvā rājataruṃ śritaḥ / bhāṣayā māgadhenaiva dharmarājagṛhe vadan / kāṣāyavastrasaṃvīto muṇḍitaḥ śukladantavān / śuddhodanasuto buddho mohayiṣyāmi mānavān / śūdrāḥ śrāddheṣu bhojyante mayi buddhatvam āgate / bhaviṣyanti narāḥ sarve muṇḍāḥ kāṣāyasaṃvṛtāḥ / anadhyāyā bhaviṣyanti viprāś cāgnivivarjitāḥ / […] tataḥ kaliyugasyānte brāhmaṇo haripiṅgalaḥ / kalkir viṣṇuyaśaḥputro yājñavalkyapurohitaḥ / sahāyā brāhmaṇāḥ sarve tair ahaṃ sahitaḥ punaḥ / mlecchān utsādayiṣyāmi pāṣaṇḍāṃś caiva sarvaśaḥ //. Translation González-Reimann 2002: 171 (slightly modified according to the interpretation [Alexis Sanderson, personal communication, May 2010] of yājñavalkyapurohita as a bahuvrīhi compound).

133

See Saindon 2004: 29–30 for the Sanskrit text and a French translation of the passage.

134

Note Klostermaier 1989: 54: “Some authors seem to think that, among the avatāras of Viṣṇu, Buddha would have been received in a spirit of tolerance. But the way in which this Buddha-avatāra is described in the ViP and the general Hindu attitude of considering both good and evil as coming from the same Supreme Being would suggest that we have here an early and unmistakably hostile Hindu text dealing with Buddhism.”

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

65

chas remain abundant in Purāṇic literature, especially in connection with Kalki(n) (or Pramiti/Pramati) and the famous Purāṇic motif of the dynasties of the kaliyuga.135 What I intend to stress here is that considered in themselves, the prophecies concerning the kaliyuga mirror significant shifts regarding the elements at work during the last age. According to Doniger O’Flaherty, the Māyāmoha myth, a “Gupta myth,”136 mirrors “the Paurāṇikas’ hope that the Gupta kings would exterminate heretics.”137 It has even been conjectured that these conservative segments of society transferred the messianic myth connected to the advent of Kalki(n) (or Pramiti/Pramati) on historical rulers such as Candragupta II Vikramāditya (r. 375–415)138 or Yaśodharman of Malvā, who defeated the Hūṇas around 530.139 As is testified to by the vāyuproktaṃ brahmāṇḍapurāṇam, heretics and heresy seem to enter apocalyptic discourse and anxieties in the early Gupta era and have clearly become their central concern as the Gupta empire starts to crumble. This is likely to reflect the Brahmanical confidence with, and sociopolitical expectations towards Gupta, bhāgavata rule, i.e., with a partly restorative dynasty noted, e.g., for the repeated performance of the aśvamedha. But this may also betray Brahmanical exasperation with regard to this dynasty’s policy of toleration—rather than “tolerance”—towards the pāṣaṇḍas.140 Thus, by the turn of the sixth century, the Brahmanical

135

See Parasher 1991: 241–244.

136

Doniger O’Flaherty 1983: 117.

137

Doniger O’Flaherty 1983: 125. A few pages earlier, Doniger O’Flaherty terms this story “the most important of the anti-Buddhist myths of this period” (1983: 108). According to her (1983: 108), “[i]n the Purāṇic texts, the Buddha was depicted as deluded and destructive, for these texts of the Gupta period were weapons in a battle between the instigators of the Hindu revival and the still thriving establishments of Jainism and Buddhism.” As Yuko Yokochi has rightly pointed out to me, however, one should refrain from considering the Purāṇic milieux in their entirety (especially the Śaiva) as socially conservative and promoting Vedic restoration.

138

See Agrawala 1963: 228–231 and Doniger O’Flaherty 1983: 122–127.

139

See Jayaswal 1917.

140

For a thorough critique of the traditional, nationalistic account of the Gupta era as India’s golden age and the Guptas as religiously tolerant and egalitarian rulers, see Verardi 2011: 128–196. According to Verardi (2011: 128), it is not possible “to speak of a patronage dispensed with equal generosity to brāhmaṇas and śramaṇa-s—a still current opinion which is not only contrary to historical truth, but to every historical likelihood.”

66

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

orthodoxy’s hostility towards Buddhism and Jainism crystallizes in a complex network of behavioural prescriptions, apocalyptic and messianic prophecies, aetiological myths and literary tropes deriding Buddhist monks and nuns.141 It is easy to show that this complex network had a direct bearing on at least one decisive segment of Indian philosophy during the sixth century CE, Kumārila Bhaṭṭa’s Mīmāṃsā. 1.2.8. Kaliyuga and Heresiology in the Mīmāṃsā By the end of the fifth century CE, the bulk of the Mīmāṃsā still consists in a complex system of ritual exegesis introduced by an increasingly large but philosophically shaky section dedicated to Vedic apologetics. Sometime in the middle of the sixth century, however, Kumārila Bhaṭṭa provides the system with a new shape and purport by turning it into a socially conservative and even restorative philosophical enterprise centred on the intrinsic and unconditioned validity of the Veda, as well as the normativity of all that can claim to be “rooted” in the Veda (vedamūla).142 Kumārila’s endeavour is not only apologetic (showing that the Veda, especially its positive and negative injunctions, provides the only access to the supersensible dharma), but also, perhaps even mainly, heresiological (refuting all extra-Vedic accounts of religious authority, reality, and epistemology). Since Buddhism radically challenges Vedic authority, the caste system, and Brahmanical prerogatives in symbolic and ritual matters, it becomes Kumārila’s favourite target. In his ŚV, however, Kumārila rarely ventures beyond the limits of carefully argumented and well informed philosophical debate when criticizing Buddhist doctrines. Things are quite different in the TV, Kumārila’s commentary on adhyāyas 2 and 3, which, being only marginally concerned with philosophy proper, is very likely addressing a different, more ritually oriented audience. Here, Kumārila leaves the well-mannered ground of philosophical exchange to give free rein to his abhorrence and hostility towards Buddhism, allowing us to look More specifically (2011: 139), the “Guptas fulfilled the political answer that the authors of the early Kali Age literature were overtly expecting and pressing for.” 141

See Bloomfield 1924 and Kher 1979.

142

As McCrea (2012) rightly insists, by no means does this imply that Kumārila’s philosophy can be adequately characterized as orthodox in the primary, purely doctrinal sense of the word. Quite to the contrary, Kumārila’s philosophy is of a decidedly innovative and thought-provoking character.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

67

deeper into the real, non-philosophical motivations of Kumārila’s radical anti-Buddhist stance. And Purāṇic apocalyptic eschatology is likely to bear a significant responsibility on Kumārila’s heresiological attitude and rhetorical strategy. Kumārila’s familiarity with at least part of the apocalyptic prophecies considered above is clear from the following statement: It is recorded in the Purāṇas that in the kali [age, heretics] such as the Buddhists 143 will cause the ruin of the dharma. Who deigns to listen to their speech?

This important verse encapsulates all we already know about the Purāṇic emphasis on heterodoxy as a sign of the last age, on the prohibition of verbal and physical contact with the heretics, and on the latter’s responsibility for the ruin of the dharma. Consider now the following passage: By arguing and exhibiting that they contradict the śruti, we show [in the present section] that the [treatises] that are not adopted by those who are familiar with the triple [Vedic science] cannot be relied upon [in order to cognize dharma. These consist of] the treatises of [mixed] religion-cum-irreligion that are adopted by the Sāṅkhya, the Yoga, the Pāñcarātra, the Pāśupata, the Buddhists and the Jains; they aim at nothing but proselytizing people, acquiring [wealth and gaining public] veneration and renown; their composition is [entirely] based on reason(ing) [but such a reason(ing)] that is contrary to the triple [Vedic science], where the greed and other [vices of their authors] are manifest, and that resorts mainly to perception, inference, analogy and presumption; by exhibiting the occasional success of a few [magic] formulas and decoctions capable of [things] such as curing poison, subjugating, ruining and maddening [other people], they teach various things that pertain mainly to [mere] livelihood [but] are perfumed by a few elements agreeing with the śruti and the smṛtis, such as non-violence, 144 truthfulness, self-restraint, giving, compassion.

143

TV ad MīSū 1.3.7/II.123,24–25: smaryante ca purāṇeṣu dharmaviplutihetavaḥ / kalau śākyādayas teṣāṃ ko vākyaṃ śrotum arhati //. According to Verardi 2011: 275 and 319, n. 58 (referring to Kane 1997: 721–722), Kumārila did not recognize the Buddha as an avatāra of Viṣṇu. I am, however, not aware of any explicit statement of Kumārila to this effect.

144

TV ad MīSū 1.3.4/II.112,17–24: yāny etāni trayīvidbhir na parigṛhītāni […] lokopasaṅgrahalābhapūjākhyātiprayojanaparāṇi trayīviparītāsambaddhadṛṣṭalobhādi&pratyakṣānumānopamānārthāpattiprāyayuktimūlopanibaddhāni sāṅkhyayogapāñcarātrapāśupataśākyanirgranthaparigṛhītadharmādharmanibandhanāni viṣacikitsāvaśīkaraṇoccā-

68

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

As Halbfass has repeatedly pointed out, Kumārila’s endeavour must be understood as an attempt to neutralize ethical and religious relativism by demonstrating that only the Veda and the Veda-based smṛtis can supply behavioural and practical maxims.145 In this passage, Kumārila identifies the pseudo-, heterodox smṛtis challenging the śruti and smṛti by laying down norms that, for the most part, are found to contradict the Veda-based dharma. And as we can see, Kumārila is far from attempting to federate all the denominations that post-classical India and Western scholarship have classified as “hinduistic” or even “orthodox.” According to him, Sāṅkhya, Yoga, Pāñcarātra and Pāśupata are no less heterodox than Buddhism and Jainism. Interestingly, Kumārila’s emphasis on their vile and deceitful seduction echoes the ViP’s depiction of the heresiarch Māyāmoha. Moreover, most of the clichés that had long been associated with heretics recur in Kumārila’s description, the most significant one being their resorting to autonomous, non-scripturally based reason(ing) and argumentation. As Kumārila says, these are the revilers of the Veda (vedanindaka), the “sophists” (haituka) and the deniers (nāstika) who according to Manu should be carefully avoided by righteous Brahmins. In another passage of his TV, Kumārila adduces three arguments in order to deny the Buddhist dharma any Vedic foundation: With the exception of a few statements about [things] such as self-restraint and giving, all the statements of [heretics] such as the Buddhists contradict all the fourteen [authoritative] disciplines of knowledge and have been uttered by [people,] such as the Buddha, whose conduct was swerving from and contradictory to the path of the triple [Vedic science. And] since they have been delivered to deluded [people] who are external to the triple [Vedic science] and, for the most part, representatives of the fourth caste-class [i.e., śūdras], they cannot be rooted 146 in the Veda.

ṭanonmādanādisamarthakatipayamantrauṣadhikādācitkasiddhinidarśanabalenāhiṃsāsatyavacanadamadānadayādiśrutismṛtisaṃvādistokārthagandhavāsitajīvikāprāyārthāntaropadeśīni […] teṣām evaitacchrutivirodhahetudarśanābhyām anapekṣaṇīyatvaṃ pratipādyate /. &°lobhādi° em. (Alexis Sanderson, May 2010): °śobhādi° Ed. My translation of this passage is tentative (I owe the translation of dharmādharma as “religion-cum-irreligion” to Alexis Sanderson [personal communication, May 2010]). 145

See, e.g., Halbfass 1991: 87–129.

146

TV ad MīSū 1.3.4/II.113,22–114,1: śākyādivacanāni […] katipayadamadānādivacanavarjaṃ sarvāṇy eva samastacaturdaśavidyāsthānaviruddhāni trayīmārgavyutthitaviru-

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

69

First, most of the Buddhist tenets contradict the Veda-based vidyāsthānas, a point he makes again elsewhere when spelling out the names of the authoritative disciplines: Indeed, the disciplines of knowledge accepted by the learned [ones] as authoritative regarding dharma are strictly limited in number, fourteen or eighteen, and are called “Veda,” “Secondary Sciences” (upaveda), “Limbs [of the Veda]” (aṅga), “Secondary Limbs [of the Veda]” (upāṅga), “Eighteen Religious Collections” (aṣṭādaśadharmasaṃhitā), Purāṇas, “Recensional Phonetics” (śākhāśikṣā), and “Politics” (daṇḍanīti). But the works of [heretics] such as the Buddhists or the 147 Jainas are neither recorded nor accepted among them.

Second, the Buddha, considered a kṣatriya, is found guilty of having transgressed his svadharma through his teaching activities. And third, he preached his dharma to śūdras whom the Brahmanical orthodoxy denies any contact with the Veda. In short, the Buddhist dharma cannot lay claim to vedamūlatva. The threat of ethical, doctrinal and ritual relativism is stressed in yet another interesting passage of the TV. But here, Kumārila explains the decisive role that accrues to the orthodox literati and philosophers in the gloomy last age:

ddhācaraṇaiś ca buddhādibhiḥ praṇītāni / trayībāhyebhyaś caturthavarṇaniravasitaprāyebhyo vyāmūḍhebhyaḥ samarpitānīti na vedamūlatvena sambhāvyante /. 147

TV ad MīSū 1.3.7/II.122,2–5: parimitāny eva hi caturdaśāṣṭādaśa vā vidyāsthānāni dharmapramāṇatvena śiṣṭaiḥ parigṛhītāni vedopavedāṅgopāṅgāṣṭādaśadharmasaṃhitāpurāṇaśākhāśikṣādaṇḍanītisañjñakāni / na ca teṣāṃ madhye bauddhārhatādigranthāḥ smṛtā gṛhītā vā /. Note also TV ad MīSū 1.3.4/II.114,26–115,10: eta eva ca te yeṣāṃ vāṅmātreṇāpi nārcanam / pāṣaṇḍino vikarmasthā haitukāś caita eva hi // etadīyā granthā eva ca manvādibhiḥ parihāryatvenoktāḥ / yā vedabāhyāḥ smṛtayo yāś ca kāścit kudṛṣṭayaḥ / sarvās tā niṣphalāḥ proktās tamoniṣṭhā hi tāḥ smṛtāḥ // tasmād dharmaṃ prati trayībāhyam evañjātīyakaṃ prāmāṇyenānapekṣyaṃ syād iti siddham /. “And these are the ones whom no [one should] honour even by mere words, for these are the heretics, the ones who persist in wrong action and the sophists [mentioned by Manu]. And [authoritative legislators] such as Manu have said that their works must be avoided. All the smṛtis that are external to the Veda and those which exhibit wrong views have been declared useless, for they are recorded as resting on darkness. It is therefore established that something like that, external to the triple [Vedic science] as it is, cannot be relied upon as authoritative with regard to dharma.” The first stanza is adapted from MDhŚ 4.30= ViP 3.18.100.

70

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

If we did not declare, without any respect [for them], these [treatises] as nonauthoritative, other [people], thinking that [their non-authoritativeness] is impossible, would share their views. Or, due to the brilliant facility [manifested by them], their [constantly] stating reasons [for what they say] or the kaliyuga [which we are living in, these people] would fall into the error of renouncing, for instance, the violence [done] to animals [as it is] prescribed for the ritual. Or, basing themselves on the [idea] that, like those of Manu and others, [these treatises] are rooted in the śruti, since there is no difference between being composed by a Brahmin and by a kṣatriya, they would even bona fide (sacetaso 'pi) regard [what these treatises teach] as a mere alternative to [that which is] laid down in the śruti. [But] even if one came across a certain contradictory smṛti professed by [authoritative legislators] such as Manu, still this [possibility of following one or the other] only applies within the [Vedic corpus]. We shall [never re]gain the purity of the dharma unless we refute all those [pseudo-paths] 148 that totally contradict the [well-]established path of the triple [Vedic] science.

According to Kumārila and the Purāṇas he alludes to, heretics, especially the Buddhists, are responsible for the degeneration of dharma during the kaliyuga. But as the YP and the messianic myth of Kalki(n) make clear, we should refrain from considering this eschatological degeneration as a purely mechanical, inescapable process. At least in certain places and under proper political circumstances (note Kumārila’s insistence on the king in his discussion of the caste-classes, e.g., at ŚV vanavāda 27–29), the repristination of the threatened dharma (and even the rise of a new kṛtayuga) is possible, and this is likely to be the meaning of Kumārila’s philosophical endeavour to eradicate Buddhist heresy. 1.2.9. Conclusion To be clear, my aim is certainly not to regard Gupta apocalyptic eschatology itself as responsible for the conspicuous heresiological turn of the 148

TV ad MīSū 1.3.4/II.113,1–10: yadi […] anādareṇaiṣāṃ na kalpyetāpramāṇatā / aśakyaiveti matvānye bhaveyuḥ samadṛṣṭayaḥ // śobhasaukaryahetūktikalikālavaśena vā / yajñoktapaśuhiṃsādityāgabhrāntim avāpnuyuḥ // brāhmaṇakṣatriyapraṇītatvāviśeṣeṇa vā mānavādivad eva śrutimūlatvam āśritya sacetaso ’pi śrutivihitaiḥ saha vikalpam eva pratipadyeran / tena yady api labhyeta smṛtiḥ kācid virodhinī / manvādyuktā tathāpy asminn etad evopayujyate // trayīmārgasya siddhasya ye hy atyantavirodhinaḥ / anirākṛtya tān sarvān dharmaśuddhir na labhyate //.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

71

Mīmāṃsā during the sixth century, and even less as responsible for the striking new directions taken by Indian philosophy from this century on. 149 Rather, my use of apocalyptic prophecies is aimed at showing the growth of a Brahmanical hostility150 that may, at least in part, explain why Brahmanical schools such as Nyāya and Mīmāṃsā turned their attention towards Buddhism,151 and why the Buddhist epistemologists changed their habits and the meaning of Buddhist philosophy radically during the sixth century. Brahmanical orthodoxy and Buddhism had been familiar with each other for many centuries, but in spite of the development of eristicdialectical (vāda) traditions in both of them and the narratives of debates between their representatives, there are very few textual or otherwise documentable hints at there having been a sustained philosophical confrontation before the (end of the) fifth century.152 The change occurring at this time

149

As Kyo Kano has pointed out to me, such a one-sided interpretation of the rise of sixthcentury philosophy would certainly fail to explain intrabrahmanical debate, e.g., between Nyāya and Mīmāṃsā.

150

Note already Verardi 2011: 18: “The Kali Age literature is an unequivocal sign of Brahmanical hostility towards the śramaṇa-s and the social sectors they represented.”

151

Note that, besides countless other fascinating testimonies, Verardi (2011: 130) refers to an interesting passage of the MMK (53.703ab [III.633,9]) to the effect that Samudragupta’s “government [was(/will be)] overspread with wicked Brahmin logicians” (dvijair ākrāntatadrājyaṃ tārkikaiḥ kṛpaṇais tathā /). Concerning the thriving state of Brahmanical schools during the early fifth century, see Faxian’s testimony in Verardi 2011: 132.

152

For an overview of narratives about debates in ancient and early medieval India, see Bronkhorst 2007 (with important epigraphic evidence), Cabezón 2008 (mainly Chinese and Tibetan accounts), and Verardi 2011: 204–214; for an overview of extant eristicdialectical literature, see Kang 2003 and 2006; for a paradigmatic account of debate, royal courts and academies and the rise of eristic-dialectical treatises, see Frauwallner 1984. For tentative accounts of the controversies between Buddhist and non-Buddhist intellectuals during the “Middle Period of Indian Buddhism,” see Eltschinger 2012: 439–479, and below, Chapter 2, §§2.3.1.3–4. This is certainly not to say that direct encounters between Brahmins and Buddhists of different persuasions never took place before the turn of the sixth century. What I mean is that the extant philosophical and eristic-dialectical literature, as well as narratives of debates, provides surprisingly little conclusive evidence to this effect. As far as the narratives are concerned, however, we may look for alternative interpretations. In this connection, Cabezón (2008) provides insightful remarks (but note that Cabezón [2008: 72] considers it “foolish […] simply to lump all this material under the category of ‘legend’ and to summarily dismiss it”). As Cabezón (2008: 90) rightly points out, “[g]iven the importance of debate/argument to

72

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

can hardly be explained purely incrementally in terms of a problemgeschichte. In this case, such an explanation would be just as good as no explanation at all. To cite the transition from a “systematic” period of Indian philosophy, in which all schools, gradually but separately, built their own doctrinal identities, to a “post-systematic” period, in which these schools engaged in disputation, still requires an explanation as to why this transition occurred and why all these schools were simultaneously ready and eager to debate with one another. The factors responsible for this sudden outburst of philosophical confrontation cannot be seriously looked for within the competing traditions themselves, since here the reasons are most likely to be of a non-philosophical and sociohistorical character. But by looking closer at the evolution of the Brahmanical apocalyptic eschatology, I hope I have been able to uncover one part of the ideological background against which these philosophical shifts and many other things make sense. What we need is a general explanation as to why so many innovations in Buddhist India took place slightly before or after 500 CE: the foundation of these mahāvihāras or vihāramaṇḍalas “mimicking feudally grounded fortresses,”153 the nearly contemporaneous rise of Buddhist Tantrism and epistemology, the strong decline of Abhidharmic creativity and controversy,154 etc. Needless to say, the present chapter does not provide an explanation of this type, but it highlights one sine qua non factor of this explanation, viz. a dramatic increase of hostility towards Buddhism as the Gupta dynasty starts to crumble.

Indian and Tibetan scholastic identity, it is not surprising to see the narratives of the great debates as central features of the lives of these individuals [Nāgārjuna, Āryadeva, Asaṅga, Vasubandhu, Dignāga, Dharmakīrti, etc.].” In other words, these stories, or what Cabezón (ibid.) aptly calls “argumentational lives,” have a “powerful force in the construction of scholastic communal identity” (Cabezón 2008: 91). Whatever their value as witnesses of actual, historical debates, “these stories create for elite Chinese and Tibetan monastic audiences—and to the extent that they were built on Indian fragments, perhaps for late Indian monastic audiences as well—a sense of identity through varied forms of affiliation to a relatively fixed group of eminent Indian monks who together form a kind of ‘fleshly canon,’ if you will” (Cabezón 2008: 90). A similar “fleshly canon” is recorded in the very interesting Jaina inscription (twelfth century, near Sravana Belgola) discussed by Bronkhorst (2007: 269–272). 153

Davidson 2002: 167.

154

On the latter two aspects, see below, Chapter 2, especially §2.3.3 concerning the decline of Abhidharmic creativity.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

73

1.3. INDIAN BUDDHIST APOCALYPTICISM 1.3.1. Introduction In spite of a few recent groundbreaking studies, research on Indian Buddhist apocalypticism is still in its infancy. Improving on the work of Przyluski and Lamotte, Nattier has provided a detailed picture of the Buddhist “timetables of decline” through her analysis of the “Prophecy of Kauśāmbī” embedded in the CGSū and other sources. More recently, Boucher’s in-depth investigation of the RPSū has drawn attention to the apocalyptic dimensions of the early Mahāyāna’s vitriolic criticism of mainstream Buddhism in its concessions to secular society. However, assuming that apocalyptic rhetoric was nearly always associated with present-day events and/or situations of a social, political, economic or institutional nature, much research remains to be carried out regarding the historical background of at least some of the most significant prophecies of decline. There is indeed evidence to suggest that major political breaks in Indian history—the rise of Kuṣāṇa rule, the Gupta empire and its gradual collapse, the alleged Hūṇa terror and the rise of Islamic domination—brought about an increase (if not always a complete reassessment) of Indian Buddhist apocalypticism. Whereas Indic sources are scarce in number and most often are difficult to interpret, the Chinese Buddhist canon and Dunhuang materials seem to provide valuable information in this regard. However, as Lévi and especially Zürcher have pointed out, most of the relevant materials are of an apocryphal character, or at least can be shown to have been reworked and/or interpolated in early medieval China (albeit through the agency of Indian translators like Narendrayaśas) under purely Chinese religio-political circumstances. One should, then, be extremely cautious with such materials when tackling Indian apocalypticism and its historical background. 1.3.2. The Timetables and the Causes of Decline 1.3.2.1. Two things are striking about the Indian Buddhist accounts of time and history with regard to the duration of Buddhism itself. First, the surprisingly short life-span ascribed to Śākyamuni’s dispensation. As already pointed out by Nattier, the timetables of decline prophesising the demise of

74

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

the Good Law (saddharmavipralopa) range from 500 to 5,000 years after the Buddha’s parinirvāṇa,155 and seem to have been gradually updated according to the Buddhists’ awareness of the time elapsed since the passing away of their founder. Second, the strong emphasis laid on purely internal causes of decline. According to Nattier, the Pali canon points to seven causes connected with the “laxity within the saṅgha” (Nattier 1991: 120): 156

1. the admission of women into the monastic community; 2. lack of respect toward various elements of the Buddhist tradition; 3. lack of diligence in meditation practice; 4. carelessness in the transmission of the teachings; 5. the emergence of divisions within the saṅgha; 6. the emergence of a false or “counterfeit” 157 158 Dharma; and 7. excessive association with secular society.

155

These figures may or may not be further subdivided into distinct phases that account for the gradual decline (see Nattier 1991: 27–64). Among the many interesting results of Nattier’s groundbreaking study, one may note that, contrary to what the standard rendering (“counterfeit doctrine”) might suggest, Skt. saddharmapratirūpaka hardly ever seems to hint at a second, degenerate period of the Law (SN II.224,5–6 being the only exception; see Nattier 1991: 87). The Indian Buddhist sources also do not seem to provide any textual foundation for the East Asian third period of Buddhism, the socalled mofa/mappo (ᮎἲ). According to Nattier, the most likely candidate as an equivalent of mofa, Skt. paścimakāla, does not mean “the final period,” but “the latter period.” This point has been challenged by Boucher (2008: 66) who contends that, at least in the context of the RPSū, “final” provides a more satisfactory rendering than “latter.” Note, however, that Boucher does not contest Nattier’s results regarding the complete lack of an Indian Buddhist basis for the triple division of dharma into [saddharma/]zhengfa/shobo (ṇἲ), [(sad)dharmapratirūpaka/]xiangfa/zobo (ീἲ) and [paścimakāla/]mofa/mappo. For a critique of Nattier’s view of mofa as a Chinese “apocryphal word,” see Hubbard 1993: 141–142.

156

As Nattier convincingly shows, the famous story according to which the admission of women into the regular community was responsible for the shortening of the dharma’s duration can be found only in Mahīśāsaka, Dharmaguptaka, Theravādin, Haimavata and Sarvāstivādin sources, i.e., in sources originally belonging to the Sthavira line. In other words, “[t]he date of emergence of this tradition should […] be placed during the period from 340–200 BCE” (Nattier 1991: 32–33). On the Mahāvibhāṣā’s exegetical accommodation of this tradition implying a 500-year timetable in contrast to the Sarvāstivādin 1,000-year timetable, see Nattier 1991: 43–44 and Lamotte 1976: 212–213. For a summary of the story, see Chappell 1980: 124–125.

157

See above, p. 74, n. 155.

158

Nattier 1991: 120. See also Chappell’s overview of the causes of decline in Chappell 1980: 123–130.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

75

Interestingly enough, the Pali canon shows no awareness of any external cause of the decline and/or demise of the Good Law. Like the sources to be considered below, the Pali canon does not hold outsiders (tīrthika) as instrumental in the final collapse of Buddhism. Among these internal causes, factors 5 and 7 recur with considerable frequency in Mahāyānist as well as non-Mahāyānist sources. In the ThaG, for instance, the elder Pārāpariya complains that in earlier times, monks were not as eager for the necessities of life, for medicines and supplies, as they were for the annihilation of the outflows (āsava). In the forest, at the foot of trees, in caves and grottos, devoting themselves to seclusion, they lived making [the annihilation of the outflows] their aim […] Now, these elders […] have all passed 159 away.

In his Manorathapūraṇī, Buddhaghosa declares that monks “will enter into secular occupations and ordinary family life.”160 The shorter Parinirvāṇasūtra,161 which divides its thousand-year timetable into ten hundred-year periods, states that [d]uring the first several periods believers are still devoted to the practice and preservation of the Dharma, but as time goes on they become more preoccupied with the externals of the Buddhist tradition rather than its essence. A turn for the worse begins during the seventh and eighth centuries, and during the ninth and tenth periods Buddhist monks turn their attention almost exclusively to worldly activi162 ties.

As for the Tibetan version of the CGSū,163 it proclaims that

159

ThaG vv. 924–925 and 928 in Nattier’s revised translation (1991: 124). According to Nattier (1991: 125), Pārāpariya condemns “the excessive involvement of monks with secular society;” in the following stanzas, he also vituperates “the variety of tactics used by monks to acquire worldly goods and fame” (vv. 936–944) .

160

See Nattier 1991: 57. Timetable: 5,000 years divided into 5 x 1,000 years (the basis of the Theravāda doctrine). Note that this statement concerns the end of the fourth 1,000year period.

161

See Nattier 1991: 43, n. 40 and Lamotte 1976: 213.

162

Nattier 1991: 45.

163

For a critical edition and English translation of the Tibetan text, see Nattier 1991: 228– 277. Timetable: 2,000 years divided into 4 x 500 (with the last 500-year period divided into two further periods).

76

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

during the second sub-period even monks will not practice the Dharma in accordance with the Buddha’s teachings. They will engage in worldly pursuits, occupy164 ing themselves in a variety of secular trades.

The issue of quarrels and dissension within the Buddhist saṅgha is no less recurrent. According to the AN, then [monks] revile one another, accuse one another, quarrel with one another, and repudiate one another; those of no faith do not find faith there and the faithful 165 fall away.

The Prophecy of Kātyāyana166 predicts that “there will be anger and fighting.”167 The Chinese version of the CGSū168 foresees that “in the latter five hundred years, quarrelling and debate will be well-established.”169 1.3.2.2. The emphasis on internal divisions finds its paradigmatic expression in a story that would eventually transcend sectarian affiliations and geographical/cultural borders, the “Prophecy of Kauśāmbī.” According to Nattier, the “Kauśāmbī decline prophecy appears to have emerged in northwest India sometime between the second century CE and the middle of the third,”170 under Kuṣāṇa rule. The archetype of the story171 as reconstructed by Nattier can be summarized as follows: Three foreign kings (the Greeks, Sakas, and Parthians) are named as the cause of 172 disruption within the Indian Buddhist community. A pious Buddhist king 164

Nattier 1991: 53.

165

AN III.180 in Nattier’s revised translation (1991: 123). According to Nattier (1991: 123), “[m]ore specifically, the issue in this passage seems to be the hostility and argumentation that can emerge in sectarian rivalries.” On early Buddhist attitudes towards debate and argumentation, see Collins 1982: 127–131 and 139–141, and Eltschinger 2012: 432–439.

166

T. 2029. Timetable: 1,000 years divided into 3 x 300 and 1 x 100 years.

167

T. 2029, 12b26–c1 in Nattier’s translation (1991: 45).

168

T. 397[15]. Timetable: 2,000 years divided into 4 x 500 years.

169

T. 397[15], 370b8–11 in Nattier’s translation (1991: 52).

170

Nattier 1991: 226. On the Kauśāmbī story, see also Chappell 1980: 128–129.

171

This archetype probably consisted only of verses. Of uncertain sectarian affinity, it may have been recorded in a northwest Prākrit in the Kharoṣṭhī script.

172

In later recensions of the story, these Yavana, Śaka and Pahlava kings are reported to have invaded northwest India and to have ruined Buddhist stūpas and temples.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

77

(unnamed), ruling at Kauśāmbī, invites all the monks of Jambudvīpa to a mahā173 dāna feast at the suggestion of his advisor, the monk Śiṣyaka. A dispute then breaks out between an Arhat, Sūrata, who insists on strict observance of all 174 monastic rules, and a group of more complacent monks; Sūrata is killed by a 175 disciple of Śiṣyaka, who in turn is killed by a yakṣa.

This prophecy of decline might be the earliest among those ascribing external causes to the demise of the dharma. But this story reveals more than that. As Nattier points out, it is not the invasions themselves that are portrayed as bringing an end to the Dharma […] but the munificence of a well intentioned Buddhist king. Moreover, it is the Buddhist community itself […] whose actions result in the death of an 176 Arhat, and ultimately in the disappearance of the Dharma from the world.

According to her, this story “makes good sense as the product of a Kushan environment,” in a time where “the Buddhist subjects of […] a cosmopolitan realm [were] presumably enjoying all the material and spiritual benefits afforded by the long-lasting pax kushanica.”177 1.3.3. Monastic Degeneration and Early Mahāyāna If Nattier’s hypothesis is correct, one might extrapolate the following explanation: the “Prophecy of Kauśāmbī” perhaps mirrors the worries aroused in a disciplinary conservative fraction of northwestern Buddhist 173

In later recensions of the story, the king is named Mahendrasena and is said to have crushed the three foreign invaders; the invitation to the feast (sometimes a pañcavarṣa feast) is sometimes accounted for as expressing the king’s feelings of guilt with regard to the death of so many people in war.

174

In various recensions of the story, the tripiṭaka master Śiṣyaka himself decides to abridge the prātimokṣasūtra recitation on the grounds that the precepts are no longer respected by anybody. Sūrata protests this, arguing that the full text should be recited, which arouses the wrath of Sūrata’s disciples.

175

Nattier 1991: 215 and 217. In later recensions of the story, this murder is followed by a struggle involving all the monks gathered in Kauśāmbī (i.e., all the monks of Jambudvīpa). This results in the death of all of them, which coincides with the final demise of the dharma. At any rate, the story ends with the murder of both the last tripiṭaka master and the last arhat.

176

Nattier 1991: 227.

177

Nattier 1991: 227.

78

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

communities by the saṅgha’s laxity under Kuṣāṇa rule. Indeed, the Kuṣāṇa period coincides with major institutional and economic breaks within Buddhist communities (among them, the “emergence in India of a new type of social institution with considerable clout: the fully institutionalized, permanently housed, landed monastery”178). Put in other words, the early centuries CE lay the foundations of what Gregory Schopen has termed “Mainstream Buddhism,” and inaugurate the “Middle period” of Indian Buddhism. But if the “Prophecy of Kauśāmbī” in statu nascendi still gives voice to traditionalist claims within “nikāyic” communities (Dharmaguptaka, Sarvāstivādin?), it certainly did not take long for this new form of institutional Buddhism to arouse the criticism of yet another segment of its members, those who were eventually responsible for at least one of the trends of early Mahāyāna. Daniel Boucher provides the following summary of recent research and reflections on this issue: What Gregory Schopen has termed the Middle Period of Indian Buddhism, roughly the first half of the first millenium, can be characterized by a highly organized, sedentary Buddhism with a complex administration governed by an equally complex legal system. Monks living in these monasteries were bound in a tangled web of relationships to lay donors and their fellow monks, relationships that required the constant negotiation of property rights and ritual obligations. The monastic disciplinary codes give every indication that the monks governed by them were fully entrenched in the socioeconomic milieu of contemporary society and were “preoccupied—if not obsessed—with avoiding any hint to social criticism 179 and with maintaining the status quo at almost every cost.”

As scholars such as Schopen, Silk, Nattier and Boucher have strongly argued in recent years, it seems nearly certain that at least some of the promoters of early Mahāyāna reacted against Mainstream Buddhism by clamouring for a return to wilderness dwelling and the strict observance of the purificatory ascetic qualities (dhutaguṇa). And this gave rise to new apocalyptic voices. Boucher’s work on the RPSū provides a unique opportunity to take a closer look into the development of this locus classicus of the Mahāyānist criticism of the Mainstream’s deep interaction with secular society and upper-class donors. According to Boucher, the authors of the RPSū “reject accommodation with the socioeconomic milieu by insisting 178

Schopen 1994: 553(/81).

179

Boucher 2008: 67, quoting Schopen 1995: 478(/96).

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

79

on a rigorist interpretation of monastic discipline.”180 The RPSū is a “sectarian product at the level of intramonastic tension and fractionalization”181 where the status quo relationship between monks and laity, a relationship founded on the exchange of material and symbolic commodities […] had become disordered. 182

As Finot already noticed,183 the RPSū’s depiction of contemporary Buddhism is perhaps unique in its severity. Its authors bemoan fellow monastics for behaving like householders, roaming “among upper-class patrons in towns” and performing “services for people in the household.”184 They have wives, sons and daughters, delight in visiting relatives, and think about them day and night. They give themselves up to agriculture, keep cows, horses and asses. Always preoccupied with profit, they practise trades, have stores and many goods, and mix up what belongs to the stūpa, to the saṅgha and to themselves. Engaged in the affairs of the monastery, they have male and female slaves and lots of attendants. As a predictable corollary to their intimate association with secular society, they have given up meditation, study and the virtues of the Buddha; deficient in the virtues of a monk, they are nothing but hypocrites and deceivers. These monks have no regard for the rules of training or for the prātimokṣa or for Vinaya; given to immorality and devoid of prescriptions, they are undisciplined, arrogant, and consumed by jealousy and pride. Last but not least, they are surrounded by a coterie of students who venerate them, but who themselves are also undisciplined. The description given here is far from exhausting the depiction in the RPSū, and provides a lively Buddhist counterpart to the Epic and Purāṇic accounts of the kaliyuga/yugānta. No less interesting is the historical stratigraphy of the text. The RPSū was first translated into Chinese in 270 by the Yuezhi monk Dharmarakṣa, 180

Boucher 2008: 78.

181

Boucher 2008: 80.

182

Boucher 2008: 83.

183

See Finot 1901: ix–xi. On p. ix: “La prophétie est en réalité un tableau satirique des mœurs relâchées du clergé bouddhique. La vivacité et la précision de cette peinture, qui reflète sans doute des faits réels, en font un intéressant document d’histoire religieuse.”

184

The following examples are borrowed from Boucher’s translation (2008: 137–141) of RPSū vv. 177–216 (see also Finot’s French translation in Finot 1901: ix–xi). For similar Chinese (apocryphal) statements, see Zürcher 1982: 17–18.

80

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

and a second time around 585–592 by the Gandharan monk Jñānagupta.185 As Boucher observes, [t]he crucial period, then, for understanding significant developments in the history of the text can be placed roughly between 270 and 550 CE, corresponding ap186 proximately to the north India Gupta period.

Interestingly enough, although the overall charge is basically the same in the two earliest Chinese versions, the text as translated by Dharmarakṣa seems to contain no hint of apocalyptic rhetoric. The references to the pitiable behaviour of contemporary monks as a sign of the imminent demise of the Good Law only appears in the broadly expanded text translated by Jñānagupta. These Gupta or slightly post-Gupta interpolations consistently link apocalypticism with the disruptive forces undermining the saṅgha: Hostile to one another, they are constantly engaged in seeking each other’s faults […] Monks, who are unrestrained during the final period [of the Dharma] are far removed from morality and virtue. They cause this Dharma to disappear as a 187 result of their contentiousness, strife, and envy.

Or, a few pages later: Always honoured but devoid of virtue, [such monks] will be divisive, treacherous, and fond of quarrelling. They are venerated as the Teacher by the people, but they will be consumed by pride and conceit […] The destruction of the Dharma occurs during the very dreadful final period. And such undisciplined monks as these will 188 cause the ruin of this teaching of mine.

185

Respectively, T. 170 and T. 310[18]. There exists a third Chinese translation of the RPSū, executed in 994 by the Song translator Dānapāla (T. 321). See Boucher 2008: xvii–xix.

186

Boucher 2008: 109.

187

RPSū vv. 84ab and 85 (Finot 1901: 17,12 and 16–17): te ca parasparam eva ca dviṣṭā chidragaveṣaṇanityaprayuktāḥ / […] evam asaṃyata paścimakāle bhikṣava śīlaguṇeṣu sudūre / te ’ntarahāpayiṣyant’ ima dharmaṃ& bhaṇḍanavigrahaīrṣyāvaśena //. &’ntarahāpayiṣyant’ ima dharmaṃ em. (Alexis Sanderson, May 2010): ’ntara hāpayiṣyanti madharmaṃ Ed. Translation Boucher 2008: 127.

188

RPSū vv. 204 and 207 (Finot 1901: 31,19–20 and 32,5–6): sada satkṛtā guṇavihīnā bhedasūcakāḥ kalahakāmāḥ / te śāstṛsammata janasya te ca bhaviṣyanti mānamadadagdhāḥ // […] etādṛśāś& carimakāle dharmavilopa vartati sughore' / ta cāpi bhikṣava adāntā nāśayitāra śāsanaṃ mamedam //. &etādṛśāś em. (Boucher, MS, RPSūT): etādṛśaś

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

81

1.3.4. The Buddhist Appropriation of the kaliyuga 1.3.4.1. There is at least some evidence to the effect that the chaotic final decades of the Gupta rule as well as the early post-Gupta period not only witnessed an increase in apocalyptic rhetoric, but also recorded significant shifts in strategy, contents and emphasis. In this regard, the tenth chapter (sagāthaka) of the LASū provides a very interesting testimony. Whereas its chapters 2–8 had already been translated into Chinese in 443,189 chapters 1, 9 and 10 are only recorded in Bodhiruci’s 513 translation.190 The late composition of the relevant passages can hardly be questioned since they refer explicitly to the Guptas. Most interesting is the fact that this text resorts exclusively to the Hindu cosmological and apocalyptic system of the four ages in order to account for the imminent demise of the dharma. In verse 804, the former buddha Viraja, who claims to belong to the “golden age” or kṛtayuga, speaks as follows: Neither in the dvāpara[yuga], nor in the tretā[yuga], nor afterwards in the kaliyuga do [buddhas,] the protectors of the world, arise. They [only] get awak191 ened in the kṛtayuga.

With no fear of contradicting himself, Viraja proclaims, in v. 794: [There are four ages:] the kṛtayuga, the tretā[yuga], the dvāpara[yuga] as well as the kali[yuga]. Myself and others [arose] in the kṛtayuga, [whereas] the Lion of 192 the Śākyas [will arise] in the kaliyuga.

No less contradictory but even more interesting is, however, the properly apocalyptic kernel of the passage: Once I [shall] have passed away, afterwards, there will be [teachers] such as Vyāsa, Kaṇāda, Ṛṣabha, Kapila and the Guide of the Śākyas. Once I have passed away, within one hundred years there will be Vyāsa as well as [his] Bhārata, the

Ed. 'sughore em. (Boucher, MS): sughoraḥ Ed. Translation (slightly modified) Boucher 2008: 140. 189

T. 670.

190

T. 671.

191

LASū 10.804: na dvāpare na tretāyāṃ na paścāc ca kalau yuge / sambhavo lokanāthānāṃ sambudhyante kṛte yuge //.

192

LASū 10.794: kṛtayugaś ca tretā ca dvāparaṃ kalinas tathā / ahaṃ cānye kṛtayuge śākyasiṃhaḥ kalau yuge //. On kalina, see BHSD 172b s.v.

82

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS 193

Pāṇḍavas, the Kauravas, Rāma and, afterwards, Kṛṣṇa. [There will be] the Mauryas, the Nandas and the Guptas, and then, the mlecchas [will be] the vilest among rulers. At the end of the mlecchas, [there will be] an armed conflagration (śastrasaṅkṣobha), and at the end of the warfare (śastrānte), the kaliyuga [will open up]. And at the end of the kaliyuga, the Good Law will no longer be culti194 vated by people.

Although the sequence hinted at is in itself perfectly clear, it is of course difficult to assess the identity of the mlecchas (Hūṇas?) as well as the exact circumstances presiding over the saddharmavipralopa. One may tentatively interpret the sequence linking mleccha invasion–fight–demise of the dharma along the lines of the “Prophecy of Kauśāmbī,” in which the sequence of events is the following: foreign invasion–victory over invaders– demise of the dharma due to the agency of strictly internal disruptive forces.195 1.3.4.2. But the LASū is not the only late fifth-century to sixth-century Sūtra to reflect a Buddhist appropriation of the Brahmanical yugas.196 In an 193

According to BHSD 441b s.v. maurin, the MS reads here saurī and not maurī, “and this or śauriḥ is surely to be read (Skt. Śauriḥ = Kṛṣṇaḥ).”

194

LASū 10.784–786: vyāsaḥ kaṇāda ṛṣabhaḥ kapilaḥ śākyanāyakaḥ / nirvṛte mama paścāt tu bhaviṣyanty evamādayaḥ // mayi nirvṛte varṣaśate vyāso vai bhāratas tathā / pāṇḍavāḥ kauravā rāmaḥ paścāc chaurī bhaviṣyati // mauryā nandāś ca guptāś ca tato mlecchā nṛpādhamāḥ / mlecchānte śastrasaṅkṣobhaḥ śastrānte ca kalir yugaḥ / kaliyugānte lokaiś ca saddharmo hi na bhāvitaḥ //. Let it be noted that the events following the disappearance of the Good Law, and more specifically what is likely to be interpreted as the final cataclysm and rise of a new kṛtayuga, betray a strong alignment with non-Buddhist standards (LASū 10.787cd–789ab: vahnyādityasamāyogāt kāmadhātur vidīryate // punaḥ saṃsthāsyate divyaṃ tasmi l lokaḥ pravartsyate / cāturvarṇā nṛpendrāś ca ṛṣayo dharmam eva ca // vedāś ca yajñaṃ dānaṃ ca dharmasthā vartsyate punaḥ /): “Due to the conjunction of fire and sun, the realm of desire is [then] torn to pieces. [But] heaven will come again, and within it the world proceeds [again]. The four caste-classes, the kings, the ṛṣis and the dharma, the Vedas, the sacrifice and giving will re-arise conformable to dharma.”

195

The same pattern can be seen in a prophecy translated and most probably adapted by Narendrayaśas. See below, §1.3.5.

196

The kaliyuga is indeed the subject matter of a thirteen-verse work, the KP, ascribed to the famous second-century Buddhist scholar Mātṛceṭa. According to Dietz (2000: 173), “[d]ie Authentizität der Autorschaft des Mātṛceṭa läßt sich nicht mit Sicherheit nachweisen,” and no parallel to the Stotras of Mātṛceṭa can be identified. In my opinion, this short text is strongly indebted to Epic and Purāṇic accounts of the kaliyuga due to its

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

83

extremely suggestive passage,197 the KVSū, a Mahāyānist text dealing with Avalokiteśvara’s miraculous endeavours for the sake of the living beings, strong emphasis on the cosmological, social, political and ethical conditions that are supposed to characterize the kaliyuga. Cosmic and natural disorder is best represented (vv. 3, 4b, 6d, 7cd, 8cd, 9ac, 10cd), at times even caused by human immorality and disregard for dharma (in its Brahmanical version!). Second in number are the references to the humans’ vices and depravities, ignorance and immorality (vv. 1, 5cd, 8b, 11b and d, 12b, 13a). Issues recurring in the Brahmanical sources include women’s misbehaviour (vv. 9d, 11a), plagues and illnesses (vv. 4a, 8a), the injustice of kings (vv. 7a, 10a, 10d), warfare and other threats (vv. 5b, 6c, 7ab), breaks in social and family order (vv. 10a, 11a), the importance of money and wealth (vv. 2b, 5a), and misplaced veneration (vv. 2a, 5ab, 6d, 9d). While these purely secular concerns represent more than ninety per cent of the whole, allusions to Buddhist motifs (bhadrakalpa vv. 4c, 4d, 11bc, 12; Māra v. 13) and to the decline of the Buddhist Law are remarkably few in number: (1) “Die Betrüger werden die Wahre Lehre (saddharma) zugrunde richten” (v. 5d: g.yon can rnams kyis dam chos bsnub par 'gyur; translation Dietz 2000: 180); (2) “Nur wenig sind es, die in Zucht lebend sich der höchsten Lehre zuwenden. Zweifellos wird die Lehre der Seher nach kurzer Zeit verfallen” (v. 6ab: gaṅ źig sdom pa la gnas bstan pa'i mchog la 'jug par byed pa ñuṅ / the tshom yod min draṅ sroṅ bstan pa riṅ por mi thogs nub par 'gyur /; translation Dietz 2000: 180–181); (3) “Zusammen mit dem Dharma fliehen die Parivrājakas und Asketen” (v. 10c: chos daṅ rab byuṅ dka' thub rnams 'bros; translation Dietz 2000: 183). I am inclined to consider the KP as a rather late (?) work modelled on Brahmanical prophecies and providing a very weakly “buddhicized” account of the kaliyuga, one, at any rate, that lacks an equivalent in Buddhist literature. The earliest Buddhist references to the kaliyuga I am aware of belong to the YBh. While dealing with the doctrine of ritual violence (hiṃsādharmavāda), the YBh (145,20–146,4) claims that Brahmins are supposedly allowed to kill animals and eat their flesh when the kaliyuga is at hand (pratyupasthite kaliyuge). As for YBh 155,11, it ascribes the claim of social superiority (agravāda) to the so-called kaliyugikā brāhmaṇāḥ. On these two passages, see Eltschinger forthcoming b, §2.2.8 and 2.2.14. Worth noticing is also Vasubandhu’s reference to the kaliyuga in AKBh 266,25–267,2 (ad AK 4.110d), though it does not occur in an apocalyptic context: śākyamunir nāma samyaksambuddhaḥ pūrvaṃ babhūva / yatra bhagavatā bodhisattvabhūtenādyaṃ praṇidhānaṃ kṛtam / evamprakāra evāhaṃ buddho bhaveyam iti no 'py evaṃ kaliyuga evotpannavān āryavat tasyāpy evaṃ varṣasāhasrāntaṃ śāsanaṃ babhūva /. “Formerly there was a perfectly awakened [buddha also] named Śākyamuni under whom [our] Blessed One, [still] in the state of a bodhisattva, made his initial vow[, saying]: ̒May I become a buddha of exactly this kind!’ Just like ours, he also arose during a kaliyuga, [and] like the Noble One his teaching also lasted one thousand years.” According to AKVy 432,7–8 the bodhisattva was named Prabhāsa and was the son of a potter (kumbhakārakumāra). On this earlier Śākyamuni, see Kośa III.228. 197

I owe this reference to González-Reimann (2002: 171–172). On the KVSū in general, see Studholme 2002, and below, Chapter 2, §§2.2.5 and 2.2.7.2–3.

84

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

Hindu deities, most importantly Śiva Maheśvara, are reported to originate from Avalokiteśvara. To Śiva, the great bodhisattva delivers the following prophecy: O Maheśvara, you will be [active] when the kaliyuga arrives. Born as the foremost of the gods in the realm of suffering beings, you will be called the creator and the agent [of the world]. All beings who hold the following discourse to(/among) ordinary people will be deprived of the path to enlightenment: “It is said that space is [Maheśvara’s] liṅga, [and that] the earth is [his] pedestal; it is the receptacle of all beings, [and it is] because [they] merge(/fuse) [into it that it] is called 198 liṅga.”

This passage somehow echoes the later Buddhist Tantric apologetic strategies, according to which even the non-Buddhist, Śaiva esoteric procedures (which readers or practitioners of the MVASū or the MMK might suspect as hiding behind these Tantras) were actually revealed by the Buddha.199 Most telling is our passage’s explicit association of Śiva and Śaivism to the kaliyuga. Indeed, as recent and ongoing research by Alexis Sanderson shows, Śaivism appears to be the most successful Indic religion from at least the sixth century onwards, with its increasing appeal to royal patrons by extending and adapting its repertoire to contain a body of rituals and theory that legitimated, empowered, or promoted key elements of 198

KVSū 265,4–8: bhaviṣyasi tvaṃ maheśvara& kaliyuge pratipanne / kaṣṭasattvadhātusamutpanna ādideva ākhyāyase sraṣṭāraṃ kartāram / te sarva'sattvā bodhimārgeṇa viprahīṇā bhaviṣyanti ya īdṛśaṃ pṛthag(janeṣu sattveṣu) sāṅkathyaṃ kurvanti // ākāśaṃ liṅgam ity āhuḥ pṛthivī tasya pīṭhikā / ālayaḥ sarvabhūtānāṃ līyanāl* liṅgam ucyate //. & maheśvara em. (KVSūT P231b3): maheśvaraḥ Ed. 'sarva° with no equivalent in KVSūT P231b4. (īdṛśaṃ pṛthag° em.: īdṛśapṛthag° Ed. (īdṛśa with no equivalent in KVSūT P231b4). )sattveṣu with no equivalent in KVSūT P231b4–5 (so so'i skye bo'i naṅ du). *līyanāl em. (Regamey 1971: 431): līlayā Ed. Translation (slightly modified) González-Reimann 2002: 172. On this passage, see Studholme 2002: 30–31, 44–45 and 123–124; on the verse quoted here (and especially līyana), see Studholme 2002: 19–20 and 28–29, Regamey 1971. This statement, which occurs at the beginning of the KVSū, is not represented in the Gilgit manuscripts (G1 lacks folios 1–x). However, considering that the only known significant divergence between the Nepali and the Gilgit versions concerns a very neatly delineated section (Sarvanīvaraṇaviṣkambhin’s quest for the ṣaḍakṣarī vidyā), I see no compelling reason to doubt the presence of this passage in the textual tradition reflected in the Gilgit manuscripts. On this passage, see also below, Chapter 2, §2.2.7.3.

199

See Sanderson 2009: 128–132, and below, Chapter 2, §2.2.7.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

85

the social, political and economic process that characterizes the early medieval 200 period.

To put it in other words, from this period Śaivism is the most dangerous religio-political challenge to Buddhism. Buddhism does not respond to this challenge on a philosophical level (at least not until its tenth century confrontation with Kashmir Śaivism), but by developing an esoteric synthesis of its own (if one may describe it this way!). As the KVSū makes clear, contact with Śaivas is as detrimental to a Buddhist’s soteriological expectations as the contact with pāṣaṇḍins is to a pious Brahmin’s purity and religious merit. 1.3.5. Narendrayaśas Worth mentioning in connection with late Gupta prophecies of decline is the Buddhist monk and translator Narendrayaśas.201 Born in Uḍḍiyāna in 490 or 516, Narendrayaśas has often been credited with the transmission of Indian Buddhist prophetic/apocalyptic literature from India (where he travelled for many years) to Central Asia and China.202 And indeed, Narendrayaśas did not only provide new Chinese translations of scriptures such as the CGSū, but is also reported to have translated works such as the Dehu zhangzhe jing (ᚫ嫊㛗⪅⥂)203 and the Lianhua mian jing (ⶈ⳹㠃⥂).204 Narendrayaśas has long been noted for his “tendance à natu200

Sanderson 2009: 253. See, more generally, Sanderson’s extremely detailed account in Sanderson 2009: 252–303, and below, Chapter 2, §2.2.3, for a summary. Note also Davidson 2002: 86: “Overall, Buddhist institutions could not effectively compete for patronage from militaristic princes, who increasingly found that they were best represented by Śaiva values and rhetoric. Śaiva systems made allowance for forms of behavior that Buddhist syntheses could not support, since even the most syncretic Buddhist systems were not as open to negotiation about issues of violence, power, and self-aggrandizement as were the medieval Śaiva representatives.” For a summary of Davidson’s position, see below, Chapter 2, §2.2.2.

201

On Narendrayaśas, see Yamada 1955, Chappell 1980: 145–147, Zürcher 1982: 22–33, Kuwayama 1989: 97–111 and Wang-Toutain 1994: 75–78. Parts of Yamada’s hypothesis have been challenged by Nattier (1991: 112–117).

202

See Lévi 1905: 292–305, Kuwayama 1989: 91, Wang-Toutain 1994: 70, Hubbard 2001: 62, n. 24.

203

T. 545.

204

T. 386.

86

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

raliser les sūtra dans le monde hindou-chinois.”205 And since Zürcher’s more recent studies, it has become increasingly clear that the zealous translator was inclined to add very significant interpolations when it came to legitimizing Central Asia and especially China as the new repositories of the Good Law, thus engaging in “religious propaganda.”206 These interpolations are best understood against the background of Buddho-Taoist eschatology, a “typically hybrid complex of ideas”207 mixing up the ideology of mofa, the imminent advent of a saviour “rescuing the true believers from the impending holocaust”208 (generally Maitreya, transformed from a future buddha to a messiah, but also the bodhisattva Candraprabhakumāra, in this case rather a “great revivalist” or “temporary saviour”209) and “kalpa-disasters” or cosmic crises “heralded by countless signs of clerical corruption, political chaos, and natural disasters.”210 This is especially true of our monk’s translation of the Dehu zhangzhe jing, in which Narendrayaśas borrows from, and expands upon considerably, a prophecy already contained in the anonymous Shenrijing (⏦᪥⥂)211 but nowhere to be found in the earlier translations of Dharmarakṣa and Guṇabhadra. According to Narendrayaśas’ version of this prophecy, Candraprabhakumāra will be reborn as King Daxing (኱⾜) and patronize Buddhism on a grandiose scale, notably by the reproduction and spread of holy texts, the making of Buddha images of every kind, and the establishment 212 of countless sanctuaries in all parts of the empire.

205

Lévi 1905: 292.

206

Zürcher 1981: 48. A few lines below Zürcher adds, concerning the Dehu zhangzhe jing: “Needless to say that such legitimistic masquerades are not more than artificial constructions made for propagandistic purposes.” See also Lévi 1905: 295.

207

Zürcher 1982: 1.

208

Zürcher 1981: 44.

209

Zürcher 1982: 27.

210

Zürcher 1981: 44.

211

T. 535, “probably fifth century” (Zürcher 1981: 46).

212

Zürcher 1982: 26. On the rise and religio-political uses of Candraprabhakumāra (“Prince Moonlight”), see Zürcher 1981 and 1982. The close association between Candraprabhakumāra and the Buddha’s pātra (on the alms bowl, see below) is attested in China as early as 365 CE. As for this (hitherto obscure) bodhisattva’s “special relation to China, [it] is for the first time alluded to around the middle of the fourth century” (Zürcher

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

87

The prophecy is built on the popular but also politically legitimizing story of the Buddha’s travelling alms bowl (pātra) (coming now from Kashgar to China). Faxian (ἲ㢷) was already familiar with this story, which he puts, without the least hint of embarrassment, in the mouth of a Buddhist teacher of Śrī Laṅkā.213 Much more interesting, however, is Narendrayaśas’ translation of the Lianhua mian jing. As already suggested by Lévi, the text is replete with the same kind of Buddho-Taoist hybridization. One part of the prophecy, however, deserves closer attention. In Nattier’s short depiction, the Sūtra records the story of Mihirakula in jātaka-style text in which the Hephthalite ruler is portrayed as a devotee of the Purāṇas […] who vows to destroy the Buddhist religion in a future lifetime. He is reborn as an Indian king, and succeeds in overturning the Buddha-Dharma and in smashing the Buddha’s begging bowl. Karmic justice must be done, however, and Mihirakula is reborn in the Avīci hell in 214 recompense for his actions.

1982: 24). According to Zürcher (1982: 26–27) and Wang-Toutain (1994: 76), the king hinted at is Yang Jian (᳿ሀ), the founder of the Sui (㝳) Dynasty, who adopted the name Wendi (ᩥᖇ) , became king in 581 and emperor in 589. According to Zürcher (1982: 26), “[i]t may well be that Narendrayaśas himself was responsible for inserting this passage into his translation, which he made only one year after he had been summoned by Yang Jian himself to the Daxing shan si […]—the most prestigious Buddhist institution in the empire […]—to perform his translation work under imperial auspices.” The Dehu zhangzhe jing was translated in 584 and the Lianhua mian jing in 585. 213

See Kuwayama’s original translation (1989: 106–107). Note that Zürcher (1982: 31) seems not to call the authenticity of Faxian’s testimony into question, though he describes it as a “curious passage.” This is, in Faxian’s version of the story, the itinerary of the precious relic: Vaiśālī–Yuezhi state–Khotan–Kucha–Han (₎) China–Śrī Laṅkā– mid-India–Tuṣita heaven […]–newly awakened Maitreya. On the historically wellattested relic in Gandhāra (Puruṣapura), see Kuwayama 1989: 105–111 and Wang-Toutain 1994: 63–64. On the alms bowl as a symbol of the saddharma and an instrument of political legitimization in newly converted countries, see Wang-Toutain 1994. As Wang-Toutain (1994: 69) remarks, the appearance of the alms bowl “étant liée à la venue d’un Buddha en ce monde, il était donc logique d’affirmer qu’avec la disparition de la Loi, le bol disparaîtrait lui aussi […] Plus précisément, c’est la disparition du bol qui annoncera le déclin de la Loi.” See also Deeg 1999.

214

Nattier 1991: 113.

88

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

Contrary to Nattier’s interpretation, the Lianhua mian jing is more likely to allude to the rival teacher Pūraṇa Kāśyapa than to the Purāṇas:215 Pūraṇa is often portrayed as closely related to Śrīgupta (by the way, the father of Candraprabhakumāra), who attempted to murder the Buddha and committed suicide as a consequence of the Śrāvastī miracle. Note should also be made that the text does not prophesize the final demise of the saddharma, but its gradual degradation in India in the aftermath of Mihirakula’s “harshly anti-Buddhist activities,”216 and its revival in Serindia and China. The Lianhua mian jing’s account of the ensuing degeneration is worth quoting in full: Après la mort de cette brute d’homme, il y aura sept devaputra successivement qui renonceront à leur corps pour naître dans le royaume de Ki-pin. Et ils consolideront à nouveau la droite loi du Tathāgata et ils feront de grandes offrandes. Ô Ānanda, à cause de la destruction du pātra, mes disciples peu à peu souilleront les défenses pures. Aux premiers temps que mon pātra sera détruit, les bhikṣu, quoique violant les défenses pures, sauront encore assez pour pouvoir, comme le taureau, détruire les hérésies. Mais, les premiers temps une fois passés, les bhikṣu dans le Jambudvīpa détruiront les défenses pures, ils ne se plairont qu’à faire le mal. Eux-mêmes, ils pilleront, voleront, laboureront les champs, cultiveront la terre. Ils convoiteront et entasseront en masse le bétail, les bons vêtements, les bons vases; ils n’aimeront pas à réciter les Sūtra, le Vinaya, l’Abhidharma. Oui, ô Ānanda, les hommes qui aiment à lire, qui aiment à réciter, les hommes qui savent auront tous disparu. Et alors il y aura beaucoup de bhikṣu qui flatteront, qui envieront, qui pratiqueront l’impiété. Et comme les bhikṣu n’agiront plus selon la loi, les rois eux aussi ne se conformeront pas aux lois des rois; et comme les rois ne gouverneront plus selon les lois des rois, le peuple en masse accumulera les dix mauvaises actions (akuśala). Et, par suite de ces mauvaises actions, la terre produira en abondance des plantes épineuses, des plantes vénéneuses, du sable et des 217 cailloux.

As we can see, this prophecy of decline closely conforms to the narrative patterns already encountered above. Like the “Prophecy of Kauśāmbī” and 215

Yamada 1989: 80: “In the future, there will appear […] a disciple of the heretic Fu-lanna [Pūraṇa],” as well as Wang-Toutain 1994: 71: “Après le nirvāṇa de ces cinq hommes apparaîtra un disciple du maître hérétique Pūraṇa.” See also T. 386, 1075b16–c21.

216

Hubbard 2001: 62, n. 24.

217

Translation Lévi 1905: 298–299.

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

89

the LASū, the text “translated” by Narendrayaśas does not make political events directly responsible for the decline/demise of the Good Law, but, rather, for an intramonastic degeneration accounted for in terms of moral and disciplinary laxity and/or internecine quarrels leading in turn to the decline/demise. The present state of research does not allow a decision to be made as to whether Narendrayaśas made use of a pre-existing Indian text/story/Sūtra prophesizing the decline/demise of the dharma after Mihirakula’s allegedly violent anti-Buddhist campaigns,218 adapting it in order to have it fit the Chinese religio-political context. 1.3.6. Conclusion Indian Buddhists seem never to have made external events or persons directly responsible for the demise of the dharma (in spite of their drive to regard external, mainly political causes as launching the process). Their 218

Together with Meiji Yamada (1989), I am strongly inclined not to accept the almost systematic (since Fleet, Stein and Marshall) but uncritical association of the Hūṇas of Indian inscriptions with the Hephthalites of northwestern India. According to the Japanese scholar, the Hūṇas who enter Indian history around 455–456 and were defeated by Bhānugupta (around 512) and Yaśodharman (531–532) came (via the Kāthiāwār, like the Śakas before them) from eastern Iran or southern Afghanistan and ruled over central India (Mālwā, including Eraṇ, Gwālior, Mandasor), i.e., a territory entirely different from the one ruled by the Hephthalites. The issue of Mihirakula’s strong anti-Buddhist leaning is not without problems either. Yamada suspects Xuanzang’s description of Mihirakula’s anti-Buddhist atrocities in Śāgala/Siālkot as mixing up narrative elements borrowed from the legends of the Indo-Greek King Menander and the Śuṅga (and antiBuddhist) ruler Puṣyamitra. Considering that Kalhaṇa’s Rājataraṅgiṇī does not hold Mihirakula’s cruelty to have been specifically directed against the Buddhists, we must conclude that the only descriptions of Mihirakula as a persecutor of Buddhism are Chinese apocryphs strongly indebted to mofa ideology (the Lianhua mian jing and the Fufazangyinyuanchuan [௜ἲ⸝ᅉ⦕ബ, T. 2058], see Yamada 1989: 81–82 as well as Maspero 1911). Moreover, as Kuwayama (1989) has pointed out, Xuanzang never associates this king with the bad condition he witnessed of Buddhist institutions in northwest India. Kuwayama has made a strong case that the Hephthalites cannot be made responsible for this decay, which must, then, have taken place sometime in between the Hephthalites’ disappearance around 550–560 and Xuanzang’s account roughly 80 years later. As Deborah Klimburg-Salter kindly informs me, Giuseppe Tucci drew the hypothesis of a massive natural disaster (an earthquake?) having occurred during the midsixth century, a hypothesis that seems to have been confirmed by the IsIAO (= Istituto Italiano per l’Africa e l’Oriente) team after fifty years of archaeological excavation in Swat.

90

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

apocalyptic rhetoric rather mirrors the warnings of Mahāyānist and nonMahāyānist “rigorist” (Boucher) and “contestant” (Zürcher) sectarian fractions within the Buddhist communities. In this context, the factors most frequently alluded to are internecine quarrels, moral and disciplinary laxity, decline in knowledge and spirituality, and, in early Mahāyāna especially, excessive concessions to secular society and its values. This is true of the Pali canon, the “Prophecy of Kauśāmbī” and early Mahāyāna scriptures such as the RPSū, but also of late Gupta and early post-Gupta sources such as late interpolations and/or additions in the RPSū, the LASū, and some among Narendrayaśas’ “translations.” Note should also be made that, contrary to contemporary Brahmanical rhetoric (Kalkin) and Chinese eschatological beliefs (Maitreya, Candraprabhakumāra), Indian Buddhists seem not to have resorted to messianism in the hope of an at least provisional restoration of Buddhism. As we have seen, Gupta and early post-Gupta Brahmanical apocalypticism exhibits significant shifts in emphasis. Though to a lesser extent, the Buddhist prophecies of decline seem to undergo an analogous evolution. First, at least two important Sūtras testify to the fact that certain Buddhist literati took up the four yugas pattern, and especially the kaliyuga, in order to account for their historical environment. As Nattier rightly points out, Indian Buddhists who, contrary to their East Asian coreligionists, lacked any structured periodization of decline, might have borrowed the ready-made and widespread yuga scheme in order to make up for this lack. But they are no less likely to have taken up a scheme whose gloomily flavoured fourth item, the kaliyuga, strongly appealed to imagination. Interestingly enough, these two occurrences are explicitly associated with well identifiable historical circumstances. Whereas the LASū locates the kaliyuga at the end (brought about through the agency of mlecchas who might be interpreted as the Hūṇas) of a sequence of dynasties (Mauryas, Nandas, Guptas) well disposed, or at least not hostile, towards Buddhism, the KVSū links the kaliyuga to the Śaiva threat. Now, these features recur with only minor changes in one among the prophecies adapted by Narendrayaśas. There, the Hūṇa king Mihirakula is made responsible for the process that eventually leads to the disappearance of Buddhism from India. Interestingly, Narendrayaśas locates the non-Buddhists both upstream and downstream from Mihirakula. The ferocious king owes his anti-Buddhist leaning to having been, in a former life, a disciple of Pūraṇa Kāśyapa, a rival master noted for his attempt to murder the Buddha. In this connection, it would be extremely interesting to inquire whether the fifth and sixth centuries record any significant increase in iconic representations of the

APOCALYPTICISM, HERESY

AND

PHILOSOPHY

91

Buddha’s great “Miracle at Śrāvastī,” for which Pūraṇa Kāśyapa was ultimately responsible. But Narendrayaśas also alludes to the Buddhists’ increasing difficulty in resisting the outsiders’ hostility after the defeat of Mihirakula. Both the KVSū and Narendrayaśas, then, bear witness to the Buddhists’ awareness of one external factor never alluded to before, i.e., the tīrthikas in the form of Śaivas, followers of Pūraṇa Kāśyapa (a metaphor of hostile non-Buddhists in general), and outsiders lato sensu. In a way, this provides an interesting echo of the Brahmanical animosity exhibited in the ViP and the TV. To sum up, late Gupta and early post-Gupta Buddhist apocalypticism seems to reflect, though rather sparsely, a Buddhist awareness of otherwise clearly identifiable threats: a loss of political footing (if not political hostility) and the enmity of non-Buddhist orthodox and sectarian milieux. Does this have any bearing on sixth-century philosophy as we know it from the works that have come down to us? In my opinion, sixth-century Buddhist philosophy entertains no connection whatsoever with political power, and thus this is very unlikely to have been its intended audience. But, as I mentioned earlier, it undergoes the same “heresiological” turn as its Brahmanical counterpart. And although this “heresiological” (as well as apologetic) endeavour is mainly directed towards the philosophical expressions of orthodox Brahmanism (Mīmāṃsā, Nyāya), it is not without a few critical hints at sectarian Hinduism. Dharmakīrti in particular criticizes Śaiva doctrines and practices on at least two occasions.219 In a way, this philosophical shift matches the Buddhist experience as reflected in apocalyptic prophecies. Note, however, that I am aware of no Buddhist philosophical text alluding to apocalyptic beliefs. Contrary to what can be observed in Kumārila’s TV, apocalypticism does not provide philosophy with any explicit ideological or rhetorical background. And other, more classic features of Buddhist apocalypticism might have played no less prominent a role in the development of Buddhist philosophy during the sixth century, at least as far as Dharmakīrti (whose works would, after all, dominate the philosophical landscape until the demise of Buddhism in India) is concerned: internecine quarrels make Buddhism incapable of 219

See Sanderson 2001: 11–13 and below, Chapter 2, §2.2.6. The Mīmāṃsā chapter (= 9) of Bhāviveka’s MHK is replete with allusions to and criticism of sectarian Hindu beliefs and practices. Alexis Sanderson kindly informs me that MHK 9.62 and 9.114–115 paraphrase PSū 5.20, 4.4–5, 5.37–38, 5.24 and 25, and 5.40.

92

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

resisting the outsiders’ attacks on Buddhism. In other words, Buddhism needs a supersectarian identity, a need that might have played its part in the contemporaneous decline in Abhidharmic creativity and polemics and in the concomitant appearance, in doctrinal texts, of such supersectarian selfdenominations as Bauddha, Śākya, and Saugata (all of them meaning: Buddhist). There is, however, a more satisfactory explanation: Buddhism attempted to answer the strongly interrelated political and Śaiva threats by reformulating its symbolic, ritual and soteriological identity. This would be the “Buddhist esoteric synthesis” (to which most later Buddhist intellectuals in the epistemological tradition came to be associated), which has been noted by both Davidson and Sanderson for its intrinsic connection with political power. At any rate, Dharmakīrti’s works make it a compelling case that Buddhist epistemology and esoterism in statu nascendi are contemporaneous phenomena, and most likely two different answers to the same sociohistorical challenge.

Chapter 2 Buddhist Esoterism and Epistemology

2.1. INTRODUCTION Buddhist apocalyptic literature from the fifth to the sixth century witnesses seemingly new anxieties about threats external to the Buddhist order. These political as well as socioreligious factors were believed to herald, sometimes as characteristic features of the kaliyuga, the final demise of the saddharma.1 Although evidence of these prophecies of decline is scanty, it corroborates testimony in early Purāṇas concerning the growth of an “orthodox” Brahmanical hostility towards Buddhism, and seems further to reflect the rise of Śaivism to dominance in the religio-political sphere. Both in time and contents, it is likely that these prophecies reveal the Buddhist perception of the sociopolitical and religious circumstances that prevailed during the collapse of the Gupta Empire. R.M. Davidson and A. Sanderson have recently interpreted this historical matrix as having been instrumental in the rise of (at least “mature”) Buddhist Tantrism. Despite various differences in their hypotheses, both scholars emphasize the role played in the process by Śaivism and its religio-political appeal to a new class of sovereigns. However, the emergence of a soteriologically valued and politically committed form of esoterism is by no means the only event characterizing Buddhism at the dawn of the early medieval period.2 Among the many changes and innovations of the time, another that must be mentioned is the appearance of large monastic institutions dedicated to learning and ritual (Nālandā around 475 CE, Valabhī around 500 CE).3 No less 1

See above, Chapter 1, §1.3.

2

On the presuppositions (and the dangers) of the conceptualization of the period as “medieval,” see Wedemeyer 2013: 58–66, and also Davidson 2002: 28. On the historiography of Buddhist Tantrism, see Wedemeyer 2013, especially pp. 17–102.

3

On the Nālandā and Valabhī vihāramaṇḍalas, see Dutt 1988: 224–232, Davidson 2002: 105–111 and Sanderson 2009: 72–73 and 87–108.

94

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

importantly, the last decades of the Gupta Empire coincide with the rise of the Buddhist epistemological school,4 a phenomenon that, in my opinion, can be interpreted as a Buddhist answer to the sociopolitical, institutional, religious and philosophical challenges of that much troubled period.5 If considering the Buddhist literary landscape during the early medieval period (500–1200/1300 CE), i.e., during the last phase of Buddhism on the Indian soil, one can legitimately be struck by the fact that its two most productive components, and at the same time its two major innovations, are epistemology (as a methodological, philosophical and apologetic programme) and Tantrism (with a symbolic, ritual, soteriological and, most probably, politically oriented agenda).6 On one side, epistemology provides Madhyamaka with new tools and issues, absorbs and renews Yogācāra philosophy, and seems to attract, against the background of a rapid exhaustion of Abhidharma creativity, the best intellectuals of the time. On the other side, Tantrism introduces a new trend within the Mahāyāna—the selfconscious mantranaya/mantracaryānaya as opposed to the pāramitānaya7—and monopolizes symbolic, iconographic and speculative innovation. One feels justified, then, in examining the extent to which epistemology and Tantrism might be interpreted as the two main directions taken by the Buddhist elites in response to unprecedented challenges of a non-philosophical and non-symbolic order.8 For as different as they can be and indeed are, Buddhist epistemology and esoterism do share interesting discursive 4

On the chronology and literature of this school, see Steinkellner/Much 1995; on the philosophical programme of its main representative, Dharmakīrti, see Vetter 1964, Dunne 2004 and Eltschinger 2010c.

5

For a first attempt in this direction, see Eltschinger 2007a: 25–65; see also Eltschinger 2010c: 398–400 and 432–433.

6

See Kapstein 2003: 235.

7

See de Jong 1984: 92–93 and Omi 2008. The eighth-century commentator *Buddhaguhya (on this name, see Sanderson 2009: 131, n. 308) distinguishes between mantraand pāramitānaya whereas the MVASū speaks of mantracaryānaya (for Buddhist uses of the word naya, see Nance 2012: 144–145). Later authors such as Advayavajra and Ratnākaraśānti distinguish between mantrayāna and pāramitāyāna. Note, however, that they hold this distinction to be internal to the Mahāyāna. “Therefore, says de Jong, Tantrism is not, as is commonly assumed, a third yāna opposed to both Hīnayāna and Mahāyāna” (de Jong 1984: 93).

8

For a discussion of Buddhist esoterism as a response to sociopolitical circumstances and a “survival strategy,” see already Verardi 2011: 304–317.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

95

features, the most prominent being, in my opinion, their common drive to construe (and, at least in the case of epistemology, legitimate) themselves in explicit opposition to specific representatives of the non-Buddhist environment: in the case of epistemology, against Brahmanical orthodoxy by means of philosophical arguments; in the case of Buddhist Tantrism, against Śaivism at the sublimated level of mythology. In other words, by the sixth century, certain segments of the Buddhist monastic elites shaped the modalities of their self-assertion by contrasting themselves no longer with dissenting coreligionists, but with non-Buddhist challengers. This may well look trivial. However, if one considers that during the first half of the first millennium CE at least (coinciding with G. Schopen’s “Middle Period of Indian Buddhism,” an era of “status quo at almost every cost”9), the different Buddhist denominations had been fighting each other with regard to their soteriological systems (Mahāyāna versus Śrāvakayāna/Hīnayāna, etc.) and doctrinal controversy (intersectarian rivalries as recorded in Abhidharma literature, Yogācāra versus Madhyamaka, etc.), thus largely neglecting the non-Buddhists, the simultaneous rise of two movements deeply involved in polemics and competition with outsiders (tīrthika) deserves closer examination. Two complementary arguments strengthen the hypothesis of a significant shift in the focus of the sixth-century Buddhist elites: first, the conspicuous decrease in Abhidharmic creativity and intersectarian polemics after the fifth century; second, the exponential increase, in doctrinal and philosophical literature, of supersectarian self-depictions such as Bauddha, Saugata and Śākya(putrīya), which are almost totally missing in Abhidharmic literature. But, one may ask, in sixth- to seventh-century India actually how new were these two movements? Would they not be better explained as the (nearly) mature outcome of two already old traditions, the so-called Buddhist dialectical tradition in the case of epistemology, and “proto-Tantrism” in the case of esoterism? Indeed, their historiography shares several basic assumptions. Both insist on epistemology and Tantrism as the mature products of gradual, linear and incremental developments within organic traditions committed to debate and magic respectively. In other words, both are inclined towards teleology in that they interpret whatever is found to resemble, structurally or morphologically, an element of the mature system as a milestone in its development. Neither trend regards external, viz. 9

Schopen 1995: 478(/96), quoted above, Chapter 1, §1.3.3.

96

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

political, social and economic factors as being in any way involved in the rise of what they purport to explain, considering these factors, at best, to be a neutral setting. To sum up, both historiographies are strongly committed to incrementalist, teleological and internalist assumptions. However, at least in the field of Tantric studies these presuppositions have been seriously challenged in the last decade, and I do believe they should be called into question with regard to the rise of Buddhist epistemology as well. This is certainly not to deny empirically observable patterns of traditional, “internal” and “vertical” dynamics (inspiration, foundation, borrowing, adaptation, quotation, commentary, legitimation, etc.), nor to claim spontaneous generation in intellectual history. Rather, it must be acknowledged that the rise of intellectual systems is located at the crossroads of both vertical and horizontal forces, in a historical matrix comprised of internal/traditional and external/non-intellectual factors. 2.2. ŚAIVISM, EARLY MEDIEVAL INDIA AND THE RISE OF BUDDHIST TANTRISM 2.2.1. Preliminary Remarks In its attempt to circumscribe the nature, the rise and the successive stages of Buddhist esoterism, Japanese and Western Tantric scholarship has long been (and to a great extent remains) indebted to two traditional classification systems. According to the Indo-Tibetan one, whose origin can be traced to the eighth-century Indian commentator *Buddhaguhya, Buddhist Tantric literature can be divided into four or five classes, viz. kriyā-, caryā, yoga- (yogottara-/mahāyoga-) and yoganiruttaratantras.10 According to 10

See de Jong 1984: 93, Sanderson 1994: 97–98, n. 1, Omi 2008 and Sanderson 2009: 145–147, n. 337. To sum up Sanderson’s rich materials: The eighth-century commentator *Buddhaguhya distinguishes between kriyātantras (specialized in *bāhyacaryā [phyi’i spyod], such as the Susiddhikara, the Vidyādharapiṭaka and the Vajrapāṇyabhiṣeka), and yogatantras (emphasizing *adhyātmayoga [naṅ gi sbyor], such as the STTS and the MVASū, the latter being alternatively classified as a kriyā° or an ubhayatantra). Still in the eighth century, Vilāsavajra distinguishes between kriyā°, caryā° and yogatantras, the latter including works such as the Guhyasamāja and the Sarvabuddhasamāyoga (concerning the origin of the caryā class, Sanderson hypothesizes a borrowing from the Śaiva division into kriyā, caryā, yoga and jñāna/vidyā). Later authors (Rāmapāla, Ratnākaraśānti, Kāṇha, Advayavajra) know of five classes of Tantras, viz. kriyā°,

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

97

the Japanese one, which goes back to Kukai (✵ᾏ)/Kobo Daishi (ᘯἲ኱ᖌ, 774–835), Buddhist esoterism (mikkyo, ᐦᩍ) is to be divided into “mixed/eclectic Tantrism” (zomitsu, 㞧ᐦ) and “pure Tantrism” (junmitsu, 11 ⣧ᐦ). Insofar as the distinctions that these two systems introduced are based on criteria of contents and soteriological value, both are of primarily taxonomic and normative/hierarchical intent, but involve at least implicit chronological assumptions. And indeed, although the Shingon (┿ゝ)/Tendai (ኳྎ) classification is not meant to include the yoginī- or yoganiruttaratantras (largely rejected by Sino-Japanese esoterism on account of their antinomian and allegedly non-Buddhist character), the two systems overlap in regarding the rise of “true,” soteriologically valued Tantrism (i.e., caryā-/yogatantras and junmitsu) as coinciding with the composition (or compilation) of the MVASū and the STTS. As a consequence, the IndoTibetan kriyātantra class and the Japanese zomitsu overlap both in their alleged non-soteriological, mundane import and their (frequent) chronological anteriority, and coincide to a large extent, both in scope and significance, with the modern academic category of “proto-Tantrism.” As this designation makes clear, “proto-Tantrism” refers to those texts and practices whose gradual development, associated with a process of systematization and reorganization, is held to have led to the coalescence of “mature” Buddhist esoterism or the Buddhist esoteric synthesis. According to Hodge,

caryā°, yoga°, yogottara° and yoganiruttaratantras. The distinction between yogatantras such as the STTS and the more esoteric yogottara° or mahāyogatantras such as the Guhyasamāja is used by Āryadeva and Dīpaṅkaraśrījñāna. As for the expression *anuttarayoga, which is apparently not to be found in Indic materials, it is a modern mistranslation of the Tibetan expression rnal ’byor bla na med pa (= yoganiruttara). 11

See, e.g., de Jong 1984: 92, Strickmann 1996: 127–133 and Shinohara 2010: 389–390 (for additional references, see Shinohara 2010: 389, n. 1). De Jong 1984: 92: “According to this terminology zomitsu indicated the Tantric doctrines not included in the systems of the Dainichikyo (኱᪥⤒) and the Kongochokyo (㔠๛㡬⤒) (T. 848 and 865). It is only much later that these two terms were used by Eko (្ග, 1666–1734) to indicate an opposition implying a value judgement. In Shingon studies zomitsu is used as a deprecatory term to indicate ritual and dhāraṇī texts which have as their chief goal worldly goods and benefits, whereas junmitsu designates Dainichikyo and the Kongochokyo texts which deal mainly with the attainment of bodhi.” According to Shinohara (2010: 389– 390), zomitsu refers to “rituals based on simple recitation of spells often accompanied by hand gestures,” while junmitsu consists of “rituals carried out through visualization […] often characterized as oriented toward attaining enlightenment.”

98

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

this “evolutionary process”12 can be characterized as a “gradual progression from external ‘mundane’ rituals and objectives to the internal and the ‘spiritual,’ from the un-systematic to the systematic.”13 According to Matsunaga, this “development”14 consisted in a “process of change from unsystematic texts to systematic ones, from simple to complicated ones, and from external rituals to serious internal meditation.”15 From the third to the seventh century CE, this process can best be seen in the Chinese “translations” of dhāraṇīsūtra literature, whose evolution reflects the gradual enrichment of the ritual syntax and elements surrounding the use of mantras, vidyās and dhāraṇīs.16 Characteristic of this historiographical stance is the work of the Japanese scholar Yukei Matsunaga,17 to whom we owe much of our knowledge as to how, where and when hand gestures (mudrā), fire sacrifices (homa), maṇḍalas, (self-)visualization techniques, consecrations (abhiṣeka), buddha families and iconographic prescriptions successively entered Mahāyānistic ritual. Matsunaga’s studies inform us about the chronology of Indian Buddhist ritual trends and innovations (despite being seen through the lens of Chinese “apocrypha,” which, as Matsunaga and Strickmann have both argued, are usually based on Indic materials) and the sixth- to seventh-century turn to soteriologically valued esoterism. At times he even hypothesizes a large-scale “absorption,” on the part of the Buddhists, of the “rituals of Brahmanism [and] the deities of Hinduism.”18 According to scholars like Matsunaga, then, the composition/compilation of the MVASū and STTS, which sets the divide between 12

Hodge 2003: 8. Note also Hodge: 2003: 5: “process of synthesis and development;” Matsunaga 1978: xix: “gradual process of development;” Matsunaga 1978: ix: “development which resulted in the emergence of Buddhist Tantrism.” Shinohara 2010: 391: “process of evolution;” Shinohara (ibid.): “stages in an evolutionary process.”

13

Hodge 2003: 5.

14

Matsunaga 1978: viii, xi, xii, xiii, xiv, and passim.

15

Matsunaga 1978: xix. Note also Shinohara 2010: 400, pointing to a “process in which the simpler recitation of dhāraṇīs first evolved into more complex rituals around specific deities.” Shinohara’s excellent study aims to trace “how the simpler ritual practice of recitation of spells gradually evolved into complex esoteric initiation ceremonies” (2010: 390).

16

Recent studies on vidyās and dhāraṇīs include Davidson 2009, Shinohara 2010, and Hidas 2012.

17

For a review and detailed presentation of Matsunaga 1980, see de Jong 1984.

18

Matsunaga 1978: xi.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

99

proto-Tantrism and Tantrism as a self-contained soteriological method, is the outcome of a continuous and mostly internal incrementation process. However, be it on the level of its systematicity or its supposedly new soteriological orientation, the rise of “pure Tantrism” is at the same time depicted as involving a radical break with regard to proto-Tantrism.19 This tension between a cumulative process and a sudden shift in scale and finality, according to which “on est tenté d’imaginer tous les éléments anciens en suspension, sans attaches précises, attendant leur cristallisation,”20 has been criticised by Strickmann. According to him, the older practices already functioned within their own coherent framework, so that “il devient évident […] qu’il n’y a jamais eu de continuité totale entre les anciennes doctrines et les nouvelles.”21 Moreover, Strickmann holds the zomitsu-junmitsu distinction, at least insofar as it is based on a dichotomy between mundane and supramundane expectations, to be “artificielle,”22 for it is

19

De Jong 1984: 98: “The texts of this period can be distinguished from those of the first period by the following five characteristics. 1. The goal of the Tantric rituals is no longer worldly gain but the attainment of buddhahood (jobutsu [ᡂ௖]; abhisambodhi). 2. In the early Tantras, mudrās, dhāraṇīs and samādhis were not systematically arranged. In the middle period they are grouped together systematically as the three secrets (sanmitsu [୕ᐦ] of body, speech and mind [kāyavākcitta]). 3. The content of many early Tantras was magical, with no direct relation to Buddhist teaching. In the Tantras of the middle period, Mahāyāna ideas are incorporated in Tantric ritual. M[atsunaga] proposes to call this the ritualisation (gikishika, ൤ᘧ໬) of Mahāyāna ideas. 4. A special characteristic in the Tantras in the middle period is the formation of maṇḍalas, namely the Mahākaruṇāgarbhodbhavamaṇḍala which is based on the Dainichikyo and the Vajradhātumaṇḍala which is described in the Shinjitsushokyo (┾ᐿᨡ⥂, STTS, T. 882). 5. In the early Tantras the preacher is Śākyamuni, but in the Tantras of the middle period the preacher is the Tathāgata Mahāvairocana. According to the tradition of Japanese Tantrism, the preaching of Mahāvairocana is that of the dharmakāya as opposed to the preaching of the nirmāṇakāya by Śākyamuni.”

20

Strickmann 1996: 133. Note also Davidson 2002: 145 (concerning the dhāraṇīsūtra literature and the constitution of a dhāraṇīpiṭaka): “This is the material that will come to be designated as ‘proto-Tantric’ in critical literature, although it provides us with a misleading sense that somehow these collections understood that they were anticipating the later, mature system.”

21

Strickmann 1996: 131.

22

Strickmann 1996: 131.

100

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

difficile d’isoler dans le prototantrisme des exemples de réalisation de souhaits “mondains” tout en négligeant des passages identiques du tantrisme “pur”, sous prétexte que ces derniers devraient s’interpréter métaphoriquement ou anagogiquement.23

But Strickmann is also to be credited for drawing the political into the picture by arguing that Tantrism, far from being a bottom-up movement, involves a move “vers le bas et vers l’extérieur, des cours vers les provinces et le peuple.”24 According to him, Tantric rituals concentrate on power; the mythology connected with them is of an almost exclusively martial character, “une épopée martiale à la conquête des dieux païens.”25 Moreover, “[l’]une des principales fonctions du tantrisme a toujours et partout été la protection de l’État,”26 a function involving issues of political patronage and merit transference. 2.2.2. Buddhist Tantrism as the Internalization of the Early Medieval Political Landscape The issues of power, state, aesthetics and patronage form the cornerstones of R.M. Davidson’s influential (and controversial) Indian Esoteric Buddhism (2002), a deeply innovative study interpreting the rise of Buddhist esoterism at the end of the seventh century as reflecting the Buddhists’ internalization and sacralization of the early medieval political landscape. Most of Davidson’s book can be interpreted as a reaction against the dominant incrementalist and teleological historiography of Buddhist Tantrism. Expressions such as “esoteric turn” (2002: 163) reveal Davidson’s commitment to “discontinuist” historiographical models; as for his interpretation of Buddhist Tantrism as a “consequence” (2002: 167) or an “extension” (2002: 123) of the early medieval sociopolitical reality, it betrays 23

Strickmann 1996: 131. Note also Snellgrove 2004: 235–236: “It is sometimes suggested that while the tantras, later classified as inferior, cater for the more mundane requirements, the superior ones are concerned with more truly religious objectives. In fact all tantras are interested in precisely the same objectives, whether supramundane or mundane.” For a similar remark concerning the zomitsu-junmitsu distinction, see Shinohara 2010: 389–390.

24

Strickmann 1996: 43.

25

Strickmann 1996: 41.

26

Strickmann 1996: 41.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

101

strong “externalist” assumptions. Davidson’s account is, however, far more subtle than my gross methodological characterization suggests. As a “mature synthesis” (2002: 114), Buddhist esoterism certainly presupposes and reinterprets most of the ritual, magical and meditative elements that have come to be labelled “proto-Tantric.” But far from being the outcome of a gradual, linear, incremental and purely internal process, it owes its existence as a synthesis to sociopolitical circumstances. Were it not for a complex matrix of political, economic and social factors, it would not have come into being despite the availability of these “proto-Tantric” elements.27 The early medieval period that opens up at the collapse of the Gupta Empire is characterized by a political fragmentation entailing a strong movement toward “regionalization” (2002: 29, passim) and the rise of “new regional centers” (2002: 70) of political authority.28 According to Davidson, the early medieval basic political structure is best described as the “segmentary state”29 (2002: 135) or more generally as “sāmanta-feudalism”30 (2002: 137), with rapidly changing relations of hierarchical subordination between core/nuclear areas and peripheral/vassal areas. For the Buddhist institutions, this period coincided with unprecedented challenges and duress. For quite a long time, they had “depended as much on their symbiotic relationship with the guilds of Indian tradesmen and merchants as on their attractiveness to princes” (2002: 77). But in the aftermath of the Guptas, the new political structures, attitudes and aesthetics as well as the redistribution of trade routes in Central Asia destabilized the guilds. In particular, “much of the wealth that had previously been in the hands of long-distance merchants and caravan directors became accrued by those in 27

See Davidson 2002: 145 (quoted above, p. 99, n. 20) and 2002: 166.

28

See Davidson 2002: 29.

29

Davidson 2002: 135: “A segmentary state is a political system with numerous centers, each of which has separate administrative systems, frequently clan-based; power is wielded by a single overlord whose subordinates in the lesser centers recognize his authority through ritual means.”

30

For a discussion of the expression “sāmanta-feudalism,” see Davidson 2002: 137, especially: “[T]he designation sāmanta is only one of many employed—also including hierarchical levels—such as mahāsāmanta, mahāmaṇḍaleśvara, maṇḍaleśvara, mahārāja, and rāja. These were all used to indicate various levels of subordination to a lord, who was often given the title Great King of Kings (mahārājādhirāja) or Overlord (paramabhaṭṭāraka), and who maintained a circle of subordinate quasi-states (sāmantacakra, sāmantamaṇḍala).” See also Verardi 2011: 197–204.

102

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

positions of political power” (2002: 78). As a consequence, the Buddhist institutions became increasingly dependent on political power and appreciation, so that they were “secure primarily in those areas where the Buddhist tradition was not seen as the adversary of the state or statesponsored social systems.31 Yet this was a difficult proposition for Buddhists, as Hindu divinities and Purāṇic ritual systems appeared to offer a closer fit in needs and values espoused by medieval states” (2002: 83–84). And indeed, this context of “military adventurism” (2002: 67, passim), “opportunism” (2002: 62) and “endless warfare” (2002: 42) induced new needs regarding issues of legitimacy and identity as well as substantial transformations in the representation of rulers, now considered divine or manifestations of divinities. This “apotheosis of kings” (2002: 68) entailed a concomitant feudalization or “sāmantization” (2002: 71) of the gods32 as well as an “eroticisation of warfare” (2002: 70; see also 2002: 169) which could only promote Śiva’s activity as “the model for medieval terrestrial monarchs, because of both his violent manifestations and the erotic nature of his aesthetics and iconology” (2002: 72). Unsurprisingly, the period “saw a very rapid buildup of temple construction, most frequently for reasons of political expediency and cultural authority” (2002: 72), the development, sponsored by those rulers, of “massive Śaiva and Vaiṣṇava temples” (2002: 29). Moreover, “seeking legitimacy and identity, Indian kings from all areas began to increase their patronage of literature and to strategize their support for religion, searching for religious counselors that could bolster their political and military agendas. They also developed patronage relations with those who could validate their local aesthetics and valorize their locales” (2002: 26). But “Buddhist virtues did not present compelling reasons for patronage, when a king could just as easily reformulate his image in favor of the model of Śiva, who was, after all, represented as a killer divinity with a permanent erection” (2002: 90). Or, as Davidson says, feudal cultures engaging in opportunistic belligerence presumed that religious groups receiving patronage would valorize their behavior and vindicate their authority. This brings us to […] the problems Buddhists had with the evolving ethics 31

Davidson 2002: 167: “Institutional Buddhism responded by contracting into regions of strength and into edifices mimicking feudally grounded fortresses, which mirrored in legal behavior the activities of the kings they cultivated.”

32

On these expressions, see Davidson 2002: 71–72.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

103

of patronage. While it is true that certain warlords who succeeded in putting together a confederation of belligerents ultimately found their way into the Buddhist fold, this was relatively rare […] Overall, Buddhist institutions could not effectively compete for patronage from militaristic princes, who increasingly found that they were best represented by Śaiva values and rhetoric. Śaiva systems made allowance for forms of behavior that Buddhist syntheses could not support, since even the most syncretic Buddhist systems were not as open to negotiation about issues of violence, power, self-aggrandizement as were the medieval Śaiva representatives.33

This point is worth noticing, for contrary to what certain accounts of the book suggest, Davidson does indeed lay strong emphasis on Śaivism when analyzing the factors responsible for the rise of Buddhist Tantrism. He does, however, reject “the model that Buddhist esoterism is the pale imitation of Śaivism” (2002: 113).34 According to him, [e]soteric Buddhism […] was not simply a reaction to the new environment, but stemmed from the palpable sense of institutional duress. The decline of guilds and international trade relations, the rise of militant Śaivism and its capacity to appropriate patronage […] all set the institutions adrift.35

And [i]t is in this new world of rapidly changing feudal alliances and evolving religious orders that the great Buddhist monasteries and lay associations would find 36 themselves challenged on all sides.

33

Davidson 2002: 86.

34

In his Tibetan Renaissance (2004), Davidson admits that “[t]he principal religious influences of the period stemmed from the Śaiva cultus supported by the predominantly Śaiva southern states” (2004: 27). Or: “Buddhists consistently represent Śaiva ascetics— along with the ordinary, or garden-variety Brahmans—as the principal antagonists to the Buddhadharma. Especially notable were the extreme groups among the Śaivas, those employing human bones or extraordinary behaviors (e.g., omophagy, scatophagy) as part of their ritual practices. A few groups (especially the Kaulas) engaged in ritual copulation, and the tribal peoples frequently offered human sacrifices to the goddesses of the earth or the locale. All these activities tended to impress Buddhist authors as objects of fascination, derision, imitation, or other responses” (2004: 27).

35

Davidson 2002: 167.

36

Davidson 2002: 74. Note also Davidson 2002: 167: “The narrative use of violence in the context of Buddhist institutions allowed institutional esoterism to compete with Śaivism,

104

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

According to Davidson, Buddhist esoterism came as “a response to the difficult medieval environment and a strategy for religious reaffirmation” (2002: 113) which allowed the Buddhist institutions to “interact with the warlords and princes, the military generals and the emerging tribal leaders” (2002: 114). The Buddhists’ reaction, i.e., the mature esoteric synthesis, is to be interpreted both as an “internalization” (2002: 115, passim) and a “sacralization” (2002: 165, passim)/“sanctification” (2002: 160–161) of the “structure, aesthetics, and ideology of medieval Indian feudalism” (2002: 115). Davidson’s endeavour consists, then, in revealing the “implicit political model of the Mantrayāna” (2002: 123), the way in which Buddhist institutions “model[led] the new meditative practices after systems of political power” (2002: 166) by internaliz[ing], appropriat[ing], reaffirm[ing], and rearrang[ing] the structures most closely associated with the systems of power relations, ritual authentication, aesthetics, gift-giving, clan association, and sense of dominion that defined post37 Gupta Indian polities.

In other words, the monastic institutions “reacted to a predatory feudal system by emulating its form while subverting its goals” (2002: 132–133). This internalization of the sociopolitical environment is best seen in the remarkable “bivalence” (2002: 121) manifest in Tantric vocabulary, where “so many significant terms found in the standard esoteric ritual manuals and the Buddhist Tantras have political and military significance as well as religious” (2002: 121). This vocabulary reveals the sustaining metaphor of Buddhist esoterism, the “imperial metaphor,” a “paradigm of dominance, hierarchy and regal power” (2002: 122). Here, “the metaphor of the esoteric to appeal to the worst instincts of the warlords of the medieval period and yet to delimit the nature of approved violence.” Davidson 2002: 339: “[E]soteric Buddhism demonstrated tenacious success in India for more than five centuries at a time when the dynamics of the subcontinent were rapidly changing, and Buddhist institutions were in retreat in the Kṛṣṇa River Valley and elsewhere. It assisted the maintainance of the great monasteries and stemmed the Śaiva tide sweeping up from the south.” Davidson 2004: 29: “Because they were unable to come to terms with many of the southern kings and lords, Buddhist communities in the Kṛṣṇa River Valley—a site of extraordinary Buddhist activity for almost a thousand years—slowly disappeared in the rising tide of militant Śaivism. As a consequence, many Buddhist communities contracted into larger, fortresslike monasteries in North India, West India, the far south, and especially the areas of royal patronage in the east.” 37

Davidson 2002: 115. For a qualification of this position, see Davidson 2002: 160–161.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

105

meditator becoming the Rājādhirāja is sustained through the multiple forms of ritualization that are equally applicable to kings and tantrikas” (2002: 122). According to Davidson, “the different practices were synthetized into a nexus whose overarching narrative was that of divine kingship in early medieval feudal world of India” (2002: 122). The “mature esoteric synthesis” (2002: 117) consists, then, in the coalescence, around the “metaphor of an individual assuming kingship and exercising dominion” (2002: 121), of numerous esoteric materials that had hitherto been “individual aids, not a unified system” (2002: 117). These mantras, visualisation techniques, consecrations, maṇḍalas and hand gestures were previously “not considered constituents of a self-contained path and did not contribute a sense of identity within the community” (2002: 117). In other words, the “mature synthesis of esoteric Buddhism” is the form “defined as a separate method or vehicle employing mantras” (2002: 114), and which insisted on an immutable master-disciple bond, employed royal acts of consecration, and used elaborate maṇḍalas in which the meditator was to envision himself as the Buddha in a field of subordinate Buddhas.38

According to Davidson, the way of secret mantras that emerges between 650 and 700 is a synthesis, a system, but one that was made possible by Buddhism’s “acculturat[ion] to the medieval Indian landscape,” by the “politicization of Buddhism” (2002: 144).39 Note, however, that in spite of its being a “ritualization of external circumstances” (2002: 165) and emulating Śaiva ways of addressing the sociopolitical reality, the new esoteric system is unmistakably Buddhist: for its creators, the Buddhist path was still the content—and most of the institutional tantras are explicit in this—but its ritualization put that content in a new vessel, with accoutrements of power and a new, self-confident authority. And because it was much the

38

Davidson 2002: 114.

39

According to Davidson (2002: 117), the “self-description of mature esoteric Buddhism as the way of secret mantras (guhyamantrayāna) is analogous to Mahayana’s selfdescription as the way of the bodhisattva (bodhisattvayāna). The nomenclature and ideology of bodhisattvas (Siddhārtha, Maitreya) had been around long before it became embedded in a different ritualized ‘way,’ with new vows and a new path topography. In a similar manner, the identification and use of mantras (or dhāraṇī or vidyā) had existed for centuries before there arose a new ritualized synthesis of different factors.”

106

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

same content, and since it was constructed using building blocks largely present in advance (mantras, maṇḍalas, homas, etc.), the transition was relatively smooth, with a remarkable lack of protest from others in the Buddhist community about 40 the new turn to ritual.

2.2.3. Buddhist Esoterism as an Adoption-cum-Adaptation of Śaiva Tantrism For nearly two decades, A. Sanderson has stressed Buddhist Tantrism’s strong indebtedness to both Śaiva Siddhānta and Śāktism, a claim that has long been almost exclusively based on the comparative evidence provided by textual criticism41 and mainly linked to the so-called Yoginītantras.42 Sanderson’s monumental “Śaiva Age” (2009) now provides this hypothesis with a detailed religio-political foundation and extends it to works classified as caryā- and yogatantras. By emphasizing the historical matrix characterizing the early medieval period, this context’s instrumentality in Śaivism’s “rise to dominance” (2009: 252) and hence its decisive impact on the formation of the Buddhist Mantranaya, Sanderson not only breaks with “continuist” and incrementalist models, but also rejects previous accounts of Buddhist Tantrism as a mature synthesis of formerly unsystematic ritual elements. At first sight, Sanderson’s statement about the “causes of the dominance of Śaivism” (2009: 252) do not seem to diverge significantly from Davidson’s, for according to him,

40

Davidson 2002: 163. As we shall see below (§2.2.6), Dharmakīrti is likely to have reacted against certain Tantric doctrines, such as the natural efficacy (bhāvaśakti) of mantras and the soteric relevance of initiation (dīkṣā).

41

Note, however, Sanderson 2004: 230: “[Inscriptional evidence] at best […] instantiates or challenges the model of possibilities conveyed by the prescriptive literature. It does not enable us to go beyond its range into detailed social history. Nonetheless it is possible, I would say necessary, to read the literature and inscriptions with the sort of questions in mind that a social historian would wish to ask.”

42

Note, however, Sanderson 1994: 96: “When we consider Tantric Buddhism in terms of its origin we see Śaiva influence at almost every turn; and the higher one goes up the hierarchy of the Buddhist Tantras, the most pervasive these influences become. However, Tantric Buddhism is, of course, entirely Buddhist in terms of its function and selfperception; and in transforming Śaiva elements it gave them meanings which obscure these origins.”

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

107

the fundamental reason for the religion’s success, underlying and structuring the mass of particulars now lost to view, was that it greatly increased its appeal to royal patrons by extending and adapting its repertoire to contain a body of rituals and theory that legitimated, empowered, or promoted key elements of the social, political and economic process that characterizes the early medieval period.43

Instead of insisting on Śaivism’s appeal to the new monarchs on account of its martial and erotic aesthetics, Sanderson stresses the Śaivas’ ability to match new religio-political needs by providing sovereigns with a concrete set of ritual, sacerdotal, intellectual and socioreligious services.44 According to Sanderson, the “emergence of numerous new dynasties at subregional, regional, and supraregional levels” (2009: 253) is typical of the early medieval period.45 This process entailed an “extension of the reach of the state” (2009: 284) and a “progressive penetration of the authority of the centre into new territories” (2009: 253). These “expanding states” (2009: 285) created “new urban settlements from above”46 (2009: 281) that required an “expansion of the agrarian base” (2009: 253). And in spite of their “personal religious commitment to […] the deities of the new initiatory religions” (2009: 43), the early medieval Indian kings remained the 43

Sanderson 2009: 253.

44

According to Sanderson 2004: 231–233, the success of Śaiva Mantramārga from the sixth century on is due to three factors. (1) “[F]ollowing the example of the Buddhists,” this religion “cultivated the support of a wider community of uninitiated lay devotees” with a “lesser religion of merit gathering” (2004: 231). (2) It accommodated “the brahmanical religion that it claimed to transcend” (2a) by having its devotees “accept that the brahmanical tradition alone was valid in the domain it claimed for itself” (2004: 231) and (2b) by “tak[ing] over, preserv[ing], and regulat[ing] in accordance with the expectations of the uninitiated laity a much wider range of ancillary deities” (2004: 232). (3) Śaivism “succeeded in forging close links with the institution of kingship and thereby with the principal source of patronage” (2004: 232), (3a) by having Śaiva officiants occupy the office of Royal Preceptor; (3b) by “promoting […] the practice of displaying and legitimating a dynasty’s power by their officiating in the founding of Śaiva temples” (2004: 233); (3c) by providing “a repertoire of protective, therapeutic and aggressive rites for the benefit of the monarch and his kingdom” (2004: 233); (3d) by developing “Śaiva rituals and their application to enable a specialized class of Śaiva officiants to encroach upon the territory of the Rājapurohita” (2004: 233; “warding off all manner of ills from him through apotropaic rites, using sorcery to attack his enemies,” etc.).

45

See Sanderson 2009: 254–273.

46

As against “commercial centres that grew from below through a process of agglomeration” (2009: 253).

108

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

guardians of the varṇāśramadharma and promoted the penetration of the “values of brahmanical society” (2009: 285).47 In other words, the early medieval period saw the “imposition, reinforcement, and expansion” (2009: 43) of the “monarchical model of government” (2009: 253), of Brahmanism (2009: 43), “rural economy” (2009: 253) and urban centres. The process involved issues of religio-political legitimacy, patronage and socioeconomic welfare that Śaivas were the first and probably the best able to address. First,48 this new class of sovereigns endeavoured to give “material form to the legitimacy and solidity of their power” (2009: 274) by erecting “royal temples in nuclear areas” and “lesser temples in peripheral zones” (2009: 253), which they “endowed with land and officiants to support their cult” (2009: 274) and where they had images of their chosen deities installed. According to Sanderson, “the ceremonial repertoire of these temples included special rituals for the king’s protection,” and “the texts place a great emphasis on the connection between the temple and the welfare of the ruler and his kingdom” (2009: 279). The Śaivas of the Mantramārga “provided specialized officiants and rituals,” the most prominent among the former being the sthāpaka assumed to be “competent not only in the Śaiva domain but also on all the levels that the Śaivas ranked below it” (2009: 274, including in certain sources Buddhist, Jaina, Vaiṣṇava and Vaidika cults and images49). Second,50 the “demonstration of [the king’s] sovereignty” (2009: 281) did not only concern the building of landowning temples, but also the foundation of new planned urban centres which saw the use of “Śaiva officiants on a purely civic level” (2009: 282). 47

Note Sanderson 2009: 41: “The early medieval period, from about the fifth century to the thirteenth, saw a decline in the role of Śrauta sacrifice in the religious ceremonies undertaken by Indian rulers. But it was not that kings turned aside from the brahmanical tradition in a fundamental sense. They continued to uphold the brahmanical social order of the castes and disciplines (varṇāśramadharmaḥ) and they were commonly commended in inscriptions from the fifth to the eighth centuries for having rigorously imposed it on their subjects.” Note also Sanderson 2009: 43: “Moreover, there is abundant epigraphical evidence of kings throughout this time bringing Vaidika brahmins into their kingdoms by making them grants of tax-exempt land, thereby extending the penetration of brahmanical culture while facilitating the administration of their territories and promoting agricultural development.”

48

See Sanderson 2009: 274–279.

49

See Sanderson 2004: 232, n. 4, and 2009: 274–275, n. 652.

50

See Sanderson 2009: 280–282.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

109

And indeed, Śaivas were involved in “both urban and rural planning” (2009: 282) insofar as their competence ranged from the “choice and consecration of the site” to the “layout of the royal palace” and the “urban settlement […] complete with markets and segregated areas for the dwellings of the various castes and artisans” (2009: 281). Third,51 “agricultural development” (2009: 43) required the “construction of water-reservoirs, wells, and other means of irrigation” (2009: 253). Though on the basis of Brahmanical prototypes, here too the Śaivas sought to “add this important domain to their ritual repertoire” (2009: 283), providing “irrigation rituals” (2009: 282–283) and the “consecration of irrigation works undertaken by their royal patrons” (2009: 284). Fourth,52 the Saiddhāntika and Śākta socioreligious models, extending recruitment “into lower social strata” (2009: 290), proved to be invaluable in furthering “cultural and religious assimilation” (2009: 253) and “social integration” (2009: 284). By opening “initiation to candidates from all the four caste-classes” (2009: 285, especially among the landowning agriculturalists, classed as sacchūdras due to their abjuring alcohol) and claiming that “all devotees of Śiva form a single community” (2009: 289),53 the Saiddhāntikas in particular “provided a means of articulating a social unity that transcended the rigid exclusions of the brahmanical social order” (2009: 285). No less importantly, the Śaivas are likely to have strongly contributed to the “incorporation of the many local deity-cults of the regions being drawn into the orbit of the state and its patronage of religion” (2009: 301).54 According to Sanderson, the dominance of Śaivism in the religiopolitical sphere can be seen from two sets of evidence. First, the epigraphic records reveal that the kings’

51

See Sanderson 2009: 282–284.

52

See Sanderson 2009: 284–301.

53

See Sanderson 2009: 289–290.

54

Note Sanderson 2009: 301: “There is another manner in which Śaivism is likely to have played a significant part in the process of social integration during this period, one which I wish to touch on only briefly and tentatively at this stage. This was in the incorporation of the many local deity-cults of the regions being drawn into the orbit of the state and its patronage of religion. In this it seems that it was the non-Saiddhāntika traditions of the worship of Bhairavas, goddesses, and Yoginīs, with their indifference to caste-status and brahmanical criteria of purity and their cults of possession that are likely to have provided the avenue of assimilation.”

110

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

personal religious commitment generally took the form of Buddhism, Jainism, or, more commonly, devotion to Śiva, Viṣṇu, the Sun-God (Sūrya/Āditya), or the Goddess (Bhagavatī),55

and that “[a]mong these alternatives devotion to Śiva was the most commonly adopted” (2009: 44). Second, [t]he preponderance of Śaivism during this period is also revealed by evidence that all the other religious traditions competing for patronage were colonized or profoundly influenced by it,56

an evidence that Sanderson’s subtitles make abundantly clear: “Etiolation and Subsumption of the Cult of the Sun-God,”57 “The Decline of Vaiṣṇavism and the Rise of the Tantric Pañcarātra Following Śaiva models,”58 “The Development of Tantric Buddhism Through the Adoption and Adaptation of Śaiva and Śākta Śaiva Models,”59 “The Jains’ Adaptation of the Śaiva Mantraśāstra,”60 and “Śaivism in the Brahmanical Substrate.”61 Sanderson’s detailed analysis of the epigraphic evidence concerning royal patronage of Buddhism provides a suggestive picture of the contrasting economic and political circumstances faced by the monastic institutions during the early medieval period. As one could expect, eastern India under Pāla rule houses Buddhism’s most flourishing establishments of the time (Nālandā, Vikramaśīla, Somapura, Trikaṭuka, Uddaṇḍapura,62 Jagaddala). But as Sanderson makes clear, neither was “Buddhism, for all the liberal support it received from the Pālas, [in a] position to oust or diminish Śaivism” (2009: 116),63 nor did

55

Sanderson 2009: 43.

56

Sanderson 2009: 44–45.

57

See Sanderson 2009: 53–58.

58

See Sanderson 2009: 58–70.

59

See Sanderson 2009: 124–243.

60

See Sanderson 2009: 243–249.

61

See Sanderson 2009: 249–252.

62

On this name, see Sanderson 2009: 92–93, n. 169.

63

Note Sanderson 2009: 116–117: “The monasteries themselves reflect this symbiosis. The excavations at Somapura revealed an abundance of non-Buddhist deities, particularly Śiva, among the stone relief sculptures around the base of the central temple and the very numerous terracotta plaques that decorated its walls. Excavations of the Vikra-

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

111

royal devotion to the Buddha […weaken…] the traditional commitment of Indian rulers to the imposition and preservation of the caste-based brahmanical social 64 order in which Śaivism was embedded.

There is ample evidence that eastern Indian Buddhist vihāras “were state monasteries […] rather than autonomous, self-governing institutions” (2009: 107-108). And indeed, rituals for state protection were performed on behalf of the monarch at Vikramaśīla. We have seen above Tāranātha’s report of the fire-ritual performed for the benefit of the dynasty by the Vajrācāryas of that monastery; and two important texts on the ritual of initiation written by two major Tantric authorities under the early Pālas, the Sarvavajrodaya of Ānandagarbha and the Guhyasamājamaṇḍalavidhi of Dīpaṅkarabhadra, the successor of Buddhajñāna at Vikramaśīla, insert ancillary rites specifically for the averting of danger from the monarch. Moreover, Tāranātha relates several occasions on which Buddhist Tantric masters were believed to have used Tantric rituals to good effect against the enemies of their patrons in times of danger.65

But the reason why Buddhism managed to compete with Śaivism for royal patronage can most probably to be seen in the fact that it emulated the Śaiva ritual repertoire in order to meet the new religio-political needs.66 maśīla monastery also uncovered a mix of Buddhist and predominantly Śaiva nonBuddhist images, the latter Śiva, Umāmaheśvara, Śiva and Pārvatī, Bhairava, Mahiṣāsuramardinī, Pārvatī, Kaumārī, Cāmuṇḍā, Gaṇeśa, Kārtikeya, the Navagraha, Vṛṣabha, Viṣṇu, and Sūrya.” 64

Sanderson 2009: 115. Note also Sanderson 2009: 115–116: “[I]n the Neulpur grant of the Bhauma-Kara king Śubhākara I his grandfather Kṣemaṅkara is described both as a Buddhist and as having ensured that the members of the caste-classes and disciplines observed their prescribed roles; in his Teruṇḍiā copper-plate inscription Śubhākara II, the grandson of Śubhākara I, is given the epithet paramasaugataḥ yet is also commended for having ‘propagated the system of uncommingled caste-classes and disciplines proper to the [perfect] Kṛta Age following the unexcelled [brahmanical] scriptures;’ the Pāla Dharmapāla is described in a grant of his son Devapāla both as a paramasaugataḥ and as taking measures to ensure that castes that erred were made to adhere to their respective duties, thereby discharging his father’s debt to his deceased ancestors; and Vigrahapāla III is described in his Āṃgāchi copper-plate as the support of the four caste-classes.”

65

Sanderson 2009: 105–107.

66

Note Sanderson 2004: 231: “[T]here can be no doubt that for several centuries after the sixth [Śaivism] was the principal faith of the élites in large parts of the Indian subconti-

112

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

Most characteristic of this process is certainly “[t]he adoption of the Śaiva practice of Maṇḍala initiation,” which created a further line of access to patronage and was propagated vigorously, as it was by the Śaivas, as a means of the recruiting of social élites both in the subcontinent and beyond. Among the Buddhist Tantras at least two major texts teach rituals of initiation, or consecration (abhiṣekaḥ) as it is called in these sources, in which it is kings in particular and royalty in general that are envisaged as the primary initiands. These are the Mañjuśriyamūlakalpa and the Sarvadurgatipariśodhanatantra.67

Initiatory rituals did not exhaust the kings’ expectations, however, for these were basically concerned with belligerence as well as the socioeconomic welfare of their subjects. According to Sanderson with regard to eastern India, this co-existence of Buddhism and Śaivism under royal patronage was surely facilitated by the fact that the form of Buddhism adopted and developed was one that had equipped itself not only with a pantheon of ordered sets of deities that permitted such subsumptive equations but also with a repertoire of Tantric ceremonies that paralleled that of the Śaivas and indeed had modelled itself upon it, offering initiation by introduction before a Maṇḍala in which the central deity of the cult and its retinue of divine emanations have been installed, and a system of regular worship animated by the principle of identification with the deity of initiation (devatāhaṃkāraḥ, devatāgarvaḥ) through the use of Mantras, Mudrās, visualization, and fire-sacrifice (homaḥ); and this was presented not only as a new and more powerful means of attaining Buddha-hood but also, as in the Śaiva case, as enabling the production of supernatural effects (siddhiḥ) such as the averting of danger (śāntiḥ), the harming of enemies (abhicāraḥ), and the control of the rain (varṣāpaṇam and ativṛṣṭidhāraṇam), through symbolically appropriated inflections of the constituents of these procedures. The latter is particularly important from the point of view of Buddhism’s relations with its royal patrons, since such

nent and in both mainland and insular Southeast Asia. Only Mahāyāna Buddhism was able to rival it during this period; and when it achieved success in this rivalry, either equalling or excelling Śaivism as the beneficiary of patronage, it was in a form led by the Way of Mantras, a system of ritual, meditation and observance in which Buddhism had redesigned itself, if not in essence, then at least in style and range of functions, on the model of its rival.” 67

Sanderson 2009: 125.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

113

rituals enabled it to match the Śaivas by promising kings more tangible benefits than the mere accumulation of merit through the support of the Buddha, his teaching, and the Saṅgha.68

In order to attract patronage, the Buddhist monasteries were compelled to cover the whole array of their benefactors’ institutional and personal ritual needs, including the erection and maintenance of temples as well as the execution of funeral rites. In this, they followed the example of the Śaivas by devising Tantric ceremonies for patrons in the public domain: for the consecration (pratiṣṭhā) of temple images (pratimā), paintings of deities on cloth (paṭaḥ), manuscripts of sacred texts (pustakam), monasteries (vihāraḥ), shrines (gandhakuṭī), Caityas, reservoirs (puṣkariṇyādi), gardens and the like (ārāmādi). It also adapted the Śaiva procedures for funerary initiation to produce a Tantric Buddhist funeral rite (antyeṣṭiḥ) for initiates, in which, as in the Śaiva case (antyeṣṭidīkṣā), the officiant draws the consciousness (jñānam) of the deceased back into the corpse from the other world, takes it again through the initiatory process of consecration and the rest (abhiṣekādi) before a Maṇḍala, and then sends it out through the top of the head to ascend to liberation or a pure Buddha-field such as Sukhāvatī.69

2.2.4. Summary Beside a common drive to question standard historiographical assumptions on a sociopolitical basis, Davidson and Sanderson have more in common than one would first expect. As we have seen, they both interpret the early medieval political landscape as instrumental in the rise of Buddhist Tantrism and regard the competition for royal patronage as the driving force behind religious change. Both scholars insist on similar features of early medieval India: a movement toward regionalization, the rise of numerous new dynasties and urban centres, the massive buildup of temples, and a process of integration of peripheral populations and cults. Moreover, Davidson and Sanderson regard Śaivism, especially its appeal to the new sovereigns, as the main threat as well as major source of inspiration for Buddhism. However, while Davidson insists on symbolic issues of ideology and aesthetics, Sanderson emphasizes the concrete set of ritual 68

Sanderson 2009: 124.

69

Sanderson 2009: 126–127.

114

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

and institutional services performed by Śaiva officiants; and while Davidson regards institutional Buddhist esoterism as an internalization and sacralization of the political environment, Sanderson interprets it as the product of the Buddhists’ adoption and adaptation of the Śaiva answer to this environment. Be that as it may, we can safely take Śaivism, with its appeal to princes and concomitant economic success, to be the most serious religio-political challenge to the early medieval Buddhist communities. 2.2.5. Remarks on the Chronology of Early Buddhist Tantrism Most scholars would agree that the “definitive esoteric system” (Davidson 2002: 117) had emerged by the second half of the seventh century CE. Let us recall here Wuxing’s (↓⾜) oft-quoted statement according to which “recently the Mantra Method has come to be venerated throughout the land.”70 By 674, Wuxing had secured copies of several Tantric works, such as the Subāhuparipṛcchāsūtra, the Susiddhikaratantra and the earliest known version of the MVASū. At about the same time, Yijing (⩏῕) records the use of maṇḍalas in Nālandā and seems to have searched for Tantric initiation, though in vain.71 This also corresponds to the time when Śubhakarasiṃha (637–735) and his ācārya Dharmagupta (an expert in meditation and mantra practice) sojourned in Nālandā,72 where or in the vicinity of which the MVASū is likely to have been composed/compiled sometime between 600 and 650.73 Note that in the present case also,

70

T. 431a in Hodge’s translation (1994: 65 = 2003: 11). See also Lin Li-Kouang 1935: 83, Matsunaga 1978: xvii and Davidson 2002: 118.

71

See Hodge 1994: 65 and Gray 2007: 79, n. 212. T. 2066, 7a11–12 in Gray’s translation (ibid.): “When I, Yi-jing, was at Nālandā, I repeatedly tried to enter the maṇḍala, but my hopes were in vain as I was unable to produce sufficient merit, which shattered my aspiration to propagate these extraordinary teachings.” On the translation of tan chang ቭሙ as maṇḍala, see Hodge 1994: 65 and Gray 2007: 79, n. 212.

72

Hodge 1994: 67: “[In Nālandā,] Śubhakarasiṃha was taught the mantras, mudrās, maṇḍalas and samādhis connected with the Vairocanābhisambodhi lineage by Dharmagupta and was given the initiations (abhiṣeka) by him.”

73

On the chronology and localisation of the MVASū, see Matsunaga 1978: xvii (“during the period extending from the early to the middle part of the seventh century”), de Jong 1984: 100–101 (p. 100: “it is therefore possible that this text had already been written in the first half of the seventh century;” Nālandā); Hodge 1994: 66–74 (p. 65: “it is likely that the Vairocanābhisambodhi was composed or ‘revealed’ some time around 650 AD

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

115

Xuanzang (⋞ዔ, who travelled to and in India between 629 and 645) provides the generally accepted chronology with an argumentum a silentio: as Snellgrove (2004: 184–185) has it, “there is no indication that he ever encountered any form of ritualized Buddhism that might be described as Tantric.”74 Or, to put it in other words, there is [a] difference in attitude evident between Hsüan-tsang’s dismissal of ignorant users of spells in 646 C.E. and I-ching’s personal involvement with eso75 terism during his Indian sojourn between 671–692 C.E.

Note, however, that Xuanzang’s biographer, Huili (្❧), reports that when king Harṣa Śīlāditya was travelling in Orissa he was reproached for his support of Nālandā by “Hīnayāna” monks, who argued that “the Monastery of Nālandā and its sky-flower doctrine is not different from the Kāpālika sect.”76

give or take a decade either way;” in the region between Nālandā and the Himalayan foothills). 74

Snellgrove 2004: 185: “It is possible that this was a matter of his own spontaneous choice and that he noted little of forms of Buddhism of which he disapproved, or maybe he did not recognize such religious forms as truly Buddhist. Thus, writing on the Swat Valley (Uḍḍiyāna) in the far northwest, he says that the people have great reverence for the Buddhist Dharma and believe in the Mahāyāna, listing the various schools of the Vinaya tradition that were known, but he also makes the general comment that while learning may be appreciated, there is a lack of application, and that in particular the art of spells is practiced.”

75

Davidson 2002: 118. Davidson alludes here to Da tang xi yu ji (኱၈すᇦグ) T. 2087, 882b13–14 (Beal 1983: 120) and Song gao seng zhuan (Ᏽ㧗ൔബ) T. 2061, 710b8– 711b4.

76

Gray 2007: 81, quoting Beal 2003: 159. Note also, a few lines earlier (Beal 2003: 159), these Orissan monks’ sarcastic remark: “We hear that the king has built by the side of the Nâlanda convent a Vihâra of brass, a work magnificent and admirable. But why did not your majesty construct a Kâpâlika temple, or some other building of that sort?” Note also, for a slightly later period (under the reign of Dharmapāla, about 775–812), Gray 2005: 67 (and n. 88 for references): “There is evidence that the Buddhist tantras in general, and particularly the transgressive Yoginītantras, were resisted by Buddhists adhering to the more conservative Nikāya traditions such as the Theravāda. According to Tāranātha (1575–1634), ‘In the temple of Vajrāsana there was a large silver image of Heruka and many treatises on mantra. Some of the Saindhava and Siṃhala śrāvakas said that these were composed by Māra. So they burnt these [texts] and smashed the image into pieces and used the pieces as ordinary money.’” On similar testimonies, see also Verardi 2011: 371.

116

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

That “pure/mature” Buddhist Tantrism had emerged and gained widespread acceptance by 650–670 certainly does not mean, however, that it arose at that time. First, as Strickmann informs us, full-fledged ritual systems with soteriological valuation must have been in existence by the turn of the sixth century.77 Second, the political environment regarded by Davidson and Sanderson as a necessary condition of the rise of Buddhist Tantrism may have been in place as early as 500–550. Third, the earliest extant Tantric Śaiva scripture (the NTS) might go back to 450–550,78 and Sanderson has argued that a Śaiva Tantric scriptural corpus must have been in existence by the beginning of the seventh century at the latest.79 One feels justified, then, in asking whether coherent Tantric systems were not already in existence around 550–575, i.e., at the time of Dharmakīrti’s acme if Krasser’s new chronology is to be accepted.80 77

See also Kapstein 2003: 236–237.

78

See below, §2.2.6.2, and p. 123, n. 101.

79

Sanderson 2001: 8–11: “I propose that a scriptural corpus of the kind we can find later in the Saiddhāntika scriptures must have been in existence by the beginning of the seventh century. There survive inscriptions recording the Saiddhāntika Śaiva initiation of three major kings during the second half of that century, and during its first half the Buddhist philosopher Dharmakīrti (c. 600–660) goes to the trouble of attacking the Tantric Śaiva practice of initiation as a means to liberation. These facts reveal that Tantric Śaivism of this relatively public and strongly soteriological variety was not only present in the seventh century but well established. And this implies the existence of Tantric Śaiva scriptures. For while innovation in religious practice must have preceded its scripturalization, it could not have survived without it, far less reached such prominence.” Sanderson 2001: 18: “So by the beginning of the seventh century at the latest there existed a Tantric Śaivism of the kind known from the early surviving literature […] It is quite possible that by the seventh century most of the literature available to Śaiva scholars in the tenth was already in existence.”

80

Kumārila, Dharmakīrti and Candrakīrti have long been considered, ever since Frauwallner’s influential “Landmarks in the History of Indian Logic” (1961), roughly contemporary philosophers belonging to the first half of the seventh century CE (for the state of the art before Krasser’s new chronology, see Eltschinger 2007a: 25–28, and Krasser 2012a: 583–585; see also Steinkellner 2013: I.xxix–xxx). According to Krasser, however, Bhāviveka, who can be assigned with a fair amount of certainty to 500–570 (see Krasser 2012a: 580–582), presupposes both Kumārila and Dharmakīrti. As a working hypothesis, Krasser (2012a: 587) proposes “the time of activity of Kumārila and Dharmakīrti to be the middle of the sixth century.” Hypothetical (and impopular) as it may be, Krasser’s chronology relies in my opinion on much stronger arguments than Frauwallner’s argumentum a silentio.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

117

Strickmann and Matsunaga have paid close attention to the Buddhist dhāraṇīsūtra literature that, either in the form of apocrypha mirroring Indian ritual practices or in the form of genuine translations, has been preserved down to us in the Chinese Buddhist canon. T. 1331 (“Livre de consécration”), composed or compiled in the middle of the fifth century, was “assemblé comme un guide complet de la vie religieuse, depuis l'initiation jusqu'à la mort et l'au-delà” (Strickmann 1996: 84). It contains the earliest known references to an officiant’s self-identification with the divinity as well as injunctions pertaining to consecration applying not to the crowned bodhisattva but to human initiands. While its chapter eleven deals with rebirth in pure lands, its chapter twelve concerns the elimination of defilements and liberation. As Strickmann (1996: 84) has it, these two chapters “exposent des directives sur les rites de salut.” As for T. 1336 (“Mélange de dhāraṇī”), composed during the first part of the sixth century, it betrays “un puissant courant de renouveau rituel venu de l'Inde” (Strickmann 1996: 78). According to Strickmann, this text exhibits the first known Chinese allusions to mudrās, an outline of the Buddhist homa linked with the recitation of dhāraṇīs dedicated to Avalokiteśvara, and detailed iconographic indications. But later texts, translated or compiled between 625 and 650 on the basis of Indic materials, are likely to shed light on Indian Buddhist esoterism at the turn of the seventh century. This is the case of four works translated by Zhitong (ᬛ㏻) between 627 and 649, each centring on a manifestation of Avalokiteśvara and the corresponding mantras, maṇḍalas and mudrās. T. 901 (“Collection de dhāraṇīsūtra”), compiled in China around 653–654 under the direction of the Indian monk Atikūṭa, represents, according to Strickmann (1996: 135), “une systématisation tantrique complètement aboutie,”81 with “une hiérarchie divine au complet et un programme rituel cohérent.” This text outlines a full-blown maṇḍala anticipating the garbhamaṇḍala of the MVASū, detailed indications pertaining to the abhiṣeka as well as a fire-sacrifice “grâce auquel les divinités sont invoquées, apaisées et utilisées à des fins diverses” (Strickmann 1996:

81

Note, however, Davidson 2002: 117–118: “Literature designated as ‘proto-Tantric’ […] was still the exclusive form of Buddhist esoterism through the middle of the seventh century. Examples of this are the Gilgit manuscript of the Kāraṇḍavyūha-sūtra, which was written sometime around or before 630 C.E. and the 653–54 composition and translation of the Dhāraṇīsaṃgraha of Atikūṭa.”

118

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

135).82 Worth mentioning in this connection is the KVSū, whose two Gilgit manuscripts cannot be later than 630.83 This important scripture, mainly dedicated to the worship of (the name of) Avalokiteśvara and to the Sukhāvatī, is well known as an early source of the mantra OṂ MAṆIPADME HŪṂ and provides a very interesting testimony regarding early Buddhist Tantrism and its intimate connection with Śaivism. Here, the ṣaḍakṣarī mahāvidyā(rājñī), i.e., OṂ MAṆIPADME HŪṂ (MS G1 46a1/Mette 1997: 87), the supreme essence/heart (paramahṛdaya, MS G1 1620/46a2, Mette 197: 87) of Avalokiteśvara, is said to eliminate sin (pāpakṣayakārin, MS G1 1625/48b1, Mette 1997: 92), to allow one to become an avaivartika bodhisattva (MS G1 1621/46b8, Mette 1997: 88), to free its devotees from rebirth, old age and death (jātijarāmaraṇa, MS G1 1621/46b8, Mette 1997: 88) and bring about liberation (mokṣa, MS G1 passim; anuttarasamyaksambodhi, MS G1 1624/48a6, Mette 1997: 91). The celebrated mantra is further presented as “the purification for the best liberation” (mokṣapravaraviśuddhiḥ, MS G1 1620/46a2, Mette 1997: 87), as providing one, if it is associated with its mudrās,84 with the six perfections (ṣaṭpāramitā, MS 82

On this important work, see Shinohara 2010. According to Shinohara (2010: 392), “we have in the Collected Dhāraṇī Scriptures the earliest form of esoteric Buddhist ritual that was later elaborated in texts like the Vairocanābhisaṃbodhi sūtra and the Vajraśekhara […].”

83

On the KVSū, see de Mallmann 1967: 39–47, Mette 1997 and Studholme 2002. There are actually two Gilgit manuscripts of the KVSū (G1 and G2 in Mette’s 1997: 9–11 description), but G2 consists of only one folio (2105/81a–2106/81b, see Mette 1997: 162 for a transcription).

84

Note KVSū MS G1 1638/69a6–1640/70a4, Mette 1997: 105–107 (note that 15 to 25 akṣaras are missing at the right end of each folio): “Die Geste (mudrā), jede einzelne mudrā… Die mudrā bewirken für alle Wesen das Schwinden des Bösens. Täter der fünf Todsünden (ānantarya)… […] (Wer diese mudrā) im Gedächtnis trägt, der wird diese Segnungen der guten Qualitäten… (Weltherrscher) werden. Mit den sieben Kostbarkeiten ausgestattet wird er 70 Weltalter von 10 Millionen… sein. 50 Weltalter lang ist er ein Gelehrter aller Wissenschaften… wird Versenkung (samādhi) erlangen; 35 Weltalter lang alle… mit einem… Tod werden sie sterben. Für 25 Weltalter von anziehender Schönheit (prā[sādika-])… wird er sein für alle Wesen, ganz und gar frei. 20 Weltalter… wird er ein guter Freund (kalyāṇamitra) aller Wesen sein; für 15 Weltalter allseits…; in schlechte Wiedergeburten wird er nicht fallen. Für je 1000 Weltalter… Vollkommenes Glück wird er genießen, bis ihm das höchste vollkommene Erwachen (zuteil wird)… Die 6 pāramitā wird er vollführen; wird ohne körperlichen Makel sein; mit jedem Dharma… Von dieser Art, du Sohn aus guter Familie, sind die Segnungen der guten Qualitäten der beiden mudrā… die sechssilbige Zauberformel.”

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

119

G1 1622/47a1, Mette 1997: 89), and as supplying innumerable samādhis (MS G1 1622/47a1, Mette 1997: 89). No less importantly, the Gilgit manuscript of the KVSū provides detailed iconographic prescriptions (vidhi) for the painting of a maṇḍala centred on Avalokiteśvara.85 Besides many other eschatological and soteriological results, those drawing a fouredged maṇḍala for Avalokiteśvara will be reborn as cakravartin rulers (MS G1 1604b/30b7–8, Mette 1997: 44). Interestingly enough, the KVSū anticipates—in ideology, symbolism and vocabulary—several features that are typical of “true/mature” Tantrism, such as becoming the overlord of the Sorcerers (vidyādharacakravartin, MS G1 1651/75b3, Mette 1997: 118),86 ascribing mudrās to the fundamental nature (dharmatā) of the tathāgatas (MS G1 1627/49b6, Mette 1997: 94),87 numerous occurrences of vajra as 85

After a few ritual and dietary prescriptions concerning the painter (citrakara MS G1 1650/74a3, Mette 1997: 115), the iconographic indications develop as follows (note that 15 to 25 akṣaras are missing at the right end of each folio): “Nachdem dann der Maler eine quadratische (Umriß)linie für ein maṇḍala angelegt hat, […] soweit der Nāgakönig Sāgara, mit dem zweiten der Nāgakönig Anavatapta… soll er zeichnen. Mit angelegtem, festgebundenem Amulettzauber… In der Mitte des Tuches (paṭa) sind drei Paläste darzustellen, mit himmlischen Juwelen und Edelsteinen… Im mittleren Palast ist Avalokiteśvara darzustellen, auf einem Sitz aus Rohr niedersitzend… und ohne… An der rechten Seite ist der Tathāgata Amitābha darzustellen… mit Schmuck geschmückt ist er darzustellen. An der linken Seite ist der Śākyamuni… (darzustellen). Oben ist der Bodhisattva Sarvanīvaraṇaviṣkambhin darzustellen, mit himmlischem Schmuck geschmückt… (Dem) zu Füßen ist ein Vidyādhara darzustellen, einen Löffel für das Räucherwerk haltend und mit einem Blütenkranz in der Hand; …zu Füßen des… ist auch die Puṣpadantī darzustellen. Zu Füßen des Ākāśagarbha aber… Zu Füßen des Tathāgata… ist die Hārītī darzustellen samt ihren Töchtern, (unterhalb von?) Sarvanīvaraṇaviṣkambhin… Von dieser Art habe ich, meine Liebe, von dem König über die Fürsten der Welt, dem Tathāgata, die Regel (für die Erstellung) des Tuches (paṭavidhi-) gehört: …wird die Regel des Tuches (paṭavidhi-) hören” (MS G1 1650/74a7–1652/75a3, Mette 1997: 115– 117). And a few lines lower: “Tritt einer an das Tuch (paṭa, d.h. das Kultbild) heran, wird er von den fünf Todsünden frei sein … eben diesem (werden) unmittelbar nach der Anschauung des Erhabenen die fünf Todsünden…” (MS G1 1652/76a1 and 5, Mette 1997: 119). For the maṇḍala of the Nepali tradition of the KVSū, see KVSū 296,5–26, Mette 1997: 148,1–149,20, KVSūT P259b6–260b3, and Studholme 2002: 82–83 and 68– 70.

86

See Granoff 2000: 412–419, Davidson 2002: 173–176, 194–201, 330–334 and Studholme 2002: 82–84. Note also KVSū 293,5–8/Mette 1997: 142/KVSūT P256b24: yaḥ kaścit kulaputro vā kuladuhitā vemāṃ ṣaḍakṣarīṃ mahāvidyāṃ japati sa […] vidyādharacakravartyabhiṣekaṃ pratilabhate /.

87

See below, p. 132, n. 131, for a similar statement concerning mantras.

120

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

the first element of a compound (e.g., vajramayaṃ kāyam, MS G1 1622/47a6, Mette 1997: 89; vajrajvālā nāma mudrā, MS G1 1628/49a4, Mette 1997: 9388), and princely as well as military metaphors (e.g., cakravartikula, MS G1 1624/48a1, Mette 1997: 91; parasainyapramardaka, MS G1 1624/48a3, Mette 1997: 91). And given the political dimension of Tantrism emphasized above, it is worth noticing that T. 486 (Siṃhavyūharājabodhisattvaparipṛcchāsūtra, translated by Puṇyodaya around 663), while prescribing the erection of a maṇḍala, declares that “[s]i c'est un roi qui la [= this method VE] pratique ou la fait pratiquer, cela fera disparaître les maux de son royaume.”89 As Strickmann has pointed out, we should refrain from interpreting these texts and ritual/meditative practices as scattered and incomplete elements awaiting systematization. They can only refer to meaningful, selfcontained ritual systems based on, at least from the beginning of the sixth century, soteriological expectations. These practices involve visualization techniques, maṇḍalas, mudrās, mantras, (families of) deities, initiation, and are not unlikely to have been open to political use or adaptation. As Matsunaga (1978: xiii) describes it, “in fifth century India, rituals of meditation bound to Mudrās and Mantras are known to have been gradually established.” Strickmann (1996 : 78) argues for a “puissant courant de renouveau rituel venu de l’Inde” in sixth-century China. Scanty as this evidence may be, it points to the existence of something more coherent and selfconscious than haphazard “proto-Tantric” practices, something Dharmakīrti was familiar with and at least in part disagreed with. 2.2.6. Dharmakīrti and the Early History of Tantrism 2.2.6.1. As has already been pointed out by Sanderson and Gray, the Buddhist philosopher Dharmakīrti provides a crucial milestone in the history of Tantric Śaivism. At the end of PV 2, Dharmakīrti criticises the view that initiation (dīkṣā) brings about liberation, a view that can be ascribed to Tantric Śaivism in that it involves the typically Śaiva practice of being weighed on scales (tulā) as evidence for an initiand’s elimination of impurity (pāpa). And in his PVSV, Dharmakīrti alludes to a certain class of 88

Note also KVSū 293,3–4/Mette 1997: 141/KVSūT P256b2: yaḥ kaścid imāṃ dhārayet ṣaḍakṣarīṃ mahāvidyām […] sa kulaputra vajrakāyaśarīra iti veditavyaḥ /.

89

Lin Li-Kouang 1935: 94.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

121

Śaiva Tantric scriptures whose mantras are bound to antinomian practices such as killing and copulation. But besides evidencing Śaiva Tantric practices and scriptures, Dharmakīrti’s two elaborate discussions on mantras provide plentiful testimony to his familiarity with things Tantric. First, he explicitly alludes to Tantras as mantric scriptures. Second, he ascribes mantras, mantrakalpas, maṇḍalas, mudrās, rituals and visualization practices to the Buddhists. Third, the concepts underlying his explanation of a mantra’s efficacy, such as samaya(/pratijñā) and adhiṣṭhāna, are contextually best accounted for as Tantric. Fourth, Dharmakīrti’s polemic against an allegedly natural efficacy (bhāvaśakti) of mantras, a claim which he rather oddly ascribes to the Mīmāṃsā, is well documented in early Buddhist Tantric scriptures. In short, Dharmakīrti provides conclusive evidence for the existence of Śaiva Tantrism as well as suggestive materials concerning Buddhist Tantrism. If one subscribes to Krasser’s new chronology,90 according to which the PV is more likely to belong to the period around 550 than to 630–650, then these materials may lead to a revision of the chronology of early Indian Tantrism. Dharmakīrti’s seventh-century (?) commentator Śākyabuddhi provides further interesting materials in this direction. 2.2.6.2. As Franco, Sanderson and Torella have pointed out, Dharmakīrti’s PV 2.257–266 targets a Śaiva Tantric opponent whose position can be outlined as follows.91 (1) Dharmakīrti’s claim that liberation follows upon

90

See above, p. 116, n. 80.

91

See Franco 2001: 292–293 (with Torella’s comments in n. 24, p. 293) and Sanderson 2001: 10–11, n. 7. Here is the end of Devendrabuddhi’s commentary on PV 2.269 (PVP D115a7–b1/P133b8–134a1: gaṅ gi phyir 'di ltar log pa'i śes pa daṅ sred pa daṅ sdug bsṅal gyi rgyu dag& la dbaṅ bskur ba la sogs pa'i cho ga gnod pa can ma yin źiṅ gnod pa can ma yin pa'i phyir bcom pa ma yin la bdag med pa ñid mthoṅ ba gnod pa can yin pa de'i phyir de goms par gyur pa las yaṅ srid pa yod pa ma yin no //. &The Tibetan translation suggests a dvandva compound (mithyājñāna-tṛṣṇā-duḥkhahetu); however, I am inclined to interpret the relation between mithyājñānatṛṣṇā and duḥkhahetu as a karmadhāraya. “Since, in this way, a ritual procedure (vidhi) such as initiation does not annul (*bādhaka) the two causes of suffering, viz. erroneous cognition and craving and, because it does not annul [them, these] are not overcome, [it is] the perception of selflessness (nairātmyadarśana) [that] annuls [them]; therefore, rebirth (punarbhava) ceases to exist (*nāsti) thanks to the repeated practice of this [perception] (*abhyāsa, *sātmībhāva).” This suggests that Dharmakīrti’s polemic against his Tantric opponent extends up to PV 2.269, although I fail to see how PV 2.267–269 can be satisfactorily

122

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

the destruction of the personalistic false view (satkāyadṛṣṭi, as it can be seen, e.g., in PV 2.246cd92) is unsatisfactory; liberation can only be achieved through initiation (dīkṣā, the only path to/method for liberation, *mokṣamārga) as it has been revealed by Īśvara(/Śiva).93 (2) Whoever has undergone initiation (dīkṣita) is liberated,94 viz. is not reborn any longer,95 like a seed (bīja) that is made sterile or brought to lose its “seedness” by ritual procedures such as anointment (tailābhyaṅga) and burning (agnidāha).96 (3) Scales-initiation (tulādīkṣā) provides empirical evidence for connected to PV 2.257–266. On PV 2.267–269, see Vetter 1990: 159–161 and Eltschinger/Ratié 2013: 173–186. 92

PV 2.246cd: tasmāt tatkāraṇābādhī vidhis taṃ bādhate katham //. “Therefore how [can] a [soteriological] injunction [prescribing the cultivation of the thought of pain] annul this [attachment to what supposedly belongs to the self if, as it is the case, it] does not annul its cause [i.e., the vision of the self]?” Translation Eltschinger/Ratié 2013: 242– 262. PV 2.246cd is quoted in PVP D110b2/P128a3–4 and PVV 98,18 ad PV 2.257. On the satkāyadṛṣṭi in Dharmakīrti’s writings, see Vetter 1990: 22–26, Franco 2001: 289– 300, Eltschinger 2009: 170–175, and below, Chapter 4, §4.3.

93

See PVP D110b1–2/P128a3–5 and PVV 98,17–18.

94

See PVP D110b2–3/P128a5–6.

95

The dīkṣā is described as mi skye ba'i rgyu in PVP D110b6/P128b1–2, as yaṅ srid par byed pa ñid mtshams sbyor ba'i rnam par gags byed pa in PVP D111a5–6/P129a2; the dīkṣita is described as rab tu skye ba 'byuṅ ba rgyun chad pa can in PVP D110b2– 3/P128a5 and PVP D110b5/P128a7.

96

See PVP D110b7–111a1/P128b3–4 (together with Vibh. 99, n. 1), PVV 99,4–5, PVP D110b3/P128a5 (*mantrābhisaṃskṛtaṃ bījam) and PVV 98,25 (dīkṣāvidhispṛṣṭaṃ bījam). The ritual alluded to here might correspond to the one described by Abhinavagupta in TĀ 20.2–6: trikoṇe vahnisadane vahnivarṇojjvale 'bhitaḥ / vāyavyapuranirdhūte kare savye sujājvale // bījaṃ kiñcid gṛhītvaitat tathaiva hṛdayāntare / kare ca dahyamānaṃ sac cintayet tajjapaikayuk // vahnidīpitaphaṭkāraghoraṇīdāhapīḍitam / bījaṃ nirbījatām eti svasūtikaraṇākṣamam // taptaṃ naitat prarohāya tenaiva pratyayena tu / malamāyākhyakarmāṇi mantradhyānakriyābalāt // dagdhāni na svakāryāya nirbījapratyayaṃ tv imam / sa śrīmān suprasanno me śambhunātho nyarūpayat //. “Il maestro, tutto inteso a recitare i convenienti mantra, deve meditare la sua mano destra in forma di un triangolo di fuoco, tutto fiammeggiante, splendente dei fonemi del fuoco, attizzato da una ruota di venti, mettervi un seme qualsiasi e meditare come esso, nella mano e dentro il cuore, stia bruciandosi. Il seme, in tal guisa, consumato dal calore di una moltitudine di PHAṬ, vivificati dai fuochi, perde ogni qualità seminale, incapace ormai di generare. Questo seme, arso che sia, non è più in grado di svilupparsi. Allo stesso tempo, in virtù di questa stessa dimostrazione, sono bruciate, in forza dei mantra, della meditazione e dell’azione rituale, anche le tre maculazioni—minimale, māyica e karmica,—nel senso che non sono più in grado di fornire i loro prodotti. Quest’ini-

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

123

(and increases one’s faith—pratyaya – in) the fact that the initiand has eliminated impurity and hence is liberated, for impurity has weight; this in turn provides evidence for the scripture’s authoritativeness.97 (4) The dīkṣā annihilates the adṛṣṭa, i.e., merit and demerit (dharmādharma), which consists in a karmic disposition/residual trace98 (saṃskāra) in the soul (ātman) and is responsible for the organs (karaṇa) allowing one to move and to cognize.99 According to Sanderson and Torella, the tulādīkṣā is characteristically Śaiva: Those present, we are asked to believe, see that the person’s weight is greater before than after, and are thereby made to understand that the initiation has succeeded in destroying the Impurity (malaḥ) that impedes the soul’s omniscience and omnipotence, that being of the nature of a material substance (dravyam), 100 albeit imperceptible.

This practice is attested in the NTS (450–550?), maybe the earliest among the Śaiva Saiddhāntika scriptures (although the NTS, which is very close to the Atimārga/Pāśupata milieu, claims to belong to the Mantramārga and does not allude to the Siddhānta at all):101 He may purify [the retributive effects of] his own or others’ bad deeds [in such a way that their removal may be known to have taken place] by means of a set of 102 scales.

ziazione che ha come dimostrazione la combustione del seme mi è stata descritta dal venerabile Śambhunātha, nella sua benevolenza.” Translation Gnoli 1999: 458. 97

See PVP D111a2–3/P128b5–6 (where homa seems to be involved in the process) and PVV 99,7–8.

98

Or, more precisely, by the dispositions/residual traces left by twenty kinds of good and bad actions (PVP D112a1–2/P129b8 [dge ba daṅ mi dge ba rnam pa ñi śu 'jug pas skyed pa'i 'du byed], Vibh. 100, n. 1 [viṃśati kuśalākuśalāḥ]). Note also PVV 100,11: tad adṛṣṭaṃ cātmasaṃskāraḥ, and PVV 100,5–6: pūrvaśubhāśubhakarmasaṃskāraḥ.

99

See PVP D111b6–7/P129b5–6 (together with PVV 100,7). These karaṇas are reminiscent of the Sāṅkhya’s division of organs (karaṇa) into buddhīndriyas and karmendriyas.

100

Sanderson 2001: 11, n. 7. On the tulādīkṣā, see also TAK III.113, s.v. tulādīkṣā.

101

On the NTS, see Sanderson 2001: 29–31, n. 32, Sanderson 2006 and Goodall/Isaacson 2007. Note that the NTS seems never to allude to “the innate impurity (mala) that is held to cling to every soul, removable only by initiation” (Goodall/Isaacson 2007: 5 b).

102

NTS mūlasūtra 7.15ab: tulayā śodhayet pāpam ātmanasya parasya vā /. I am very grateful to Dominic Goodall for providing me with a provisional edition of NTS,

124

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

That the scales-initiation, which forms the subject-matter of the twentieth chapter of Abhinavagupta’s TĀ, was aimed at providing empirical evidence of an initiate’s immaculate condition and liberation is made clear in the following statement of Sadyojyotis: By initiation on the scales he is purified even of such great sins as brahmanicide. One may know that his bonds have been destroyed through the evidence [of one’s senses], just as [the validity of the Gāruḍatantras is proved] by the cessation of 103 poison [brought about by their Mantras].

Dharmakīrti’s critique of this early Śaiva soteric practice needs not detain us here.104 mūlasūtra 7 as well as a provisional note on and translation of NTS mūlasūtra 7.15ab (personal communication, December 26, 2010). Thanks are also due to Diwakar Acharya who first drew my attention to this passage (personal communication, February 14, 2010). According to Goodall/Isaacson 2007: 6a, “the Mūlasūtra must be the most ancient core of the text.” 103

TSa 36 (quoted and translated in Sanderson 2001: 11, n. 7): śuddhiṃ vrajati tulāyāṃ dīkṣāto brahmahatyato mukhyāt / pratyayato jānīyād bandhanavigamaṃ viṣakṣayavat //. Abhinavagupta also makes clear that the tulādīkṣā aims at providing (dull-witted persons with) evidence, e.g., in TĀ 20.1: atha dīkṣāṃ bruve mūḍhajanāśvāsapradāyinīm / (“Diremo adesso dell’iniziazione che infonde fiducia a individui offuscati.” Translation Gnoli 1999: 458), TĀ 20.11cd–12ab: tulāśuddhiparīkṣāṃ vā kuryāt pratyayayoginīm // yathā śrītantrasadbhāve kathitā parameśinā / (“Il maestro, volendo, può sottoporre il discepolo all’iniziazione che somministra l’eliminazione del peso, dotata anch’essa di una dimostrazione sensibile. Essa è stata esposta dal Signore nel Tantrasadbhāva […].” Translation Gnoli 1999: 459), TĀ 20.16ab: uktā seyaṃ tulāśuddhidīkṣā pratyayadāyinī / (“E con ciò abbiamo esposto l’iniziazione che elimina il peso, fornita d’una dimostrazione immediata.” Translation Gnoli 1999: 460). On the scriptural sources of Abhinavagupta and Jayaratha, see Sanderson 2001: 11, n. 7.

104

The first three stanzas of Dharmakīrti’s answer are as follows (PV 2.257–259): āgamasya tathābhāvanibandhanam apaśyatām / muktim āgamamātreṇa vadan na paritoṣakṛt // nālaṃ bījādisaṃsiddho vidhiḥ puṃsām ajanmane / tailābhyaṅgāgnidāhāder api muktiprasaṅgataḥ // prāg guror lāghavāt paścān na pāpaharaṇaṃ kṛtam / mā bhūd gauravam evāsya na pāpaṃ gurv amūrtitaḥ //. “He who, on the sole [authority of a] scripture [supposedly laid down by Īśvara], claims [that] liberation [is achieved through initiation] does not satisfy those [practically rational persons aiming at liberation] who fail to discern the reason why the [said] scripture is of that kind [i.e., reliable with regard to that which should be put into practice].& A [ritual] procedure (vidhi) known [to hinder the fruition] of a seed, etc., is not able to stop the human beings’ [re]birth, for [if it were the case, their] liberation [from saṃsāra] would follow from [procedures] such as anointment, burning [or pulverization, all of which are known to make a seed sterile]. It

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

125

2.2.6.3. In addition to bearing witness to early Śaiva conceptions and practices centred on initiation as the means towards liberation, Dharmakīrti’s allusion to the Bhaginītantras provides important evidence for the existence, some time during the early sixth century, of Śaiva scriptures associated with the cult of the four sisters (Jayā, Vijayā, Jayantī/Ajitā, Aparājitā) of the god (and aspect of Śiva) Tumburu. This cult, which is also reflected in a Gilgit manuscript that cannot be later than 550,105 “was one of is [certainly] not because [the one who was] heavy before [his initiation] has become lighter [after it, once he is weighed on scales], that [initiation] is [to be held] responsible for the destruction of [his] sins. Let him [simply] not be heavy [any longer]! ' [Indeed,] sin is not heavy, for it is incorporeal.” &Translated according to PVP D110b4–5/P128a7– 8 and PVṬ Ñe D145a2/P179a5. 'PVV 99,13 explains mā bhūt as kasmān na kalpyate. The point of Dharmakīrti’s polemic is to demonstrate that, contrary to the intuition and cultivation of selflessness (nairātmyadarśana and nairātmyabhāvanā), initiation cannot destroy the causes of rebirth (the personalistic false view and craving) and hence is unable to bring about liberation. Dharmakīrti spells this out in three stanzas. First, in PV 2.260–261: mithyājñānatadudbhūtatarṣasañcetanāvaśāt / hīnasthānagatir janma tatas tacchin na jāyate // tayor eva hi sāmarthyaṃ jātau tanmātrabhāvataḥ / te cetane svayaṃ karmety akhaṇḍaṃ janmakāraṇam //. “[Re]birth [i.e.,] moving to a vile place, & [is] due to the volitions (sañcetanā) associated with erroneous cognition [i.e., nescience,] and craving occasioned by this [erroneous cognition];' therefore, the one who cuts them off is not [re]born. Indeed, only these two( are capable of [causing re]birth, because [one observes, through co-presence and co-absence, that rebirth] occurs due to them only [i.e., occurs in their presence and lacks in their absence. And] since the two volitions[, which do still exist in the initiand,] are themselves [nothing but] karman,) the causes of [re]birth are complete [and can only bring about their effect for the initiand also]. *” &See PVP D111a6/P129a3–4 and PVṬ Ñe D145a3–4/P179a6–7. 'See PVP D111a7– b1/P129a4–5, PVṬ Ñe D145a5/P179b1 (which refers to PV 2.81) and PVV 99,19–20. ( PVP D111b3/P129a8 clearly interprets tayoḥ as cetanayoḥ (ji skad du bśad pa’i sems pa dag go), whereas PVV 99,24 interprets tayoḥ as ajñānatṛṣṇayoḥ; note PVV 99,25: tau ca dīkṣitasya sta iti sa jāyate. )See PVP D111b4–5/P129b1–3, PVṬ Ñe D145a6– 7/P179b2–3 and PVV 100,5–6. *See PVP D111b5–6/P129b3–5. Second, in PV 2.266: nirhrāsātiśayāt puṣṭau pratipakṣasvapakṣayoḥ / doṣāḥ svabījasantānā dīkṣite ’py anivāritāḥ //. “Even in an initiate, [moral] faults whose continuum [proceeds] from self[-homogenous] seeds are not removed, because they diminish or increase with the growth[, respectively,] of [their] antidote (pratipakṣa)[, which consists in the intuition of selflessness and not in initiation,] or [their] ally (svapakṣa)[, which consists in erroneous cognition and craving and not in non-initiation].&” &See PVP D113b5–6/P131b8–132a2 and PVV 101,10–11. For references to Kṣemendra’s rebuttal of Dharmakīrti’s arguments, see Sanderson 2001: 10, n. 7. 105

On the first two folios of the *Devītantrasadbhāvasāra, see Sanderson 2009: 50. On the god Tumburu, see Griffiths 2004–2005.

126

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

the earliest, perhaps the earliest, of the esoteric Śaiva systems.”106 According to Sanderson, [t]he text teaches the Mantras of the four Devīs, who, it says, were made manifest at the beginning of creation so that men could attain supernatural accomplish107 ments and liberation.

Together with these two folios, Dharmakīrti’s PVSV testifies to the existence, parallel to the early texts of the Śaivasiddhānta and the Mantrapīṭha (the Śaiva scriptures dedicated to the cult of the ferocious deity Bhairava), of scriptures pertaining to forms of the Goddess considered as manifestations of Śiva’s śakti (and, more precisely, belonging to the vāmasrotaḥ of the Vidyāpīṭha). In other words, the PVSV points to an early phase of Śāktism, a current embedded “in the intensely transgressive tradition of Kāpālika asceticism.”108 Let us now consider the two relevant passages of the PVSV. The first one, translated here with part of Dharmakīrti’s polemic against the natural efficacy of mantras, runs as follows: The series [of phonemes called a “mantra”] is not active naturally[, but on account of a human being’s pledge], because of the quick success [of a mantra] for a certain [person, of its being] long for another, [and] because of [yet] another [person]’s need of [something] like the practice of an observance [or fire oblation in order to secure the result of the mantra. Further, a mantra yields its results naturally, first] because one observes that even though [they perform] one [and the same ritual] action [pertaining to the mantra, such as japa and homa], one [person] profits from it and another not, [and second] because mantras, though [once] effective (vahat; Tib. bya ba byed pa), [can be] deceiving with regard to a certain [result when the same person uses them] again; [and] this feature indeed cannot belong to [the mantra’s] own nature, for [its] own nature is the same in every [case]. On the contrary, it is possible that, acting [as he does] out of his own wish, a human being [issuing a pledge] favours a certain [person and] not another either due to [their] community (sattvasabhāgatā) [with regard to morality, etc.] or on account of [this person’s] special devotion [to a certain divinity through japa, homa, etc.]. [Objection:] The success [of a mantra proceeds] naturally in the case 106

Sanderson 2009: 129, n. 301.

107

Sanderson 2009: 50–51, n. 22.

108

Sanderson 2009: 49.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

127

of [one’s] having accumulated merit due to the practice of an observance, etc., or for [someone] virtuous, and [its] non-success [proceeds] naturally in the case of [one’s] having accumulated demerit due to the break of an observance, etc., or for [someone] who is not virtuous. [Answer:] No, because in [scriptures] such as the Ḍākinītantra and the Bhaginītantra, one also encounters observances that are contrary to virtue, abundant in cruelty, robbery, sexual intercourse, vile actions, 109 etc., and because these [observances, immoral as they are, are responsible for] 110 the specific success [of the mantras to which they are bound].

According to Dharmakīrti, scriptures of mantric ritual (mantrakalpa) such as the Ḍākinītantra and the Bhaginītantra prescribe observances (vrata), i.e., behavioural and ethical restrictions (niyama) such as violence, robbery, sexual congress and other vile actions. Śākyabuddhi and Karṇakagomin shed more light on this issue: In the Ḍākinītantras, the rule of post-initiatory discipline is that one attains the power of the Mantra if one kills and devours a living creature. In the Kambukinītantras one’s attainment is said to come about through the practice of robbery. And somewhere in the Bhaginītantras the post-initiatory rule is that a certain 111 goddess bestows the attainment on one if one practises sexual intercourse.

That the post-initiatory rules (samaya) referred to in these Tantras are nonBuddhist is made clear by another passage in the PVSV:

109

Note the following explantion (PVṬ Ñe D39b4–5/P45a2–3 = PVSVṬ 578,10–11): krauryaṃ prāṇivadhaḥ / steyaṃ cauryam / dvīndriyasamāpattir maithunam / hīnakarma mārjārāśucipradānādi / ādiśabdād anyasyāpi dharmaviruddhasya grahaṇam /. “Cruelty is killing; robbery is theft; the coming together of the two [sexual organs] is sexual intercourse; vile action is, e.g., offering the excrement of a cat (?); the word ‘etc.’ includes any other [practice] that is contradictory to dharma.”

110

PVSV 162,22–163,5: na cāyam anukramaḥ svabhāvataḥ kārakaḥ, kasyacid āśusiddher anyasya cirād aparasya vratacaraṇādyapekṣaṇāt / ekasmād api karmaṇaḥ kayościd arthānarthasandarśanāt / vahatām api mantrāṇāṃ punaḥ kvacid visaṃvādāt / na hy ayaṃ prakāraḥ svabhāve yuktaḥ / vratacaryābhraṃśādinā dharmādharmopacaye dharmādharmātmanor vā prakṛtyā siddhyasiddhīti cet / na / dharmaviruddhānām api krauryasteyamaithunahīnakarmādibahulānāṃ vratānāṃ ḍākinībhaginītantrādiṣu darśanāt / taiś ca siddhiviśeṣāt /.

111

PVṬ Ñe D39b6–7/P45a4–6 ≈ PVSVṬ 578,13–16: tathā hi ḍākinītantre samayavyavasthā / yadā prāṇinaṃ hatvā khādati tadā mantrasiddhim āsādayati / tathā kambukinītantre steyācaraṇāt siddhir uktā / tathā maithunācaraṇāt siddhipradā kācid devateti bhaginītantrāntare kvacit samayaḥ /. Translation Sanderson 2001: 10, n. 10.

128

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

[Thus it is that] in two [different] scriptures of mantric ritual, the one Buddhist and the other [non-Buddhist, practices] such as violence, sexual intercourse and the contemplation of the Self are presented [respectively] as causes of misfortune 112 and otherwise [i.e., as causes of prosperity].

Dharmakīrti’s earliest work, then, points to non-Buddhist, Śākta Śaiva cults involving strongly antinomian/transgressive practices supposed to ensure mantric attainments. Here as elsewhere, the little we know about these Tantras we owe to the work of Sanderson. The Kambukinītantras (“Tantras of the Robber Wives”113) are not otherwise recorded. No Ḍākinītantras have survived down to us, but allusions to them by Kṣemarāja, Somadeva, Bhāsarvajña and maybe even an eighth-century Arabic witness corroborate Śākyabuddhi’s and Karṇakagomin’s account by pointing to human sacrifices and the ways in which Yoginīs kill their victims.114 As for the Bhaginītantras, they are, in content if not in name, better attested in surviving Śaiva literature, for diverse sources refer to individual Tantras dedicated to each of Tumburu’s sisters (Jayātantra, Vijayātantra, Ajitātantra and Aparājitātantra) as well as to a bhāgineyaṃ tantram.115 2.2.6.4. When gathered together and placed into a systematic conceptual network, the remarks scattered throughout Dharmakīrti’s PVSV provide an interesting picture of sixth-century Tantrism.116 Dharmakīrti alludes to maṇḍalas, which Śākyabuddhi and Karṇakagomin describe as “a specific [graphical] arrangement of deities, etc.,” to mudrās, defined as “figures of the hand fingers,” and to dhyāna, presented as “a thought [concentrated] on 112

PVSV 123,22–24: bauddhetarayor mantrakalpayor hiṃsāmaithunātmadarśanādayo 'nabhyudayahetavo 'nyathā ca varṇyante /. See also Davidson 1981: 8, Sanderson 2001: 10, n. 10, Gray 2007: 78, Eltschinger 2007a: 301–302.

113

Note that PVṬ Ñe D39b3/P44b8–45a1 and PVṬ Ñe D39b6–7/P45a4–6 read duṅ can ma('i rgyud) for kambukinī(tantra), i.e., “Tantra of the conch/shell possessing one(s)” (Tib. duṅ, which has the same meaning as Skt. kambuka, may according to Jäschke 253a s.v. also mean “skull”).

114

For further arguments in favour of the close association of ḍākinīs with Śiva, see Gray 2005: 51–52.

115

For texts and references, see Sanderson 2001: 10–11, n. 10. Note also PVṬ Ñe D39b3/P44b8–45a1 ≈ PVSVṬ 578,7–9: ḍākinībhaginītantrādiṣu darśanāt / ḍākinītantre catur&bhaginītantre' / ādiśabdāc cauryahetuṣu kambukinītantrādiṣu darśanāt /. &PVṬ with no equivalent of catur°. 'PVṬ reads: *tantra iti dvandvaḥ.

116

For a more detailed account, see Eltschinger 2001 and 2008a.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

129

a [deity] such as Garuḍa,” or as “a thought [concentrated] on the form of deities, etc.”117 And Dharmakīrti’s detailed discussion of mantras represents a unique attempt at understanding their nature and mysterious efficacy. According to his own definitions, a mantra consists in an utterance [that has been pronounced] by [persons] who are endowed with a [special] power [they owe to their own] truthfulness and austerity [and that is] able to realize the desired purpose,

or in “the utterance and the pledge(/covenant) of [persons] who are endowed with truthfulness, etc.”118 Far from owing their efficacy to their intrinsic nature or authorlessness, mantras yield their results through the agency of an exceptional person who empowers (adhiṣṭhā) them and prescribes, in a pledge/covenant or a promise (pratijñā), the ritual, psychological, behavioural and ethical conditions to be fulfilled on the part of the user. This person’s exceptionality lies in his cognitive and psychological qualities (perception of the supersensible [atīndriyadarśana], knowledge [jñāna], power [prabhāva], etc.) and above all in his possessing “causes for [producing] mantras” (mantrahetu, such as truth[fulness] [satya], austerities [tapas], recollection [smṛti], cognition [mati], intellectual penetration [prativedha], capacity [śakti]) or “means of producing mantras” (mantrakriyāsādhana, such as gativiśeṣa, “a specific transmigrational status,” or siddhiviśeṣa, “the particular success [of a mantra]”) which, in the commentators’ explanations, often amount to Buddhist abhijñās (“superknowledges,” with the obvious exception of āsravakṣayajñāna, the “knowledge of the destruction of the outflows,” which since it is practically identical to nirvāṇa cannot characterize non-Buddhists). On the side of this superior person, the samaya is a pledge:

117

Respectively PVṬ Je D282a2/P338b5 ≈ PVSVṬ 450,27: maṇḍalaṃ devatādiracanādi & viśeṣaḥ /. &PVṬ with no equivalent of °ādi°. PVSVṬ 450,27: pāṇyaṅgulasanniveśo mudrā / (to be compared with PVṬ Je D282a1–2/P338b4–5: phyag rgya ni srog chags dag gi sor mo rnams kyi dbyibs so //). PVṬ Je D282a2/P338b5: bsam gtan ni nam mkha' ldiṅ la sogs pa sems pa (to be compared with PVSVṬ 450: 28: dhyānaṃ devatādirūpacintanam).

118

Respectively adapted from PVSV 123,15–17 (na mantro nāmānyad eva kiñcit / kiṃ tarhi satyatapaḥprabhāvavatāṃ samīhitārthasādhanaṃ vacanam /) and PVSV 124,14– 15 (na mantro nāmānyad eva kiñcit satyādimatāṃ vacanasamayād iti /). See also PVṬ Je D283a3–4/P340a4–5 ≈ PVSVṬ 452,17–19.

130

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

I bestow this result (artha) on the [person] who makes use, in that [particular] way, of the sentence I have composed, [a sentence] which[, provided these conditions are instantiated,] is conducive to the [said] desired result. 119

On the side of the user, the samaya consists in a post-initiatory rule, an ethical and behavioural observance, or a (self-)restriction in practice (anuṣṭhāna, ācaraṇa, etc.) that ensures mantric attainment. As we have seen, these observances may be transgressive and perceived as utterly immoral (kraurya, steya, maithuna, etc.), but may also be, at least from a Buddhist standpoint, of a moral character (śīlācāra, maitrī, śauca, etc.). The prescriptions laid down by the mantrakartṛ also entail more ritually oriented components: they may pertain to the recitation itself (pāṭha, mantraviṣayajapa), to fire-sacrifice (homa), or to the invocation (āvāhana) and release (visarjana) of the deity (devatā) involved in the ritual. They pertain to a particular type of worship (sevā) that aims at propitiating (ārādhana) or satisfying (toṣa) the deity associated with the mantra, or the author of the mantra himself. Note, however, that Śākyabuddhi and Karṇakagomin introduce a strong hierarchy between the samaya as a post-initiatory rule and the other kinds of instructions: Here, the samaya is that the transgression of which implies that one must enter the maṇḍala again, etc.; [every] prescription other than this [samaya] is an instruction, [and] such is the distinction between the two.120

Of decisive importance for the ritual and devotional aspects of mantric practice are the two concepts of adhimukti and upayoga, whose range of meaning encompasses inclination, faith, confidence and devotion.121 As Dharmakīrti himself makes clear, the user’s inclination or devotion should be directed toward his master’s ākāra, svabhāva and caryā. According to

119

PVSV 155,21–23: matpraṇītam etad abhimatārthopanibandhanaṃ vākyam evaṃ niyuñjānam anenārthena yojayāmi […] /. See also PVṬ Je D282b2–4/P339a8–b2 ≈ PVSVṬ 451,18–20: ya imāṃ varṇapadaracanām abhyasyati tadvidhiṃ cānutiṣṭhati tasyāhaṃ yathāpratijñātam arthaṃ sampādayiṣyāmi /. “I shall endow with the result as [it has been] promised [by me] the [person] who repeats this arrangement of phonemes and words and who puts its prescriptions into practice.”

120

PVṬ Ñe D43a1–2/P48b8–49a1 = PVSVṬ 581,26–27: tatra samayo yasyātikramāt punar maṇḍalapraveśādiḥ kartavyo jāyate / tato 'nyad vidhānam upadeśa ity anayor bhedaḥ /.

121

On adhimukti/adhimokṣa, see BHSD 14ab and 15a s.v. and Schmithausen 1969: 179– 180.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

131

Śākyabuddhi and Karṇakagomin, ākāra or appearance should be understood in the sense of the master’s colour and shape (varṇasaṃsthāna); svabhāva or character, in the sense of this master’s being of a terrifying (raudra) or peaceful (śānta) nature; as for caryā or behaviour, it consists in the outward activity defined as a vocal and corporeal operation (kāyavāgvyāpāralakṣaṇā ceṣṭā).122 In other words, the user must concentrate his mental flux on the image of his master, a psychological disposition that is perfectly accounted for in the devotional context by the term dhyāna, used by Dharmakīrti in other passages of his PVSV.123 Last but not least, both the post-initiatory rules and the set of ritual prescriptions are laid down in ritual texts known to Dharmakīrti as mantrakalpas and Tantras. Our texts define Tantras as treatises connected with mantras (mantrapratibaddhāni śāstrāṇi),124 or simply as treatises on mantras (mantraśāstra). As for the mantrakalpas, they are characterized as instructions on the injunctions that ensure the realization of mantras (mantrasādhanavidhānopadeśa).125 Inasmuch as he subjects the use of his mantra to various prescriptions and limitations, the author of a mantra is also the author of a Tantra (tantrapraṇetṛ) and serves then as a master (prabhu, svāmin) to those who know (and follow) his Tantra (tantravid). 2.2.6.5. Dharmakīrti obviously did not mean to elaborate upon mantras in a systematic way. As I said above, he rather embedded scattered remarks in a complex network of arguments that all aim at demonstrating, against the Mīmāṃsā, that the mantras’ efficacy can only be accounted for on the basis of a human origin and a controlling power/empowerment. To his Mīmāṃsaka opponent, Dharmakīrti repeatedly ascribes the idea that mantras yield their results due to a natural capacity (bhāvaśakti),126 an idea that, to the

122

See PVṬ Ñe D41b5–6/P47b1–2 = PVSVṬ 580,20–21.

123

See PVSV 163,25–164,1 and above, p. 129, n. 117.

124

See PVṬ Ñe D42b1/P48a6 ≈ PVSVṬ 581,7.

125

PVSVṬ 449,25–26, to be compared with PVṬ Je D281a3–4/P337b1–2.

126

See PV 1.292, PVSV 156,14, PV 1.294, PVSV 156,25. According to PVṬ Ñe D21a7/P23b3 = PVSVṬ 553,27, bhāvaśaktiḥ is to be explained as śabdasvabhāvasyaiva śaktiḥ. PVSV 155,27–28 provides a useful summary of the (pseudo-)Mīmāṃsaka position: na vai puruṣasamayān mantrebhyo ’rthasiddhiḥ / kiṃ tarhi / bhāvasvabhāva eṣa yad ime kathañcin niyuktāḥ phaladāḥ /. “The realization of the [desired] purpose through mantras does not originate from a human pledge[, as you claim]; rather, it is the

132

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

best of my knowledge, cannot be found in the extant Mīmāṃsā treatises.127 Whatever the case may be, Dharmakīrti’s ascription perfectly fits the context of his polemics against the Mīmāṃsaka’s authorlessness (apauruṣeyatā) of the Veda. As is well known, this school holds human beings to be incapable of any perception of supersensible states of affairs, let alone of omniscience, which implies in turn that no human being could ever have perceived Dharma, the object of Vedic injunctions.128 Now at least in Dharmakīrti’s mind, if the Veda is authorless, then its mantras must also be such,129 which amounts to saying that they must be endowed with an ability of their own, i.e., with a “natural” or authorless capacity. But I feel justified in doubting Dharmakīrti’s repeated ascription of the bhāvaśakti-hypothesis to the Mīmāṃsā school, or to it only, for, to the best of my knowledge, the extant Mīmāṃsaka sources do not allude to it. Although the term bhāvaśakti appears in Bhartṛhari’s works and may echo medical conceptions, the only sources I am aware of are later Buddhist Tantric texts such as the Sādhanamālā. Note, however, that some much earlier Buddhist Tantric works reflect an understanding of mantras which fits Dharmakīrti’s bhāvaśakti well. The seventh-century STTS repeatedly describes mantras as prakṛtisiddha, “established naturally,”130 which is indeed close to the MVASū’s claim that no buddha ever produced, approved of, or had the mantralakṣaṇa produced. A little later, the same text resorts to the concept of a mantra’s mantradharmatā, conceiving it along the same lines as the traditional Buddhist tenet that dharmatā is eternal and independent of the presence or absence of tathāgatas in the world.131 I deem it plausible, then, very nature of the entity [i.e., of the utterance itself,] that these [mantras] yield results when used in a certain way.” 127

On the Mīmāṃsā’s account of mantras, see Taber 1989. On a tentative explanation of Dharmakīrti’s background concerning the bhāvaśakti, see Eltschinger 2001: 83–90 and 120–122.

128

See Kataoka 2011: II.60–98; see also above, §1.2.8, and below, §3.2.3.7.

129

Dharmakīrti never alludes to the distinction drawn by the Mīmāṃsaka hermeneutics between mantra and codanā (vidhi/pratiṣedha), or between mantra, nāmadheya, arthavāda and vidhi, which the Mīmāṃsaka could resort to in order to neutralize Dharmakīrti’s interpretation. On this, see Taber 2012.

130

STTS no. 22 (Horiuchi 1974: [I.]25) and no. 28 (Horiuchi 1974: [I.]28).

131

MVASū Tha P134b2–3: gsaṅ ba pa’i bdag po gźan yaṅ gsaṅ sṅags rnams kyi mtshan ñid ni saṅs rgyas thams cad kyis ma byas / byed du ma bcug / rjes su yi raṅ ba ma yin no // de ci'i phyir źe na / 'di ni chos rnams kyi chos ñid de / 'di ltar de bźin gśegs pa

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

133

that Dharmakīrti’s polemics against the Mīmāṃsā on the issue of a mantra’s bhāvaśakti also (if not primarily) mirrors his concern with a conception that was creeping into Buddhist monasteries under the growing influence of Tantric teachers, or, at least, was typical of early Śaiva Tantrism. This may in turn explain why Dharmakīrti’s position and proofstrategy in time disappeared from the Buddhist epistemologists’ treatment of omniscience and related matters: while the sixth-century Dharmakīrti was writing in a not yet fully trantricized environment, his seventh- to eighth-century epigones might have adopted a different attitude due to the institutional strength of Buddhist esoterism, if not on account of their own personal Tantric commitment.132 2.2.6.6. As we can see, Dharmakīrti represents a decisive milestone in the history of Tantrism. He is familiar with the Śaiva scales-initiation, a procedure supposed to bring about liberation from saṃsāra and provide empirical evidence for it. He knows of the cult of Tumburu’s four sisters, of various transgressive observances bound to mantric rituals, and of diverse Śaiva Tantras, some of which at least bear witness to early Śākta Śaiva scriptures and practices.133 Dharmakīrti’s views on mantras and their efficacy betray his familiarity with Tantric rituals involving pledge-cum-

rnams byuṅ yaṅ ruṅ ma byuṅ yaṅ ruṅ / chos rnams kyi chos ñid de ni ye nas gnas pa ste / de yaṅ 'di ltar gsaṅ sṅags rnams kyi gsaṅ sṅags kyi chos ñid do //. The Sanskrit original of this passage is quoted in the CMP (33,3–6): api tu (sic) guhyakādhipate mantrāṇāṃ lakṣaṇaṃ sarvabuddhair na kṛtaṃ na kāritaṃ nānumoditam / tat kasya hetoḥ / eṣā dharmāṇāṃ dharmatā yad utotpādād vā tathāgatānām anutpādād vā tathāgatānāṃ sthitaivaiṣāṃ dharmāṇāṃ dharmatā yad uta mantrāṇāṃ mantradharmatā /. “Moreover, Lord of the Secret Ones, the properties of mantras are not created by all the Buddhas, nor caused to be created by them, and nor are they sanctioned by them. Why is that? Because this is the intrinsic nature of phenomena, for whether Tathāgatas appear or not, their intrinsic nature has always been present, and so that is also the intrinsic nature of the mantras of mantrins.” Translation Hodge 2003: 131. See also above, p. 119, n. 87, and below, p. 141, n. 152. 132

See Eltschinger 2001: 123–126.

133

Note also Bhāviveka’s familiarity with the goddess Cāmuṇḍā, whose cult he deems abhorrent on account of its being “replete with coarse and unbearable affliction” (TJ D184b2/P200a1: bzod par dka' ba'i ñon moṅs pa bdo bas 'khrigs pa tsa muṇḍa […] /. Translation Kapstein 2003: 243 and 249). On Bhāviveka’s treatment of mantras, see Braarvig 1997 and Kapstein 2003: 233–255. On Cāmuṇḍā, see Renou 1985: 522–523 (§§1072–1073), Verardi 2011: 284–296, and above, p. 110, n. 63.

134

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

post-initiatory rules,134 empowerment, visualization practices, maṇḍalas, mudrās, etc., all of which are revealed in Tantras and mantrakalpas of Buddhist as well as non-Buddhist (Vedic [vaidika], Śaiva [māheśvara], Jaina [ārhata], gāruḍa) obedience.135 Moreover, Dharmakīrti develops an elaborate critique of the allegedly natural efficacy of mantras, a doctrine that is well represented in seventh-century Buddhist Tantras. There seems to be nothing problematic in ascribing this testimony to the period around 550: early Śaiva Saiddhāntika scriptures were then already widespread; independent evidence of early Śākta Śaiva scriptures is available for the relevant period; all the ritual items and conceptions referred to by Dharmakīrti are well attested in earlier or contemporaneous Chinese translations and/or apocrypha; finally, there is no reason to believe that ideas permeating the seventh-century MVASū and STTS were not already in existence a few decades earlier, be it just “in the air” or as components of primarily oral teachings. But Dharmakīrti’s testimony raises another question. Considering that the great Buddhist philosopher went to the trouble of attacking initiation as soteriologically relevant as well as the doctrine of an inherent efficacy of mantras, we have to admit that he rejected two key elements of Tantric discourse and legitimation. In so doing, Dharmakīrti acted as a guardian of Mahāyānist gradualism and ontological orthodoxy. In this regard, Dharmakīrti can be seen as a philosopher fully committed to the new historical challenges and the most prominent advocate of a new Buddhist identity and attitude. At the same time, he can be seen as an uncompromising representative of Buddhist dogmatic “orthodoxy” who felt compelled to criticize the most innovative systems of the day, those that were to ensure the survival of Buddhism in India through the next seven or eight centuries. There is, moreover, evidence that Dharmakīrti’s conservative attitude regarding things Tantric (insofar as they encroach upon traditional path-structures) was not uncommon at that time, for the

134

Bhāviveka also refers to samaya in the context of mantras/dhāraṇīs in TJ D184b1/ P200b7.

135

According to PVSV 123,19–20 (avaidikānāṃ bauddhādīnāṃ mantrakalpānāṃ darśanāt) and PVṬ Je D281a3/P337b1 = PVSVṬ 449,24–25 thereon: ādiśabdād ārhatagāruḍamāheśvarādīnāṃ mantrakalpānām /.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

135

contemporary Buddhist philosopher Bhāviveka136 introduces his apology of Mahāyānist dhāraṇīs with the following “śrāvakayānist” objection: In the Mahāyāna there are taught dhāraṇī, mantra, and vidyā, which are inappropriate in both word and meaning, but for which many benefits are claimed so as to deceive the childish, and which resemble the Vedas of the others [i.e., nonBuddhists]. Without meditative cultivation even minor faults cannot be exhausted, so how can those [mantras, etc.] exhaust evil so long as conflicting emotions are gross and their roots still present? Moreover, the dhāraṇī do not pacify evil, for 137 they counteract its basis no more than do injurious actions, etc.

Here, the soteriological relevance of incantations is questioned on exactly the same grounds as that of the dīkṣā in Dharmakīrti’s PV 2: how can a practice that does not counteract the roots of evil be said to bring about liberation? Thus like Dharmakīrti, Bhāviveka engaged in a “rationalization” (Kapstein 2003: 245) of mantric power. But in stark contrast to Dharmakīrti, the Mādhyamika philosopher developed a hermeneutic and apologetic strategy aimed at showing that the use of incantations is indeed soteriologically relevant. Bhāviveka’s conclusive statement leaves no doubt on this: [B]y reciting, reflecting upon, and contemplatively cultivating the dhāraṇī, mantra, and vidyā in accord with the Tathāgata’s instruction, evil, together with its

136

For an overview of the debate regarding the authenticity of the prose commentary (here referred to as TJ) on the MHK, see Eltschinger 1998: 57–58, n. 2. Krasser (2011 and 2012a: 569–577) provides cogent arguments in favour of Bhāviveka’s authorship (at least in the hand of one of his pupils). Krasser’s results (see especially Krasser 2012a) have convinced me to reverse the relationship between Bhāviveka and Dharmakīrti, since the former obviously knows the latter and maybe also Kumārila (see also above, p. 116, n. 80). On Śrāvakayānist resistance against Buddhist Tantrism, see above, p. 115, n. 76.

137

TJ D183a6–b1/P199b2–4: theg pa chen po las yi ge daṅ don śes par mi ruṅ ba'i gzuṅs sṅags daṅ / gsaṅ sṅags daṅ / rig sṅags la sogs pa phan yon maṅ po can byis pa'i skye bo slu bar byed pa bstan pa de rnams ni gźan gyi rig byed daṅ 'dra'o // bsgom pa med pa ni skyon phra rab tsam yaṅ zad par byed nus pa ma yin te / ñon moṅs pa bsags pa daṅ de'i rtsa ba yod na sdig pa zad pa ga la 'byuṅ bar 'gyur / gzuṅs sṅags kyis kyaṅ sdig pa źi bar byed pa ma yin te / de'i rgyu daṅ mi 'gal ba ñid kyi phyir 'tshe ba la sogs pa bźin no //. Translation Kapstein 2003: 240 and 246. For a discussion of this pūrvapakṣa, see Braarvig 1997: 34.

136

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

basis, is pacified. Because they thus accord with the path, they are like the sūtras 138 and other scriptural transmissions.

2.2.7. Buddhist Hostility towards Śaivism in Early Buddhist Tantric Scriptures 2.2.7.1. As both Davidson and Sanderson have argued, the early history of Buddhist esoterism is closely connected with Śaivism and the latter’s strong appeal to monarchs. And it is most likely that the Buddhists, often by appropriation and adaptation, emulated the practices and ideologies of their rivals in order to compete more effectively for royal patronage. Several Buddhist scriptures reflect a very clear awareness of this complex relationship to Śaivism and develop various rhetorical answers to it. First, some of them betray a clear sense of the Buddhists’ indebtedness to Śaivism and attempt to downplay it by claiming priority, inverting the direction of borrowing and resorting to the Mahāyānistic device of skill in means. Second, some of them reject that the Śaiva scriptures and practices have any kind of soteriological relevance by insisting on their divine founders’ (mostly Śiva and his consort) ignorance of salvational means, devotion, surrender and conversion to their Buddhist homologues. Third, our texts develop various versions of what has come to be known as the “Maheśvara Subjugation Myth,” a narrative aimed at providing legitimation for the emerging Buddhist Tantric scriptures. To put it briefly, these texts reflect a many-layered polemical interaction with the dominant Śaiva environment, a situation of overt religious competition involving abundant expressions of Buddhist hostility towards Śaivism. They reveal that Buddhist Tantrism did not only develop against the background of Śaivism’s success, but also against Śaivism. 2.2.7.2. The edited version of the KVSū contains an interesting account of the way in which one (here the bodhisattva Sarvanīvaraṇaviṣkambhin) should be initiated into the practice of the six-syllable mantra: Then Sarvanīvaraṇaviṣkambhin said the following to the Blessed One: “To which [place] should I go, O Blessed One, where I can obtain the six-syllable great 138

TJ D184b3–4/P201a3–4: gzuṅs sṅags daṅ / gsaṅ sṅags daṅ / rig sṅags rnams de bźin gśegs pa'i man ṅag bźin du ṅag tu brjod pa daṅ / bsams pa daṅ / bsgoms pas sdig pa rgyu daṅ bcas pa źi bar byed pa yin te / lam daṅ rjes su mthun pa yin pa'i phyir mdo sde la sogs pa'i gsuṅ rab bźin no //. Translation Kapstein 2003: 249.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

137

Queen of incantations?” The Blessed One said: “In the big city of Vārāṇasī there is, O son of good family, a dharma-preacher who causes this great six-syllable incantation to be memorized, uttered, [and who] reflects upon it correctly.” [The bodhisattva] said: “I shall, O Blessed One, go to the big city of Vārāṇasī in order to see, honour and wait upon this dharma-preacher.” The Blessed One said: “Well, well, O son of good family, do so! [But,] O son of good family, the dharma-preachers who cause this six-syllable great incantation to be memorized are difficult to find. Such a dharma-preacher must be regarded as equal to a tathāgata, as a peak of merit, as the Ganges with all [its sacred] banks, as saying what is true and real, as a mass of jewels, as a wish-fulfilling gem, as a dharmaking, as rescuing the world. But (ca), O son of good family, don’t let doubt arise once you have seen [this] dharma-preacher, don’t depart from the bodhisattva level and fall into an evil state! [For] indeed, this dharma-preacher is without morality and without right conduct, [lives] surrounded with [his] wife(/wives), son(s) and daughter(s), covered with impurity, feces and urine, [and is] of unre139 strained behaviour.”

Several things are striking in this passage. First, the salvational mantric practice is to be sought in the traditional Śaiva stronghold of Vārāṇasī. While there is no compelling reason to doubt the Buddhist persuasion of this tathāgata-like lay initiate, his being located in Vārāṇasī suggests a

139

KVSū 298,2–12/Mette 152,1–29/KVSūT P261b7–262a6: atha sarvanīvaraṇaviṣkambhī bhagavantam etad avocat – kutrāhaṃ bhagavan gaccheyaṃ yatra ṣaḍakṣarīṃ mahāvidyārājñīṃ labheyam / bhagavān āha – asti kulaputra vārāṇasyāṃ mahānagaryāṃ dharmabhāṇako ya imāṃ ṣaḍakṣarīṃ mahāvidyāṃ dhārayati vācayati yoniśaś ca manasi kurute / āha – gamiṣyāmy ahaṃ bhagavan vārāṇasīṃ mahānagarīṃ tasya dharmabhāṇakasya darśanāya vandanāya paryupāsanāya / bhagavān āha – sādhu sādhu kulaputra evaṃ kuruṣva / tathāgatasamo draṣṭavyaḥ / dīkṣante / nagnaśramaṇeṣu ca / eṣu sthāneṣu dīkṣante / na teṣāṃ mokṣaḥ saṃvidyate / […] sarvadevagaṇāś ca brahmāviṣṇumaheśvarāḥ śakraś ca devānām indraś candrādityau < vāyuvaruṇāgnayo)> yamaś ca dharmarājo catvāraś ca mahārājānas te nityakālaṃ ṣaḍakṣarīṃ mahāvidyārājñīṃ prārthayanti /. &Em. (Mette 1997: 155, n. 4): dhyūpitapaṭam N1 147,2, dhyuṣitapaṭam Ed. 'Em.: maheśvareṣu Ed, N1 147,2. (Em. (N1 147,3): bailmavegarūdreṣu Ed. )Em. (N1 147,4, KVSūT): vāyuvaruṇādayo Ed. For hypotheses regarding the meaning of paṭa (“[religious] robe/garment” or “[ritual] painting”?) in the present passage, see Mette 1997: 157, n. 4, BHSD 115a s.v. indrapaṭa; on divasanirīkṣaka, see BHSD 264b s.v. The BHSD’s entry 512b s.v. vailmavegarudra can be safely deleted. On this passage, see also Studholme 2002: 85. G1 1643/71b8–1650/75a2 (Mette 1997: 110–117).

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

143

the ritual and icononographic prescriptions associated with the salvational maṇḍala.156 The same pattern recurs at the end of the STTS.157 Here, the gods belonging to the realm of desire (kāmāvacarā devāḥ), ignorant of the true nature of the Buddha, ask him a question the answer to which they do not even understand. They report the event to their lord Indra, Indra in turn to Brahmā, the lord of the realm of subtle corporeality (rūpāvacarādhipati), and Brahmā further to Īśvara (= Maheśvara), the lord of the three worlds (trilokādhipati). Maheśvara then suggests that the Buddha, provided he agrees to conform to worldly life (lokānuvartanā), will again explain his perfect awakening: The lords of all the gods, Maheśvara, etc., went to the platform of enlightenment (bodhimaṇḍa), where the Lord was sitting on the seat of enlightenment (bodhimaṇḍaniṣaṇṇa). Once arrived, they prostrated themselves with their heads before the feet of the Blessed One and spoke as follows: “Out of compassion for us, standing up from [this] layer of grass and sitting on divine/celestial seats, teach us, O Blessed One, [how] to awaken to the supreme and perfect awakening.” Then the Blessed One spoke to the lords of the gods as follows: “Learn, friends, [how] to bestow my supreme and perfect awakening.” Then they spoke as follows: “O Blessed One, we are incapable of bestowing awakening, [for] if we were capable 158 [of it], then we would awaken to this awakening ourselves!”

The ignorance and inferiority of Śiva and his consort are stressed in yet another passage of the Gilgit and the Nepali KVSū. Here, Maheśvara and Umā are shown asking the Buddha for a prediction (vyākaraṇanirdeśa) concerning their future awakening, hence converting themselves, an obvious anticipation of the Maheśvara subjugation myth:159 156

On this maṇḍala, see above, p. 119, n. 85.

157

STTS no. 3023–3042 (Horiuchi 1983: [II.]403–413).

158

STTS no. 3031–3034 (Horiuchi 1983: [II.]409–410): 3031: atha maheśvarādisarvadevādhipatayo yena bodhimaṇḍo yena bhagavān bodhimaṇḍaniṣaṇṇas tenopajagmuḥ / upetya ca bhagavataḥ pādau śirasābhivandya bhagavantam evam ūcuḥ – 3032: pratipadyasva bhagavann asmākam anukampām upādāyāsmāt tṛṇasaṃstarād utthāya divyeṣv āsaneṣu niṣadyānuttarāṃ samyaksambodhim abhisamboddhum iti / 3033: atha bhagavān devādhipatīn evam āha – pratipadyata mārṣā mamānuttarāṃ samyaksambodhiṃ dātum iti / 3034: atha ta evam āhuḥ – na vayaṃ bhagavan samarthā bodhiṃ dātum / yadi vayaṃ samarthā bhavema tadātmanaivābhisambodhim abhisambudhyemahīti /.

159

Studholme 2002: 31: “In the Kāraṇḍavyūha Sūtra […] Śiva appears to surrender himself voluntarily to Avalokiteśvara […] The conversion of all three [= Śiva, Umādevī, Bali

144

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

Then the devaputra Maheśvara went to the [place] where the Blessed One [resided]. Upon arriving, he prostrated himself with his head before the feet of the Blessed One and said this: “May I receive, O Blessed One, the detailed statement of the prediction [concerning my future awakening]?” The Buddha said: “O son of good family, go, the bodhisattva Avalokiteśvara, the great being, will give you the prediction.” Then the devaputra Maheśvara threw himself at the feet of Avalokiteśvara and started to make special praises […] Having made, in this way, [his] special praise of Avalokiteśvara, the devaputra Maheśvara remained silent. Then Avalokiteśvara said this to him: “O son of good family, why do you remain silent?” Then the devaputra Maheśvara said this to him: “Give me a prediction concerning [my future] supreme and perfect awakening!” Avalokiteśvara said this to him: “O you son of good family, in the Vivṛtā realm you will be the tathāgata, arhat [and] perfectly awakened one named Bhasmeśvara, perfected in wisdom and right conduct, well gone, world knowing, supreme, the charioteer of human ones-that-need-to-be-tamed, teacher of gods and humans, a blessed one, a bud160 dha.”

VE] figures might be said to express the same essentially triumphalist note: the Buddhist path is the superior path, taken up with enthusiasm by Bali and, even, by Śiva himself. The conversion of Śiva and Umādevī has an additional significance, however. It shows how the power of other deities may be tamed and harnessed to the Buddhadharma. The myth accounts for the synthesis of Śaivite principles into the Buddhist system.” 160

KVSū 303,32–304,10/Mette 1997: 131,33–133,20 (G1 1662/81a8–1664/82a4)/KVSūT P268b2–269a2: atha maheśvaro devaputro yena bhagavāṃs tenopasaṅkrāntaḥ / upasaṅkramya bhagavataḥ pādau śirasābhivandya bhagavantam etad avocat – labheyāhaṃ bhagavan / bhagavān āha – gaccha kulaputra avalokiteśvaro bodhisattvo mahāsattvas te vyākaraṇaṃ dāsyati / atha maheśvaro devaputro 'valokiteśvarasya pādayor nipatya stotraviśeṣaṃ kartum ārabdhaḥ […] evaṃ maheśvaro lokadhātur bhaviṣyati / atha somādevī vyākaraṇam anuprāptā //. & KVSūT reads mṅal na gnas śiṅ śa ma gaṅ ba'i sdug bsṅal ba daṅ rtag tu 'dzin pa las. ' N1 165,2 reads tatra. 161

On the Maheśvara subjugation myth, see Tucci 1932, Iyanaga 1985, Snellgrove 2004: 134–141, Davidson 1991, 1995 and 2002: 144–153, Stein 1995, and Verardi 2011: 307– 308. On the notion of “subjugation” (French “soumission”), see Iyanaga 1985: 638–639, n. 3. On Vajrapāṇi as a tamer and other subjugation tales involving Nāgas, Yakṣiṇīs and Asuras, see Snellgrove 2004: 134–135, Traité I.188, n. 1, and 548–554, n. 1, Lamotte 1966. According to Iyanaga (1985: 723–727), the story is a Buddhist “travestissement” of the myth, made famous through the Devīmāhātmya, of Durgā’s struggle against and subjugation of the two Asuras Śumbha and Niśumbha, whose names occur in a mantra uttered by Vairocana at the beginning of the STTS episode (STTS no. 656 [Horiuchi 1974: (I.)328]: OṂ ŚUMBHA NIŚUMBHA HUṂ GṚHṆA GṚHṆA HUṂ GṚHṆĀPAYA HUṂ ĀNAYA HO BHAGAVAN VAJRA HUṂ PHAṬ). Iyanaga 1985: 727: “Ainsi, il semble possible de conclure provisoirement que le premier texte où apparaisse le récit de la soumission de Maheśvara est la Deuxième Section de STTS. ; le terminus a quo de ce texte doit être placé à la formation du mythe du Devīmāhātmya, c’est-à-dire, au plus tard, le début du VIIe siècle, et son terminus ad quem, à la venue de Śubhakarasiṃha en Chine, c’est-à-dire l’année 716.”

162

Davidson 1991: 227. Note that Iyanaga does not consider external historical factors (such as “une intention de convertir les adeptes śivaïtes-śākta, ou de confirmer la foi bouddhiste au détriment de celle des hindouistes” [1985: 731]) to have played any signi-

146

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

Gray has also emphasized the role of this and other Tantric narratives in the construction of Buddhist identity: Such subordination is often dramatized in Buddhist literature by myths that portray the outsider religious group as dangerous “heretics,” whose misdeeds trigger a cosmic Buddha such as Vairocana to subjugate them, bringing them and modified forms of their practices into the Buddhist fold. These myths are products of a process in which Tantric Buddhists, having appropriated elements of Hindu ritual, were seeking to forge an identity through a representation of a radical 163 “other,” in this case Śaiva Hindus.

The myth, which relates how Vajrapāṇi finally succeeded in taming Maheśvara and his retinue, finds its locus classicus in the late seventh-century STTS.164 The frame story that pervades several chapters of the STTS can be summarized as follows. The bodhisattva Sarvārthasiddhi is about to achieve awakening and proceeds to the bodhi-tree. All the buddhas manifest themselves and reveal to him that he will fail to secure liberation through the “ordinary” set of final meditations unless he receives consecrations. These having been bestowed upon him, Sarvārthasiddhi becomes the ficant role in the formation of the tale. Iyanaga (1985: 738, n. 25) holds such an interpretation to be “vraiment trop simpliste et historiciste.” As for the question why Maheśvara was indeed elected to this role in the story, Iyanaga (1985: 740, emphasis mine) says that “[i]l faut sans doute tenir compte de l’influence du milieu historique et géographique dans lequel ce récit a été constitué; et nous pouvons rappeler à cet égard l’impact et l’essor du śivaïsme […] Mais en dehors de ces raisons proprement historiques et extérieures au bouddhisme, il semble possible d’en dégager une, interne au développement de la religion.” According to him, the only meaning that can be legitimately assigned to the myth is that given to it by contemporary Buddhists (notably Śubhakarasiṃha), who unambiguously opted for an allegorical/symbolical interpretation (where, e.g., Maheśvara personifies the obstruction of the three poisons [rāga, dveṣa, moha], and the wrath of Vajrapāṇi, especially his foot treading on Maheśvara’s head, represents the bodhicitta). For a critical appraisal of this interpretation (at least as far as the early Indic materials are concerned), see Davidson 1991: 214–218. 163

Gray 2005: 49.

164

STTS no. 619–736 (Horiuchi 1974: [I.]315–351). Partial or complete translations can be found in Tucci 1932: 135–145 (with an edition of the Sanskrit text), Davidson 1995: 550–555 and Snellgrove 2004: 135–140. A heavily annotated summary (from the Chinese version) can be found in Iyanaga 1985: 667–782. For references to other editions and to the Tibetan and Chinese versions of the episode, see Davidson 1991: 228–229, n. 4, and 2002: 375, n. 120. Regarding the many occurrences and variants of this myth, see Davidson 1991: 203–209, Gray 2005: 46–52 and Iyanaga 1985.

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

147

buddha Vajradhātu who, in the epilogue of the Sūtra, returns to the bodhitree in order to turn the wheel of the Law. But to teach the path to liberation requires the complete integration of the worldly (i.e., non-Buddhist) deities into the maṇḍala, the “essential means for obtaining enlightenment” and “direct expression of salvific reality.”165 The episode, with its lively dialogues between Vajrapāṇi, Maheśvara and Vairocana, cannot be translated here in full. Let me quote Davidson’s summary of the relevant passage (footnotes mine): The story opens with the cosmic Buddha, Vairocana, requesting that Vajrapāṇi generate or emit his adamantine family (vajrakula) […] [H]owever, Vajrapāṇi refuses the request. Different versions of the story provide variations on the rationale, but the basic line is that Maheśvara (= Śiva) is deluding beings with his deceitful religious doctrines and engaging in all kinds of violent criminal 166 conduct. So, Vairocana asks Vajrapāṇi to bring this evil character and his entourage into compliance. He utters a mantra that drags them to his palace at the summit of Mount Sumeru. Vajrapāṇi orders them to comply with the Buddha’s 167 doctrine, and all but Maheśvara submit. Maheśvara replies that he is the lord of 168 the universe and Vajrapāṇi is but a pathetic tree spirit. The two challenge each 165

Davidson 1991: 217–218.

166

STTS no. 644 (Horiuchi 1974: [I.]324): santi bhagavantaḥ sattvā maheśvarādayo duṣṭasattvā ye yuṣmābhir api sarvatathāgatair api na neyās teṣāṃ mayā kathaṃ pratipattavyam /. “O Lords, there are beings—criminals, like Maheśvara and his ilk—who remain unsubdued, even by all of you tathāgatas. How could I accomplish anything with them?” Translation Davidson 1995: 550.

167

STTS no. 666–668 (Horiuchi 1974: [I.]324): 666: pratipadyata mārṣāḥ sarvatathāgataśāsane mama cājñāṃ pālayata / 667: atha ta evam āhuḥ – kathaṃ pratipadyāmaḥ / 668: bhagavān vajrapāṇir āha – buddhaṃ dharmaṃ ca saṅghaṃ ca śaraṇapratipattitaḥ / sarvajñajñānalābhāya pratipadyadhvaṃ mārṣā iti /. “‘Friends! accomplish the dispensation of all tathāgatas and uphold my command!’ They replied, ‘Just how do we do that?’ The Blessed One, Vajrapāṇi, answered, ‘By completion of going to the Buddha, the dharma, and saṅgha for refuge. Friends, do it for the sake of attaining the omniscience of the Omniscient!’” Translation Davidson 1995: 551.

168

STTS no. 670 (Horiuchi 1974: [I.]332): ahaṃ bho yakṣa trailokyādhipatir īśvaraḥ kartādhikartā sarvabhūteśvaro devātidevo mahādevas tat katham ahaṃ yakṣājñāṃ kariṣyāmi /. “Hey sprite! I am the lord, the overlord of the triple worlds, the creator, the arranger, the lord of all spirits, the supreme God of gods, Mahādeva! How should I take orders from you, a local tree spirit?” Translation Davidson 1995: 552. On the background of this and other declarations, see Iyanaga 1985: 728–729 and Davidson 1995: 548.

148

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

other to acquiesce, and they engage in magical combat. After successive battles, eventually Vajrapāṇi is successful by means of his superior mantras and righteousness. Maheśvara and his wife, Umā, are tread on by Vajrapāṇi’s feet, which is the greatest insult in India as the feet are the most polluted limb. As a result of his victory, all of Maheśvara’s retinue, including his wife, agree to become part of Vairocana’s maṇḍala, and they are given back their names but with the word ‘vajra’ placed before each to denote their new Buddhist status. They are now bound by new vows and accept the pledge to quit their evil ways and become good Buddhists. Maheśvara is the sole exception, for he will not relent. Accordingly he is killed, and his life transferred to another realm, where he becomes a 169 Buddha named Bhasmeśvara-nirghoṣa, the Soundless Lord of Ashes. In honor of the great victory over the demonic Maheśvara, a new maṇḍala is created of the gods and goddesses formerly in Śiva’s service. The new maṇḍala is entitled Trailokyavijaya, the Victory over the Triple World, with Vajrapāṇi appearing in the 170 center as the divinity Trailokyavijaya.

As already pointed out by Davidson and Stein, the Maheśvara subjugation myth can be read as a reworking of the episode of the Buddha’s victory over Māra (with two changes, however: first, Vajradhātu has already achieved awakening—but not yet started teaching—when he faces Maheśvara; second, Māra never surrenders and even less so achieves buddhahood). But both Māra and Maheśvara represent major obstacles to Śākyamuni’s and Vajradhātu’s desire to teach the path to liberation.171 As long as the worldly gods headed by Maheśvara are not tamed, converted and integrated into the maṇḍala, the turning of the Law can also not be effected, nor can liberation be achieved on the side of suffering human

169

On this name, see above, p. 144, n. 160.

170

Davidson 2002: 150–151. For the setting of the story in the STTS, see also Wedemeyer 2013: 82.

171

Note that in the Gilgit KVSū, it is Māra and his allies who endeavour to oppose the bodhisattva Sarvanīvaraṇaviṣkambhin’s quest for the six-syllable mantra into which he wishes to be initiated and which he desires to teach for the welfare of suffering living beings. This is done first by magically spiking the bodhisattva’s path with sharp blades (1631/51b4–5, Mette 1997: 98: [mā]reṇa pāpīmatā kṣaradhārā-r upacitapathaṃ nirmitavantaḥ […]); second, by torturing him with diamond thorns (1631/51b8, Mette 1997: 98: māreṇa pāpīmatā punar eva vajramayāni kaṇṭakā[ni] […]); third, by magically creating heaps of glowing charcoals before him (1632/52a4, Mette 1997: 99: purasthā aṅgārarāśim abhinirmitavantaḥ […]).

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

149

beings. This is likely to be the rationale behind Vajrapāṇi’s wrath, for according to another passage of the STTS, wrath (krodha) is the supreme attainment(/power-maker) in order to benefit living beings, in order to teach the evil ones [like Maheśvara and his fold], and in order to protect the teaching of the Buddha.172

2.2.8. Conclusion: Towards a New Modality of Mahāyānist Self-Assertion? To sum up, the early medieval period that was initiated with the collapse of the Gupta Empire brought about new sociopolitical and economic circumstances. At the sociopolitical level, the new historical matrix was characterized by a movement towards regionalization, the affirmation of local identities and power, and the creation of new urban centres. To these factors, one should add a culture of belligerence and the penetration of Brahmanical values into previously peripheral territories. At the economic level, the early medieval period saw the growth of rural economy and a rapid decline of guilds and international trade, a phenomenon of farreaching consequences for the Buddhist institutions insofar as it involved a transfer of economic power from the hands of the guilds and traders to the state treasury. Issues of power legitimation crystallized in a massive buildup of temples, palaces and cities named after various monarchs, while the development of both agriculture and urban settlements increased irrigation requirements. These new infrastructures created new religio-political needs, such as the provision of temple officiants, chaplains, urban and architectural planning, consecration ceremonies and the brahmanization of the newly conquered territories. Moreover, the sovereigns expected that those whom they supported would provide legitimation for their power and espouse their aesthetics and values. There is every reason to believe that the Śaivas were very successful in meeting these new religio-political needs. This they did, in addition to their leaning towards violence and eroticism, by extending their ritual repertoire, creating new classes of officiants and initiands, integrating the mundane values of Brahmanism, and assimilating local cults and deities. In the new political landscape, the Śaivas’ appeal to monarchs forced all the other denominations to adapt by appropriating their 172

STTS no. 2953 (Horiuchi 1983: [II.]381): sarvasattvahitārthāya duṣṭānāṃ śāsanāya ca / buddhaśāsanarakṣārthaṃ krodhaḥ siddhikaraḥ paraḥ //.

150

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

rivals’ repertoire in order to compete more effectively for patronage. This provides the rationale behind the “Tantricization” of these religious traditions, a phenomenon seen most prominently at work in Mahāyāna Buddhism, which, at least in its eastern strongholds, remained Śaivism’s most serious challenger throughout the period. In my opinion, there is a sizable amount of evidence that this transformation of Mahāyāna Buddhism started not during the second part of the seventh century, but some time during the sixth century. For, provided we agree that the birth of Buddhist esoterism owes at least as much to external factors as to a purely internal process of gradual development and systematization, we must admit that most of the factors responsible for this were already in existence at that time. This is the case of the historical matrix outlined above, but is also true of the Śaiva Mantramārga in both its Saiddhāntika and Śākta components. In addition to the independent evidence provided by the earliest available scriptures associated with these currents, Dharmakīrti has quite a lot to offer in this regard, for he both criticizes a Śaiva initiatory practice supposed to bring about liberation, and alludes to early Śākta Śaiva Tantras. Moreover, Dharmakīrti’s writings testify to his familiarity with ritual practices and conceptions that he labels “Buddhist” and that are best considered Tantric. Finally, other textual witnesses recommend an earlier chronology: first, Matsunaga, Strickmann and Shinohara have pointed to large-scale ritual innovations of Indian provenance in Chinese translations and/or “apocrypha” by the turn of the sixth century; second, the Gilgit KVSū outlines a self-contained and soteriologically valued Tantric system centring on Avalokiteśvara and endowed with incantations, hand gestures, maṇḍalas, secrecy, anti-Śaiva rhetoric, etc.; third, there is evidence that Buddhist esoterism was already well established in Nālandā during the first decades of the seventh century: the composition of the MVASū, the presence of Śubhakarasiṃha’s Tantric master Dharmagupta, and King Harṣa’s plausible patronage of a temple criticized for its Kāpālika-like character. From the sixth century on, the Buddhists set out to compete for patronage with the Śaivas, a situation that is reflected both in their appropriationcum-adaptation of the new practices and in the many rhetorical modalities of their hostility towards Śaivism. The Buddhists claim the anteriority of their teachings, invert the direction of borrowing, and resort to the apologetic and hermeneutic device of skill in means; they emphasize the ignorance and inferiority of the Śaiva deities and present them as venerat-

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

151

ing, surrendering and converting to the Buddha and his teaching; much in the same way as Māra in the legend of Śākyamuni, Śiva is now depicted as an obstacle to the revelation of the Tantric path, and is consequently subjugated, trodden upon and even killed by Vajrapāṇi and others. As we can see, the focus of these innovative elites is now located outside Buddhism and not, as had been the case during most of the “Middle Period of Indian Buddhism,” within the Buddhist order. To be more precise: the Indian representatives of the Mantranaya (as opposed to the Pāramitānaya) understood themselves as belonging to the Mahāyāna. But whereas the Mahāyāna had hitherto mainly developed against “mainstream” Buddhism and stressed its many superior aspects with regard to the śrāvakas, its esoteric version developed in opposition to a non-Buddhist, most prominently Śaiva environment. This shift in focus must have had a deep impact on these innovators’ self-assertion as well as on the construction of Buddhist identity. The stake is no longer, or only to a small degree, to impose a certain interpretation of the Buddhist Law on allegedly inferior Buddhist denominations, but to construe a Buddhist identity by opposing an abhorrent and impeding other. As Gray rightly points out in referring to the work of John Henderson, polemical religious discourse inevitably implies an account of both self and other, of orthodox as well as heretical; for the former positions and defines itself by reference to the latter, even arises and develops historically by constructing an 173 inversion of the heretical other.

In other words, the Buddhist esoterists were inclined to substitute the normative opposition between Mahāyāna and Śrāvakayāna, or their respective representatives, with a dialectic between the non-Buddhists and the Buddhists or, as many of our texts have it, between mundane (laukika) and supramundane (lokottara) deities and practices. Two remarks seem necessary. First, this is certainly not to say that the traditional “others” of the Mahāyāna, viz. the śrāvakas and the pratyekabuddhas, cease to be belittled, for there is much evidence to the contrary.174 Second, this new diction of

173

Gray 2005: 50.

174

See, e.g., MVASū Tha P133b6–134a1 (Hodge 2003: 128) for classic statements regarding the comparatively lesser cognitive attainments of the śrāvakas and the pratyekabuddhas; MVASū Tha P134a8–b1 and P135b6–136a1 (Hodge 2003: 129–130 and 136) for the inferiority of the mantras of the śrāvakas and the pratyekabuddhas; MVASū

152

BUDDHIST EPISTEMOLOGY

AS

APOLOGETICS

religious identity is by no means incompatible with the “culture of ritual eclecticism” described by Granoff, for according to her, “ritual eclecticism was possible because [the] groups [involved] could define themselves as different from each other very easily” and “did not need these ritual markers as boundary markers.175 In addition, conversion and subjugation myths were certainly aimed not only at emphasizing Buddhist superiority, but also (1) at “bringing [outsider religious groups] and modified forms of their practices into the Buddhist fold,”176 and (2) at demonstrating that “no depravity is irredeemable.”177 This being said, let me turn back to the issue of the new modalities of Mahāyānist self-assertion. As we have seen above, wrath like Vajrapāṇi’s against the non-Buddhist deities serves the threefold purpose of benefitting living beings, teaching evil ones the path to salvation, and protecting the Law of the Buddha.178 Allusions to outsiders (tīrthya, tīrthika, and often duṣṭa), their false views (dṛṣṭi), their being tamed (vinaya) or even destroyed (vināśa) can be found throughout the STTS.179 The same kind of rhetoric is seen at work in an early Tantric work translated by Puṇyodaya no later than 663. Among the vows to be made while paying homage to the Buddha, one should, with one’s left hand touching the ground, wish that all outsiders who are difficult to vanquish be drawn to the saddharma by the four articles of attraction (saṅgrahavastu).180 In the same Tha P183b1–4 (Hodge 2003: 337), where śrāvakas, though starkly contrasted to the outsiders, are limited to the literal meaning and the realization of partial knowledge. 175

Granoff 2000: 421.

176

Gray 2005: 49.

177

Davidson 1991: 227.

178

See above, p. 149, n. 172.

179

See, e.g., STTS no. 2156 (Horiuchi 1983: [II.]230): tīrthikaiḥ; STTS no. 2490 (Horiuchi 1983: [II.]291): tīrthikānām; STTS no. 2496 (Horiuchi 1983: [II.]292) and no. 3021 (Horiuchi 1983: [II.]403): tīrthyadṛṣṭiḥ; STTS no. 2498 (Horiuchi 1983: [II.]292): tīrthikānāṃ vināśāya dhruvaṃ siddhim avāpnuyāt; STTS no. 2598 (Horiuchi 1983: [II.]310): duṣṭānāṃ vinayārthena.

180

See Lin Li-Kouang 1935: 98. The work is entitled *Vimalajñānabodhisattvaparipṛcchā (T. 487). On the four saṅgrahavastus, see BHSD 548b s.v. Note also Shinohara’s (2010: 396) comment on the Collected Dhāraṇī Scripture (see above, p. 118, n. 82): “The assumption is clearly that the teaching of the Eleven-faced Avalokiteśvara, coming last, is superior to the teachings presented by other beings in the audience, including nonBuddhist teachers.”

BUDDHIST ESOTERISM

AND

EPISTEMOLOGY

153

way, statements reflecting the federation of all Buddhists in opposition to outsiders can certainly be encountered, although this needs further investigation. This is the case of a passage in the Nepali recension of the KVSū that occurs at the close of the iconographic instructions pertaining to the layout of the maṇḍala. Through this ritual practice, it is said, all the bodhisattvas are [to be] living in their last existence (caramabhavika); all will rid themselves of human suffering and quickly awaken in the supreme and perfect awakening. Therefore, the teacher should not bestow [the six-syllable incantation] unsuitably (asthāne): it should be bestowed either on the one who is of zealous [Buddhist] persuasion or on the one who is of zealous Mahāyānist persua181 sion, but it should not be bestowed on an outsider.

This federative identity involving the greater dignity of the śrāvakas and pratyekabuddhas also appears in various passages of the MVASū, an interesting fact if one considers that this text (at least its first chapter) has often been considered more genuinely Mahāyānist than the following Tantras. One of these passages is concerned with the respective qualities of the mundane and supramundane mantras and mudrās. Here, the jinas, the pratyekabuddhas, the śrāvakas and the bodhisattvas side together against Maheśvara and the mundane deities.182 The same is repeated with regard to the samādhis of “the great sage buddhas and their children” which, contrary to their mundane counterparts, have no perceptual forms.183 Of other textual evidence of this attitude,184 let me quote an intriguing passage from the MVASū:

181

KVSū 296,17–21/Mette 1997: 149/KVSūT P260a6–8: te sarve caramabhavikā bodhisattvā bhavanti / duḥkhena viprahīṇā bhaviṣyanti / kṣipraṃ :°Í+Í$® ̹¼³»¼•»Ä¶Í+Í$ ½Á³¹¹•½¿³» • »¼ ̹¼³»Á¶½Á³¹¼•¹À • »ÄÀ ̽¹³¹¾•¹¼¶¾¿³Ä•»º • »ÄÄ ÌÀ¿³»¾•»¼¶ÄÀ³¹¹•¹¾ • »ÀÁ ÌÀ¿³»À•»Á¶ÄÀ³¹¾•¹À • »ÀÁ ÌÁ¾³»»•»½»º¼³»½•»À • »ÀÁ ÌÁÀ³¿•ÁÁ³»¶»º¿³¹•»ºÄ³¼ • ¹ºÁ ÌÁÀ³»»•»¹¶»º¿³¿•Ä • ½»Ä ÌÁÀ³»¿•»Ä¶»º¿³»¿•¹º • ½»Ä Ì»ÀÁ³¼¶¹¾½³»º"" • ¹ÄÁ Ì»Á¼³»¾•»ÁÀ³Á¶¹¼À³¹½•¹¼¿³½ • ¹º¼ @59:D2EEG239ą>:GŚEE:°Í+Í$Ƥ® Ì»Á¼¾•À¶ ¹¹»¼•¿ • ¹ºÀ Ì»Á¼Á¶ ¹¹¹» • ½»Ä Ì»Á¼Á•»ÁÀ»¶ ¹¹»¹•½ • ¹º¾ Ì»ÁÀ½¶ ¹¹»À • ½»Ä @59:D2EEG239ą>:GJèè59:DąEC2 °Î Ă® ¹Ä¿³¹¿ • ¹¼¹ 2C26=èA2AC25õA2°Î ® ½½³½•À • »½½ 2EFŮā2E2:@EE2C2AC25õA2°Ì$ .® »½³¹½•»¾³¿ • ¹¹Á ÀÁ³»¼•»À • ½¹½ À¿³¼ • ½¹» Áº³»¼•»À • ½¹¹ Áº³»Ä•¹º • ½¹º Áº³¹º•¹¹ • ½¹º 2DE2Gè=2AC2