The Routledge Handbook of Indian Buddhist Philosophy 9781138492257, 9781351030908

Citation preview

‘A splendid and comprehensive study of Buddhist philosophy in its country of origin. The Routledge Handbook of Indian Buddhist Philosophy is an important addition to the global philosophy movement.’ – Owen Flanagan, Duke University, USA ‘With an emphasis on the creative originality of key historical philosophers, and with a line-up of the finest scholars in Buddhist philosophy, this volume is sure to become an indispensable resource for teachers and scholars alike, and also serve as a magnificent introduction to the dynamism and potentiality of the field.’ – Jonardon Ganeri, University of Toronto, Canada ‘This carefully edited volume provides an understanding of Indian Buddhist philosophy through the study of the authorial figures that shaped its schools and doctrines. From the early discourses to the Fourteenth Dalai Lama, a picture emerges of an active intellectual enterprise devoted to the exploration of some of the most profound issues in philosophy.’ – Anita Avramides, University of Oxford, UK ‘An amazing resource for scholar and student alike, the chapters in this book, written by the very best scholars in the field, masterfully introduce the life and thought of the major figures of Indian Buddhist philosophy in clear and readable prose. There is no book quite like it. The next time I teach a course on this subject, this book will be atop the reading list.’ – José Ignacio Cabezón, University of California, Santa Barbara, USA ‘This Handbook is a landmark contribution to the field, offering superb coverage of the principal Indian Buddhist philosophers and showcasing the breadth and depth of one of the world’s richest philosophical traditions. It will be an essential reference work for scholars and students alike for years to come.’ – Evan Thompson, University of British Columbia, Canada ‘This is a great resource for all those who want to learn more about Indian Buddhism and its philosophical traditions. It explores in great depth many different aspects of these rich traditions and will be an invaluable resource for those interested in further exploring Indian Buddhist philosophy.’ – Georges Dreyfus, Williams College, USA

THE ROUTLEDGE HANDBOOK OF INDIAN BUDDHIST PHILOSOPHY

The Routledge Handbook of Indian Buddhist Philosophy is the first scholarly reference volume to highlight the diversity and individuality of a large number of the most influential philosophers to have contributed to the evolution of Buddhist thought in India. By placing the author at the center of inquiry, the volume highlights the often unrecognized innovation and multiplicity of India’s Buddhist thinkers, whose unique contributions are commonly subsumed in more general doctrinal presentations of philosophical schools. Here, instead, the reader is invited to explore the works and ideas of India’s most important Buddhist philosophers in a manner that takes seriously the weight of their philosophical thought. The forty chapters by an international and interdisciplinary team of renowned contributors each seek to offer both a wide-ranging overview and a philosophically astute reading of the works of the most seminal Indian Buddhist authors from the earliest writings to the twentieth century. The volume thus also provides thorough coverage of all the main figures, texts, traditions, and debates animating Indian Buddhist thought, and as such can serve as an in-depth introduction to Buddhist philosophy in India for those new to the field. Essential reading for students and researchers in Asian and comparative philosophy, The Routledge Handbook of Indian Buddhist Philosophy is also an excellent resource for specialists in Buddhist philosophy, as well as for contemporary philosophers interested in learning about the rigorous and rich traditions of Buddhist philosophy in India. William Edelglass is Associate Professor of Philosophy at Emerson College and Director of Studies at the Barre Center for Buddhist Studies, USA. Pierre-Julien Harter is Assistant Professor of Philosophy and The Robert H.N. Ho Family Foundation Professor of Philosophy in Buddhist Studies at University of New Mexico, USA. Sara McClintock is Associate Professor of Religion at Emory University, USA.

ROUTLEDGE HANDBOOKS IN PHILOSOPHY

Routledge Handbooks in Philosophy are state-of-the-art surveys of emerging, newly refreshed, and important fields in philosophy, providing accessible yet thorough assessments of key problems, themes, thinkers, and recent developments in research. All chapters for each volume are specially commissioned, and written by leading scholars in the field. Carefully edited and organized, Routledge Handbooks in Philosophy provide indispensable reference tools for students and researchers seeking a comprehensive overview of new and exciting topics in philosophy. They are also valuable teaching resources as accompaniments to textbooks, anthologies, and research-orientated publications.

Also available: THE ROUTLEDGE HANDBOOK OF LIBERAL NATURALISM Edited by Mario De Caro and David Macarthur THE ROUTLEDGE HANDBOOK OF PHILOSOPHY, POLITICS, AND ECONOMICS Edited by C.M. Melenovsky THE ROUTLEDGE HANDBOOK OF PHILOSOPHY OF SEX AND SEXUALITY Edited by Brian D. Earp, Clare Chambers, and Lori Watson THE ROUTLEDGE HANDBOOK OF PHILOSOPHY OF FRIENDSHIP Edited by Diane Jeske THE ROUTLEDGE HANDBOOK OF INDIAN BUDDHIST PHILOSOPHY Edited by William Edelglass, Pierre-Julien Harter, and Sara McClintock For more information about this series, please visit: www.routledge.com/Routledge-Hand books-in-Philosophy/book-series/RHP

THE ROUTLEDGE HANDBOOK OF INDIAN BUDDHIST PHILOSOPHY

Edited by William Edelglass, Pierre-Julien Harter and Sara McClintock

Cover image: © Getty Images First published 2023 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 605 Third Avenue, New York, NY 10158 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2023 selection and editorial matter William Edelglass, Pierre-Julien Harter, and Sara McClintock; individual chapters, the contributors The right of William Edelglass, Pierre-Julien Harter, and Sara McClintock to be identified as the authors of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data Names: McClintock, Sara L., editor. | Edelglass, William, editor. | Harter, Pierre-Julien, editor. Title: The Routledge handbook of Indian Buddhist philosophy/edited by William Edelglass, Pierre-Julien Harter, and Sara McClintock. Description: 1. | New York City: Routledge, 2022. | Series: Routledge handbooks in philosophy | Includes bibliographical references and index. Identifiers: LCCN 2022001712 (print) | LCCN 2022001713 (ebook) | ISBN 9781138492257 (hardback) | ISBN 9781032301952 (paperback) | ISBN 9781351030908 (ebook) Subjects: LCSH: Buddhism – Philosophy. | Buddhism – India. | Buddhist philosophy. Classification: LCC B162. R68 2022 (print) | LCC B162 (ebook) | DDC 294.3 – dc23/eng/20220202 LC record available at https://lccn.loc.gov/2022001712 LC ebook record available at https://lccn.loc.gov/2022001713 ISBN: 978-1-138-49225-7 (hbk) ISBN: 978-1-032-30195-2 (pbk) ISBN: 978-1-351-03090-8 (ebk) DOI: 10.4324/9781351030908 Typeset in Times New Roman by Apex CoVantage, LLC

I think those with wisdom drink something else, something that gives strength, is delicious and irresistible, they drink like travelers who gulp rain just fallen from a dark cloud. – Sukka, from the Therīgāthā, translated by Charles Hallisey Thought is not calmed without speaking words. Scholars will know which words have weight and which are drivel. Through long habituation, those who are attached to their own positions, having lapsed into hostile intent, do not believe others even when they speak what is true and beneficial. But scholars, remaining unbiased toward both the trustworthy and the untrustworthy, understand through intellectual examination the meaning of both the proponent’s and the opponent’s position. Having analyzed whether this or that word has weight or not, they should speak accordingly. . . . Fittingly, scholars will amuse themselves by analyzing your words and our words, saying “this has weight” and “this is drivel.” – Bhāviveka, the Heart of the Middle, verse 4.74, with commentary from the Flame of Logic

For all the myriad embodied hearts and minds who have conceptualized, composed, taught, remembered, copied, promulgated, edited, translated, debated, explained, wrestled with, and provided for the texts and ideas that form the ground from which this book has grown.

CONTENTS

Acknowledgmentsxvi Notes on Contributors xvii

Introduction William Edelglass, Pierre-Julien Harter, and Sara McClintock

PART 1

1

Buddhas as Philosophers

7



9

Introduction to Part 1

  1 Gotama Buddha: His Quest and His Teachings Bhikkhu Anālayo

11

  2 Siddhārtha Gautama: Beyond the Historical Figure Richard F. Nance

27

  3 The Tantric Buddha: Primordial Buddhas as Philosophical Authors Vesna A. Wallace

46

  4 Maitreya: The Future Buddha as an Author Klaus-Dieter Mathes

64

PART 2

Poet Philosophers

79



81

Introduction to Part 2 xi

Contents

  5 Aśvaghoṣa: The Dawn of Indian Buddhist Philosophy Vincent Eltschinger  6 The Milindapañha: How to Use a Philosophical Resource and Find a Literary Gem Sonam Kachru

83

97

  7 Cāttaṉār: Poet-Philosopher in Tamiḻ Anne E. Monius

113

  8 Saraha: The Anti-Philosopher as Philosopher Roger R. Jackson

124

PART 3

Abhidharma Philosophers

139



141

Introduction to Part 3

  9 The Dhammasaṅgaṇī and Vibhaṅga: The Perfectly Awakened Buddha and the First Abhidhammikas Maria Heim

143

10 Moggaliputta Tissa’s Points of Discussion (Kathāvatthu): Reasoning and Debate in Early Buddhist Thought Rupert Gethin

160

11 Kātyāyanīputra and the Large Commentary (Mahāvibhāṣā): The Development of Abhidharma Literature and of a Sarvāstivāda Self-identity 172 Bart Dessein 12 The Vātsīputrīya/Sāṃmitīya: Buddhist Personalism as a Mainstream School of Thought Peter Skilling PART 4

188

Philosophical Founders

205



207

Introduction to Part 4

13 Nāgārjuna: Dependent Arising Without Any Thing Arising Amber D. Carpenter

211

14 Āryadeva: Quietism and Buddhist Ethics Tom J. F . Tillemans

236

xii

Contents

15 Asaṅga: Great Systematizer of Yogācāra Thought Jowita Kramer

252

16 Vasubandhu: Mainstream and Mahāyāna Jonathan C. Gold

266

17 Dignāga: Early Innovator in Buddhist Epistemology Kei Kataoka

284

18 Dharmakīrti: Philosopher and Defender of the Faith John Taber

303

PART 5

Early-Period Commentators (Fifth–Seventh Century)

323



325

Introduction to Part 5

19 Buddhaghosa: Phenomenology, Hermeneutics, and Understanding Chakravarthi Ram-Prasad

328

20 Bhāviveka: Madhyamaka Dialectic, Doxography, and Soteriology Karl-Stéphan Bouthillette

346

21 Dharmapāla: A Janus-Faced Interpreter of Yogācāra? Ching Keng

361

22 Sthiramati: A Yogācāra Commentator and Innovator Roy Tzohar and Jowita Kramer

376

23 Devendrabuddhi and Śākyabuddhi: Dharmakīrti’s First Commentators Alexander Yiannopoulos

393

24 Candrakīrti: Gardener of Sky-Flowers Mattia Salvini

404

PART 6

Middle-Period Commentators (Eighth–Ninth Century)

421



423

Introduction to Part 6

25 Śubhagupta: An Externalist Outsider Within the Dharmakīrtian Tradition Serena Saccone xiii

427

Contents

26 Dharmottara: Systematic and Innovative Commentator David Nowakowski 27 Jñānagarbha: Two Truths Theory, Gradualism, and Mādhyamika Philosophy Ryusei Keira

439

450

28 Śāntarakṣita: Climbing the Ladder to the Ultimate Truth Allison Aitken

463

29 Kamalaśīla: Mādhyamika Champion of Magical Reason Sara McClintock

480

30 Haribhadra: The Voice of Perfect Wisdom Pierre-Julien Harter

497

31 Śāntideva: Virtue on the Empty Path of the Bodhisattva Stephen Harris

511

32 Prajñākaragupta: Buddhist Epistemology as the Path to the Wisdom of Non-Duality Shinya Moriyama PART 7

528

Late-Period Commentators (Tenth–Twelfth Century)

541



543

Introduction to Part 7

33 Jitāri: A Later Buddhist Master of Debate Junjie Chu

547

34 Jñānaśrīmitra: Variegated Non-Duality Lawrence McCrea

558

35 Ratnakīrti: Aligning Everyday Experience with Momentariness and Idealism Patrick McAllister

573

36 Ratnākaraśānti: The Illumination of False Forms Gregory Max Seton

587

37 Atiśa: The Great Middle Way of Mere Appearance James B. Apple

601

xiv

Contents

38 Abhayākaragupta: A Last Great Paṇḍita Kazuo Kano PART 8

615

Modern Philosophers

631



633

Introduction to Part 8

39 B. R. Ambedkar: Justice, Religion, and Buddhist Political Philosophy William Edelglass

635

40 The Dalai Lama XIV: A Modern Indian Philosopher Jay L. Garfield

650

Index663

xv

ACKNOWLEDGMENTS

We are deeply grateful to our colleagues for the excellent chapters they contributed and for their collaboration and patience. We are especially thankful to Sarah Fleming and Eva ­Seligman, graduate interns from Harvard Divinity School at the Barre Center for Buddhist Studies, who assisted us during the editing process; their tremendous skill as editors contributed greatly to this volume. We would also like to acknowledge the institutions whose support directly benefited our work: the Barre Center for Buddhist Studies, Emory University, Harvard University, and the University of New Mexico. Finally, we are profoundly appreciative of our family members, with whom we are looking forward to spending more time: Leo, Sarah, ­Jasmine, Jake, Amit, Kirstin, and Win.

xvi

NOTES ON CONTRIBUTORS

Allison Aitken is Assistant Professor of Philosophy at Columbia University, USA. Her research centers on metaphysics in the history of philosophy. She has a forthcoming monograph titled Introduction to Reality: Śrīgupta’s Tattvāvatāravṛtti, and her articles on topics in Indian and Tibetan Buddhist philosophy, as well as Early Modern European philosophy, have appeared in such publications as Philosophers’ Imprint, Analysis, Philosophy East and West, and Journal of South Asian Intellectual History. Bhikkhu Anālayo completed a Ph.D. thesis on the Satipaṭṭhāna-sutta at the University of Peradeniya in the year 2000 (published by Windhorse in the UK) and a habilitation thesis at the University of Marburg in the year 2007, comparing the Majjhima-nikāya discourses with their Chinese, Sanskrit, and Tibetan counterparts (published by Dharma Drum in Taiwan). The main focus of his more than 400 publications is on comparative studies of early Buddhist texts. He recently retired from a position as a professor at the University of Hamburg and currently resides at the Barre Center for Buddhist Studies in the USA, where he spends most of his time in meditation. James B. Apple is Professor of Buddhist Studies at the University of Calgary, Canada. His research focuses on the critical analysis of Mahāyāna sūtras and topics within Indian and Tibetan forms of Buddhism. His books include Jewels of the Middle Way: The Madhyamaka Legacy of Atiśa and His Early Tibetan Followers (Wisdom, 2019), A Stairway taken by the Lucid: Tsong kha pa’s Study of Noble Beings (Aditya Prakashan, 2013) and Stairway to Nirvāṇa (SUNY, 2008). Karl-Stéphan Bouthillette, born in Montréal, received his first M.A. in Sciences of Religions at Laval University, Canada (2011), and his second one in Sanskrit Studies at Jawaharlal Nehru University, India (2013). He obtained his Ph.D. (2018) from the Institute for ­Indology and Tibetology of the Ludwig-Maximilians-Universität, Germany. In 2020, he published his most recent book, at Routledge, titled Dialogue and Doxography in Indian Philosophy: Points of View in Buddhist, Jaina, and Advaita Vedānta Traditions. Now at Ghent University, Belgium, as an FWO post-doctoral fellow, his research focuses on lists and list-making as contemplative devices in Indian philosophy.

xvii

Notes on Contributors

Amber D. Carpenter is Associate Professor at Yale-NUS College, Singapore. She publishes in ancient Greek philosophy, especially the ethics, epistemology, and metaphysics of Plato. Her book, Indian Buddhist Philosophy (Routledge) appeared in 2014. Her current project, Buddhist Platonist Dialogues (buddhistplatonistdialogues.com) joins the two traditions in philosophical inquiry. She has taught at Oxford, Cornell, St. Andrews, and York, and held research fellowships and visiting appointments at the Einstein Forum (Potsdam), University of Melbourne, Yale University, and with the Templeton Religious Trust. Her co-edited Portraits of Integrity, arising from the work of The Integrity Project (integrityproject.org), appeared in 2021. Junjie Chu is a research fellow in the Department of Indology and Central Asian Studies at the University  of Leipzig, Germany. His current research centers on the Yogācāra  system, especially its epistemological theories, as well as Sanskrit manuscript studies. He is the author of “On Dignāga’s Theory of the Object of Cognition as Presented in PS(V) 1” (Journal of the International Association of Buddhist Studies, 29 (2), 2006 (2008) pp. 211–53.); and “Sanskrit fragments of Dharmakīrti’s Santānāntarasiddhi” (Religion and Logic in Buddhist Philosophical Analysis, ed.  by Krasser Helmut et al., Vienna, 2012, pp. 33–42). Bart Dessein is full Professor at the Department of Languages and Cultures of Ghent University, Belgium and Member of the Royal Academy for Overseas Sciences, Belgium. He obtained his Ph.D. from Ghent University in 1994 with a translation and study of Dharmatrāta’s *Saṃyuktābhidharmahṛdaya, a Sarvāstivāda Abhidharma treatise. He has published extensively on the philosophy of early Buddhism, mainly on the Sarvāstivādins and the Mahāsāṃghika schools, as well as on early Chinese Buddhism. William Edelglass is Associate Professor of Philosophy at Emerson College, USA, and Director of Studies at the Barre Center for Buddhist Studies. He publishes broadly in Buddhist studies, environmental humanities, and philosophy. William is co-editor of the journal, Environmental Philosophy, as well as Buddhist Philosophy: Essential Readings (2009), The Oxford Handbook of World Philosophy (2011), and Facing Nature: Levinas and Environmental Thought (2012). William lives on a homestead in southern Vermont. Vincent Eltschinger is Professor of Indian Buddhism at the École Pratique des Hautes Études, PSL University, France. His research focuses on the genealogy and the early history of Buddhist philosophy. He is one of the editors of Brill’s Encyclopedia of Buddhism. His numerous publications include Caste and Buddhist Philosophy (2012); Penser l’autorité des Écritures (2007), Can the Veda Speak? (2012, with H. Krasser and J. Taber), Self, No-Self, and Salvation (2013, with I. Ratié), Buddhist Epistemology as Apologetics (2014), and Dharmakīrti’s Theory of Exclusion (2018, with J. Taber, M.T. Much and I. Ratié). Jay L. Garfield is Doris Silbert Professor in the Humanities and Professor of Philosophy, Logic, and Buddhist Studies at Smith College, USA; Visiting Professor of Buddhist Philosophy at the Harvard Divinity School, USA; Professor of Philosophy at Melbourne University, Australia; and Adjunct Professor of Philosophy at the Central Institute of Higher Tibetan Studies, India. His research addresses topics in the philosophy of mind and cognitive science, ethics, epistemology, the history of Western philosophy, Indo-Tibetan Buddhist Philosophy, and the recent history of Indian philosophy.

xviii

Notes on Contributors

Rupert Gethin  is Professor of Buddhist Studies at the University of Bristol, UK, where he has taught Indian religions since 1987. His primary research interest is in the history and development of Indian Buddhist thought in the Nikāya-Āgamas and exegetical literature. He is the author of The Buddhist Path to Awakening (Brill 1992; Oneworld 2001), The Foundations of Buddhism (1998), Sayings of the Buddha: A Selection of Suttas from the Pali Nikāyas (2008) and, with R.P. Wijeratne, Summary of the Topics of Abhidhamma and Exposition of the Topics of Abhidhamma (2002). He has also published a number of scholarly articles on the theory of Buddhist meditation and Abhidharma. He has been president of the Pali Text Society since 2003. Jonathan C. Gold is Professor in the Department of Religion and Director of the Center for Culture, Society and Religion at Princeton University, USA. A scholar of Indian and Tibetan Buddhist philosophy, he is especially interested in Buddhist approaches to meaning, ethics, language, and learning. He is the author of The Dharma’s Gatekeepers: Sakya Paṇḍita on ­Buddhist Scholarship in Tibet (2007) and Paving the Great Way: Vasubandhu’s Unifying Buddhist Philosophy (2015), and is co-editor, with Douglas S. Duckworth, of Readings of Śāntideva’s Guide to Bodhisattva Practice (Bodhicaryāvatāra) (2019). Stephen Harris is Assistant Professor (Universitair Docent) at Leiden University’s Institute for Philosophy, the Netherlands. He specializes in cross-cultural and Indian philosophy, with a particular interest in Buddhist ethical texts. He has published articles in a number of academic journals, including Journal of Buddhist Ethics, Journal of Indian Philosophy, Philosophy East and West, and Sophia. Pierre-Julien Harter is Assistant Professor of Philosophy and The Robert H.N. Ho Family Foundation Professor of Philosophy in Buddhist Studies at University of New Mexico, USA. He specializes in Buddhist philosophy in India and Tibet. His research on the Buddhist concept of the path has nurtured his wide-ranging interests in different aspects of Buddhist thought, such as metaphysics and ontology, epistemology, and ethics. He works also on Indian philosophy more broadly, ancient Greek philosophy, and continental philosophy, framing his research in the larger context of philosophy by fostering conversations between different philosophical traditions and texts. Maria Heim is George Lyman Crosby 1896 & Stanley Warfield Crosby Professor in Religion at Amherst College and a Guggenheim fellow. Her most recent books are Words for the Heart: A Treasury of Emotions from Classical India (Princeton 2022) and Voice of the Buddha: Buddhaghosa on the Immeasurable Words (Oxford 2018). She is currently working on a translation of the Milindapañha for the Murty Classical Library of India. Roger R. Jackson is Professor Emeritus at Carleton College, USA, where he taught South Asian and Tibetan religions for nearly three decades. He has a B.A. from Wesleyan University, USA, and a Ph.D. in Buddhist studies from the University of Wisconsin, USA. His scholarship ranges over Indic and Tibetan Buddhist ritual, meditative practice, philosophy, and literature, as well as modern Buddhist thought. Books he has either authored or co-edited include: Is Enlightenment Possible? (Snow Lion, 1993), Tibetan Literature (Snow Lion, 1996), Buddhist Theology (Curzon Press, 2000), Tantric Treasures (Oxford, 2004), and  Mind Seeing Mind: Mahāmudrā and the Geluk Tradition of Tibetan Buddhism (Wisdom, 2019). His most recent book is Rebirth: A Guide to Mind, Karma, and Cosmos in the Buddhist World. (Wisdom, 2022).

xix

Notes on Contributors

Sonam Kachru is Assistant Professor in the Department of Religious Studies, University of Virginia, USA. He studies the history of philosophy with particular emphasis on the history of Buddhist philosophy in South Asia; topics of particular salience for him include philosophy of mind (intentionality, consciousness, attention), metaphysics, and practices of self. His first book, Other Lives: Mind and World in Indian Buddhism (2021), a study of Vasubandhu’s philosophy of mind, was published by Columbia University Press. Kazuo Kano is an Associate Professor at Komazawa University, Japan. He is working on Buddha-nature teaching in India and Tibet, and published a monograph on this issue (Buddhanature and Emptiness: rNgog Blo-ldan-shes-rab and a Transmission of the Ratnagotravibhāga from India to Tibet. Vienna: Vienna Series for Tibetan and Buddhist Studies, 2016). His research interest also includes Sanskrit manuscripts preserved in Tibet and the history of their transmission. Kei Kataoka is Associate Professor of Indian philosophy at Kyushu University, Japan. His publications include critical editions of Śabara’s commentary, Kumārila’s Ślokavārttika and Tantravārttika, Sucarita’s Kāśikā, Jayanta’s Nyāyamañjarī and Nyāyakalikā, and Aghoraśiva’s Tattvasaṃgrahalaghuṭīkā. His book, Kumārila on Truth, Omniscience, and Killing (Austrian Academy of Sciences Press, 2011) deals with a Brahmanical defense of the authoritativeness of the Vedic scripture. Ryusei Keira is Professor of Logic and Philosophy at Hosei University, Japan. After studying at Keio University (B.A. 1988) and at Tokyo University (M.A. 1991), he received his D.Litt. (2003) from Lausanne University, Switzerland. His research focuses on Mādhyamika philosophy. He is the author of Mādhyamika and Epistemology (2004). His works include “Kamalaśīla’s Madhyamakāloka, Part of Uttarapakṣa (Replies), Chapter One: Tibetan Text and Annotated Japanese Translation (1)” in Acta Tibetica et Buddhica 9 (2016) and “The description of niḥsvabhāvatā and its intentional meaning: Kamalaśīla’s solution for the doctrinal conflict between Mādhyamika and Yogācāra” in Acta Tibetica et Buddhica 2 (2009). Ching Keng is Associate Professor in the Department of Philosophy, National Taiwan University, Taiwan. His fields of research include Yogācāra, Tathāgatagarbha, and Abhidharma thought in India and China. His current research explores how, under various Buddhist cognitive frameworks, mental consciousness and sensory consciousness work together. He is the author of Toward a New Image of Paramartha: Yogacara and Tathagatagarbha Buddhism Revisited (Bloomsbury 2022) and a co-editor of Buddhist Philosophy of Consciousness: Tradition and Dialogue (Brill 2020). Jowita Kramer is professor of Indology at Leipzig University. She specializes in Indian and Indo-Tibetan Buddhism, with particular focus on the psychological concepts of the Yogācāra tradition. Her research interests also include aspects of authorship and intertextuality in Buddhist literature. She is the author of a monograph on the Yogācāra concept of the “five categories” (vastu) of reality and numerous publications on the Pañcaskandhakavibhāṣā, a sixth-century commentary by the Indian scholar Sthiramati. Before joining the University of Leipzig, she held positions at the Universities of Heidelberg and Munich, the University of California, Berkeley, the University of Oxford, and at the École Pratique des Hautes Études.

xx

Notes on Contributors

Klaus-Dieter Mathes is the Head of the Department of South Asian, Tibetan and Buddhist Studies at the University of Vienna, Austria. His current research deals with “emptiness of other” (gzhan stong) in the early Jonang tradition. He obtained a Ph.D. from Marburg University, Germany, with a translation and study of the Yogācāra text Dharmadharmatāvibhāga (published in 1996 in the series Indica et Tibetica). His habilitation thesis was published by Wisdom Publications under the title A Direct Path to the Buddha Within: Gö Lotsāwa´s Mahāmudrā Interpretation of the Ratnagotravibhāga (Boston, 2008) and his latest work, Maitrīpa: India’s Yogi of Nondual Bliss (Boulder 2021) was published by Shambhala Publications in the Series Lives of the Masters. He is also a regular contributor to Journal of the International Association of Buddhist Studies. Patrick McAllister received an M.A. in philosophy in 2005 and a Ph.D. in Buddhist studies in 2011 (supervised by Helmut Krasser) at the University of Vienna. He has been working at the Institute for the Cultural and Intellectual History of Asia since July 2016. His primary research interest is the development of Buddhist epistemological theories during the ninth–eleventh centuries (primarily in the works of Prajñākaragupta, Jñānaśrīmitra, and Ratnakīrti). He is also engaged in projects of the digital humanities, in particular EAST (http://east.uni-hd.de), a tool to collect bibliographical and prosopographical information on South Asian and Tibetan literature, and SARIT (http://sarit.indology.info), a library of electronic Indic texts. Sara McClintock is Associate Professor of Religion at Emory University, USA. Their research interests focus on questions of truth, ethics, and the production of knowledge through philosophy and narrative. They are the author of Omniscience and the Rhetoric of Reason: Śāntarakṣita and Kamalaśīla on Rationality, Argumentation, and Religious Authority (2010), and the co-editor with Georges Dreyfus of The Svātantrika-Prāsaṅgika Distinction: What Difference Does a Difference Make? (2003). They have published in such journals as Argumentation, Journal of the American Academy of Religion, Journal of Indian Philosophy, and Journal of the International Association of Buddhist Studies; they are currently working on a book on truth. Lawrence McCrea is the author of numerous papers on traditional Indian poetry, poetics, language theory, and hermeneutics. He has taught Sanskrit at the University of Chicago, Harvard University, and Cornell University, all USA, where he is currently Professor in the Department of Asian Studies. He is the author of The Teleology of Poetics in Medieval Kashmir (Harvard, 2008) and co-author (with Parimal Patil) of Buddhist Philosophy of Language in India: Jñānaśrīmitra on Exclusion (Columbia, 2010). He is currently working on a Reader on Mīmāṃsā: A Historical Sourcebook in Indian Hermeneutical Theory, under preparation for Columbia University Press. Anne E. Monius was Professor of South Asian Religions at Harvard Divinity School, USA. Her research interests lay in examining the practices and products of literary culture to reconstruct the history of religions in South Asia. Her book, Imagining a Place for Buddhism: Literary Culture and Religious Community in Tamil-Speaking South India (Oxford, 2001), examines the two extant Buddhist texts composed in Tamil. At the time of her unexpected passing in 2019, she was deeply engaged in a research project focused on the role of aesthetics and moral vision in the articulation of a distinctly Śaiva religious identity in twelfth-century South India.

xxi

Notes on Contributors

Shinya Moriyama is Professor in the Department of Philosophy at Shinshu University, Japan. His research interests lie in Buddhist epistemology in India and China, and comparative philosophy. He is the author of Omniscience and Religious Philosophy: A Study on Prajñākaragupta’s Pramāṇavārttikālaṅkārabhāṣya ad Pramāṇavārttika II 8–10 and 29–33 (LIT Verlag, 2014) and a number of articles on the theory of perception and religious philosophy of Dharmakīrti and his followers. Richard F. Nance is Associate Professor in the Department of Religious Studies at Indiana University, USA, where he teaches courses on Buddhist philosophy, rhetoric, ritual, and visual culture. He is the author of Speaking for Buddhas: Scriptural Commentary in Indian Buddhism (Columbia University Press, 2012), and has published work in various journals, including Journal of the American Academy of Religion, Journal of Indian Philosophy, Journal of the International Association of Buddhist Studies, and Revue d’Etudes Tibétaines. He is currently at work on a translation of Vasubandhu’s Vyākhyāyukti and a monograph on the transmission of Buddhist traditions in India. David Nowakowski is a public philosopher, community educator, and independent scholar. He serves as an adviser for Merlin CCC, a non-profit educational organization in Helena, Montana, USA, and as senior mentor for the Merlin Fellowship Program. He earned his Ph.D. in philosophy from Princeton University, USA, and his work has appeared in Philosophy East and West, Asian Philosophy, and Journal of Indian Philosophy. Chakravarthi Ram-Prasad is Distinguished Professor of Comparative Religion and Philosophy at Lancaster University, UK, and Fellow of the British Academy. He has written more than fifty papers on a variety of topics and seven books so far, including Knowledge and Liberation in Classical Indian Thought (Palgrave, 2001), Divine Self, Human Self: The Philosophy of Being in Two Gītā Commentaries (Bloomsbury, 2018), and Human Being, Bodily Being: Phenomenology from Classical India (Oxford, 2018). Serena Saccone is Associate Professor at the University of Naples “L’Orientale.” She holds a Ph.D. in Indology and Tibetology from the University of Turin, and was a research fellow at the Austrian Academy of Sciences from 2015–2021. Her main area of research is the intellectual history of Buddhism, focusing on South Asian authors from the early medieval period, with a specific interest in epistemology, logic, and soteriology, as well as their interconnections. Saccone’s monograph, On the Nature of Things (University of Vienna, 2018), concerns the internal Buddhist debate on cognitions and their object in the eighth century. Her second book, Tantra and Pramāṇa. Studies in the Sāramañjarī (University of Naples, 2022), ­co-authored with Péter-Dániel Szántó, deals with the interrelationship between Tantric Buddhism and the Dignāga-Dharmakīrtian tradition of logic and epistemology. Mattia Salvini studied Sanskrit and Indian Philosophy mostly in India, obtaining a B.A. and M.A. Sanskrit from RKM Vivekananda College (Madras University), India, and reading Buddhist texts with Prof. Ramshankar Tripathi (Sarnath, CIHTS). He then obtained a Ph.D. from SOAS University of London, under the guidance of Dr. Tadeusz Skorupski. He has published articles on Madhyamaka philosophy with the Journal of Indian Philosophy, Thai International Journal of Buddhist Studies, and the Bulletin de l’École française d’Extrême-Orient. Presently, Mattia is the Dean of Scriptural Languages and the Dean of the Faculty of Liberal Arts, International Buddhist College, Sadao Campus, Thailand, xxii

Notes on Contributors

and his research continues to focus on Buddhist philosophical texts and other aspects of Buddhism in India. Gregory Max Seton has been Senior Lecturer in the Religion Department at Dartmouth College, USA, since 2017. He was previously a professor of Buddhist Studies at Mahidol University in Thailand. He received his D.Phil. in South Asian Buddhist Studies from the University of Oxford, UK; a M.A. in Indo-Tibetan Buddhist studies from Naropa University, USA; a M.A. in Religious Studies from the University of California, Santa Barbara, USA, and was a DAAD research fellow at the University of Hamburg, Germany. His Sanskrit and Tibetan critical edition of Ratnākaraśānti’s Sāratamā is in press, and his monograph on Ratnākaraśānti’s system is forthcoming. Peter Skilling is Special Lecturer at Chulalongkorn University, Thailand, and an Honorary Associate, Department of Indian Sub-Continental Studies, University of Sydney, Australia. He retired as Professor of the French School of Asian Studies (EFEO) in 2016. His main field of research is the epigraphy, archaeology, history, and literature of Buddhism in South and Southeast Asia according to Sanskrit, Pali, Thai, and Tibetan sources. In 2009, he was awarded the Ikuo Hirayama Prize by the Académie des Inscriptions et Belle-Lettres, Paris. In 2010, he was elected honorary member of the Siam Society, Bangkok. In 2012, he was appointed fellow of the Khyentse Foundation. In 2017, he was elected honorary fellow of the Asiatic Society of Mumbai. Recently he published translations and studies of twenty-five sutras from the Tibetan Kanjur under the title Questioning the Buddha: A Selection of Twenty-Five Sutras (Wisdom Publications, 2021). Forthcoming is a new collection of translations from Pali and Tibetan, Buddha’s Words for Tough Times: An Anthology (Wisdom Publications, 2023). John Taber is Regents’ Professor Emeritus of Philosophy at the University of New Mexico, USA. His publications include Kumārila on Perception: A Hindu Critique of Buddhist Epistemology (RoutledgeCurzon, 2005), Can the Veda Speak? Dharmakīrti Against Mīmāṃsā Exegetics and Vedic Authority: An Annotated Translation of PVSV 164,24–176,16 (with Vincent Eltschinger and Helmut Krasser; Austrian Academy of Sciences Press, 2012), Dharmakīrti’s Theory of Exclusion (apoha): An Annotated Translation of Pramāṇavārttikasvavṛtti 24,16–93,5, Part I (with Vincent Eltschinger, Michael Torsten Much, and Isabelle Ratié; International Institute of Buddhist Studies, 2018), and diverse articles on Advaita Vedānta, Mīmāṃsā, Indian logic, and Buddhist epistemology. Tom J. F. Tillemans is a Canadian and Dutch citizen. Educated as a philosopher and ­philologist, he made his living as a professor of Buddhist Studies at the University of Lausanne, Switzerland. He was, for several years, editor in chief of a large project to translate Buddhist canonical texts from Tibetan and Sanskrit (www.84000.co). He lives on an island off the west coast of Canada. Roy Tzohar specializes in the history of philosophy with a focus on Buddhist and Brahmanical philosophical traditions in India. He is a currently an associate professor in the East and South Asian Studies Department at Tel Aviv University, Israel. He holds a Ph.D. from the Religion Department at Columbia University, USA (2011), and an M.A. in philosophy from Tel Aviv University’s Interdisciplinary Program for Outstanding Students (2004). His recent monograph, A Buddhist Yogācāra Theory of Metaphor (Oxford University Press, 2018), deals with Indian philosophy of language and experience. xxiii

Notes on Contributors

Vesna A. Wallace is Professor of Buddhist Studies in the Department of Religious Studies at the University of California, Santa Barbara, USA, where she teaches courses in South Asian and Mongolian Buddhist traditions, and advanced Sanskrit language courses. She has authored, translated, and edited six books and authored numerous articles on different aspects of Indian and Mongolian Buddhism. Her most recent book is a co-authored volume titled Text, Image, and Ritual in Mongolian Buddhism, to be published by Columbia University Press. Alexander Yiannopoulos has been studying and practicing Buddhadharma since 2005, when he took refuge under the Bodhi Tree with Chökyi Nyima Rinpoche. Dr. Yiannopoulos is a graduate of the M.A. program in Buddhist Studies and Himalayan Languages at the Rangjung Yeshe Institute in Kathmandu, Nepal, where he studied for six years, and completed his first Fulbright research fellowship. He obtained his Ph.D. in 2020 from Emory University, USA.

xxiv

INTRODUCTION William Edelglass, Pierre-Julien Harter, and Sara McClintock

The topic of this book, Indian Buddhist philosophy, is “vast and profound,” to borrow a traditional Buddhist expression. Buddhists employ this phrase to refer both to the vastness of reality itself and to a buddha’s profound and unerring realization and teaching of that reality. Ranging over approximately two and a half millennia, composed and preserved in diverse languages both in oral and written form, manifesting in myriad genres from lyric to scholastic, Buddhist philosophy in India is a field whose borders are yet to be found. Numberless manuscripts remain untranslated and even unedited, some located in India and some in monasteries, libraries, and private collections elsewhere in the world. Other works survive only in Chinese or Tibetan translations made centuries ago and are still mainly known to contemporary scholars through catalogs and by reputation. Other texts have nearly entirely disappeared and can only be glimpsed in fragmentary quotations or elusive references. No doubt, many others have vanished without a trace. The expanse of Indian Buddhist philosophy is indeed vast. But it is also profound, as those who have explored Indian Buddhist philosophy very well know. Like the rest of Indian philosophy, Indian Buddhist philosophy touches on nearly every aspect of human experience, including the nature of suffering, reality, truth, goodness, compassion, love, freedom, death, the body, the self, matter, consciousness, and divinity. Furthermore, Indian Buddhist philosophy explores these aspects of human experience with a relentless ambition to get to the heart of the matter, interrogating everything from the instruments by which knowledge can be certified to the status of knowledge itself to the problems of language, perception, inference, argument, objectivity, properties, kinds, agency, causality, morality, transformation, illusion, and much more. As the chapters that follow demonstrate, Indian Buddhist philosophers have come to a variety of conclusions about these diverse areas of inquiry. The present volume, with its forty chapters, is a window onto the vast and profound world of Indian Buddhist philosophy. It is a multifaceted stained-glass window with numerous panes, each offering a view of complex philosophical topics colored by the perspective of a particular Indian Buddhist philosopher as read and presented by a particular highly qualified contemporary scholar. Our decision to place the author at the center of inquiry represents a deliberate choice to highlight the philosophical creativity and innovation of diverse Indian Buddhist philosophers, who too often are seen as mere exegetes working in the service of pre‑existing schools. But categorizing thinkers according to schools, as if every thinker associated with that DOI: 10.4324/9781351030908-1

1

William Edelglass, et al.

school had agreed to a stable set of doctrinal commitments, obscures the reality that schools are themselves amorphous and ever-changing streams made up of the thought of many individuals over periods of time. Even as each author reflects and refracts the innovations and ideas of their predecessors and contemporaries, it is individual thinkers who make up the schools, both in terms of comprising them and, more literally, inventing them. While it is true that Indian Buddhist philosophers do generally understand themselves as working from within a tradition to which they unabashedly belong, this is not in itself anything out of the ordinary for philosophers more generally. Indeed, it is important to remember that Western philosophers, too, have undertaken their work within a series of inherited philosophical traditions, with all the resources and blinders that come along with being so situated. The tendency to see Indian philosophers – whether Buddhist or not – as so deeply enmeshed in their particular schools as to be compromised in terms of their rational capacity is an Orientalist bias. The decision to prioritize authors over schools, then, is in part a decision to resist this bias. By presenting major figures who have contributed to the evolution of philosophical thought in Indian Buddhist traditions and in Indian philosophy more broadly, this volume thus implicitly makes a case for an enlarged conception of philosophy not restricted to its European embodiment. To be clear, we are not suggesting that the study of Indian Buddhist philosophy can or should abandon reference to schools. It is important to recognize that doxographical classification, identification with a particular school, and refutation of opposing Buddhist and nonBuddhist schools were all key elements of Indian Buddhist philosophical practices, especially in later periods. Moreover, as these practices spread beyond India, so did the role of schools in providing philosophical frameworks, authority, and identity for Buddhist philosophers. In Tibetan Buddhist scholastic curricula, school affiliation is still central to the study of Indian Buddhist philosophy. Leaving schools out of the equation is therefore not an option, and the reader will find that nearly every contribution in this volume makes reference to the various schools and sub-schools of the rich Indian Buddhist doxographical tradition. Organizing a study of Indian philosophers by schools is thus not without justification. Indeed, many scholars of Indian Buddhism, both traditional and modern, have chosen to do so for a variety of reasons. Jan Westerhoff, for example, partially chooses this route in The Golden Age of Indian Buddhist Philosophy, though he also acknowledges that much of the seemingly clear division between the different schools is an ex post facto arrangement, and that the individual thinkers concerned would have been unlikely to ascribe themselves to the specific schools they are supposed to have belonged to quite so readily. (2018, 10) To make up for some of what is lost through an organization based on schools, Westerhoff introduces sections dedicated to individual philosophers, where he also notes some of the ways such thinkers transcend any easy identification with a single school. Amber D. Carpenter, in her book Indian Buddhist Philosophy (2014), also takes something of a hybrid approach. While the book narrates the progression of Indian Buddhist philosophical thought mainly through themes, important chapters are also dedicated to the Madhyamaka, Yogācāra, and epistemological schools, especially as exemplified by their most iconic representatives: Nāgārjuna, Vasubandhu, Dignāga, and Dharmakīrti. Thus, even as we recognize that some reference to schools is necessary in the study of Indian Buddhist philosophy, we need to be clear about what we mean by the term “school,” since the word can have several valences ranging from a concrete institutional reality to an 2

Introduction

abstract doxographical construct (see McClintock 2018). Many Indian Buddhist philosophers were engaged in the construction and reconstruction of schools as doxographical constructs, and some were no doubt interested in forming institutional structures to maintain particular philosophical views and lineages. But to structure our study of Indian Buddhist philosophy around these schools, however we understand them, risks giving the constructs more concrete reality than may be warranted. Doing so carries with it significant hermeneutic consequences, as it would obscure the dynamic philosophizing and original contributions of individual authors. Our focus on authors pushes back against the tendency to default to a history of reified schools – even as the individual authors themselves often make claims on behalf of a particular school affiliation. An alternative way to structure this Handbook might have been through themes, with contributions organized in sections such as metaphysics, language, epistemology, ethics, philosophy of mind, political philosophy, argumentation, and so on. Organizing the volume according to these contemporary subfields of the academic discipline of philosophy would, perhaps, have made the volume more accessible to academic philosophers, who are one of the primary audiences for this book. There are indeed advantages to this tactic, despite the challenge of figuring out which authors should be situated in which categories, as so many of the philosophers in this volume span multiple areas of philosophical thought. But those advantages would likely have a better chance at being realized had the scope of the inquiry been expanded to Buddhist philosophy beyond only India. Indeed, such a thematic approach is employed in A Companion to Buddhist Philosophy (Emmanuel 2013), which treats philosophical topics across the whole development of Buddhism in Asia. Thus, while that book does contain some short contributions on regional schools (i.e., Theravāda, Indian Mahāyāna, Tibetan Mahāyāna and Vajrayāna, and East Asian Buddhism), the bulk of the volume consists of chapters that take up specific themes in one or more of these regions so that the audience can productively read them against one another. Our decision to organize this Handbook by authors resembles more closely the approach taken by Jonardon Ganeri in The Oxford Handbook of Indian Philosophy (2017), which also focuses on figures in an effort to better highlight both chronology and innovation and to make space for nonconformists whose role in the development of philosophical thinking should not be underestimated. While that work treats Indian philosophy more generally, this Handbook somewhat artificially and with more focus highlights just those Indian philosophers who identify as Buddhists. We do so not because we imagine that Indian Buddhist philosophy can be understood apart from the broader study of philosophy in India (it most emphatically cannot!). Rather, we want to make space for our readers to engage in a philosophically rich encounter with a greater variety of Indian Buddhist philosophers in a single volume than has been possible to date. The decision to place the author at the center of inquiry in these contributions further raises the quite serious question: What is an author? While the notion of an author at first appears self-evident, closer examination reveals a complex and shifting ground. Foucault’s proposal of an author function is instructive here, as it reminds us that an author’s name is not the same as a proper name that points to an individual. Rather, an author’s name performs a set of functions that serve to classify, define, authenticate, and establish relationships among diverse texts. By identifying a text as the product of a particular author, one also indicates something about how that text is to be read and received. Thus, in specifying that the contributions in the Handbook are to be organized by authors, we are pointing not only to individual historical figures but also to powerful authorial figures whose name and reputation have been associated with particular ideas, texts, practices, lineages, schools, and spiritual attainments. We believe that by attending to these authorial figures we gain a much richer understanding of the development and practice of Indian Buddhist intellectual traditions. 3

William Edelglass, et al.

Having made the decision to organize this Handbook by authors, we were still confronted by how to choose and how to arrange the philosophers. Although we lack concrete dates for the majority of figures in this volume, there is general agreement about a relative chronology based on the ways prior texts and thinkers seem to be presupposed by a given author. We therefore decided to proceed chronologically when possible, with the understanding that each subsequent philosopher could be seen, in part, as responding to philosophers who had come before. At the same time, there are some authors who we believed needed to be placed into special groups based less on schools or time periods but more on the genres in which their philosophical material appears. The book opens with three parts organized by genre before moving on in the next five parts to a more strictly chronological approach employing a somewhat arbitrary periodization. Here again, however, we continue to prioritize the author as the central organizing principle. We begin, perhaps somewhat unusually, with a part on buddhas as philosophers, based on our contention that buddhas should be considered philosophical author figures insofar as they may be credited with initially propounding – at least in our cosmic era – the foundational doctrines and ideas that serve as springboards for later Buddhist thinkers. This is true despite the fact that the various buddhas may appear not to agree with each other (or, for that matter, with themselves, given the diversity of texts attributed to the same buddha) and may also appear not to be historical persons. The ideas of these buddha philosophers often emerge embedded in dialogical scriptures which typically have significant narrative elements. Thus, the first part of the book focuses on the genre of buddhavacana (“speech of the/a buddha”) as a modality through which Buddhist authors – in this case, buddhas themselves – make known their philosophical ideas. We next move on to another genre with a part on poet philosophers. Indian Buddhist philosophers composed their works in a wide variety of literary genres, with many of them employing poetry, narratives, and dialogues as well as treatises and commentaries. For some, the poetry, narrative, and dialogue do important philosophical work, as might also be said of the way that literary form has a philosophical function in Plato’s dialogues, Montaigne’s essays, and Nietzsche’s aphorisms. In this part, we focus on poet philosophers whose philosophy can be said to be embodied in their chosen literary form and whose philosophically significant meaning would be lost if it were articulated in the more abstract language of a treatise. The final part organized primarily on genre is concerned with abhidharma philosophers. Here we are often thwarted in highlighting the specific contributions of individual authors due to the fact that the texts themselves appear to have multiple authors. In this chapter, then, individual texts often take on the author function. But by resisting the urge to talk about “Abhidharma” or “Abhidhamma” as a school and focusing instead on the individual texts or textual traditions, we come as close as we can to highlighting the contributions of particular authors even when we are lacking a name. At this point, the parts shift and begin to focus more obviously on the iconic founding figures of Indian Buddhist philosophy. We start with a unit on philosophical founders, by which we mean philosophers whose names have become associated with the founding of a particular stream or school. The part begins with Nāgārjuna and his student Āryadeva, and is filled out with chapters on the remaining four figures making up the “six ornaments” according to Tibetan Buddhist scholasticism: Asaṅga, Vasubandhu, Dignāga, and Dharmakīrti. Following these formative figures, we consider a host of commentator philosophers, by which we mean authors who in many cases literally and in other cases more figuratively position themselves as commentators on the thought of one or more of the authors treated earlier in the volume. Because there were a great number of these, we broke them into three major 4

Introduction

groups: early period, middle period, and late period commentators. In this way, we made room for thinkers from the fifth–twelfth centuries, bringing us close to the end of the flourishing of Buddhist philosophy in India. Throughout these chapters, we find authors who are bending, contesting, reformulating, and repudiating various doxographical labels while also promoting new ideas in the domains we mentioned earlier: metaphysics, language, epistemology, ethics, philosophy of mind, political philosophy, argumentation, and so on. The volume ends with a short part that is something of an outlier in that it treats modern Indian Buddhist philosophers. Here we have just two figures, both of whom, however, have played an outsized role in revitalizing Buddhist philosophy in India and beyond. The first is B. R. Ambedkar, whose philosophical writings on Buddhism, religion, and politics have become influential among practicing Buddhists – especially in India – and scholars alike. The second is His Holiness, Tenzin Gyatso, the Fourteenth Dalai Lama of Tibet. While not an Indian by birth, the Dalai Lama has spent more than half a century living in exile in India, where he has consistently engaged with and emphasized the Indian heritage of Tibetan Buddhist philosophy. He has also been a key figure in the promotion of what he calls “the Nālandā tradition,” a style of doing philosophy with deep roots in the thought of many of the figures found in this volume. The chapters that follow illustrate the broad diversity of Indian Buddhist philosophy. They also exemplify the many ways in which contemporary scholars are engaging with Indian Buddhist philosophers and their texts, for our contemporary authors are as diverse as the Indian Buddhist authors treated in the volume, and their chapters showcase a range of approaches, some more historical, some more philosophical, some more literary, some more technical, some more playful. Like the Indian Buddhist authors who are the subject of this volume, our contributors have also inherited intellectual practices and views that are manifest in their own writings. Still, despite their differences, each chapter can serve in its own way as a window onto the major philosophical innovations and ideas of the figures we have chosen for this volume as seen through the eyes of our particular contributors.

References Carpenter, Amber D. 2014. Indian Buddhist Philosophy. London: Routledge. Emmanuel, Steven M., ed. 2013. A Companion to Buddhist Philosophy. Blackwell Companions to Philosophy. New York: Wiley-Blackwell. Ganeri, Jonardon, ed. 2017. The Oxford Handbook of Indian Philosophy. Oxford: Oxford University Press. McClintock, Sara. 2018. “Schools, Schools, Schools – Or, Must a Philosopher Be Like a Fish?” In Buddhist Spiritual Practices: Thinking with Pierre Hadot on Buddhism, Philosophy, and the Path, edited by David Fiordalis, 71–103. Berkeley: Mangalam Press. Westerhoff, Jan. 2018. The Golden Age of Indian Buddhist Philosophy. Oxford History of Philosophy. Oxford: Oxford University Press.

5

PART 1

Buddhas as Philosophers

BUDDHAS AS PHILOSOPHERS Introduction to Part 1

The opening part of this Handbook proceeds on the premise that considering buddhas as philosophers is a worthy and interesting endeavor. This volume also argues that approaching Buddhist philosophy through authors yields insights not available through other approaches. Yet buddhas are often not recognized as authors in the usual sense of the term. Rather, they are teachers whose discourse survives through the recollections of their disciples as preserved in written and oral texts. These recollections, moreover, do not consist merely of the verbatim contents of the buddhas’ teachings. Rather, they typically narrate a scene of instruction in which a buddha appears as one actor among several, responding to the questions and circumstances of various interlocutors. Thus, whatever philosophy a buddha has taught must be accessed indirectly, much as what we learn of Socrates comes through witnessing his conversations with sundry characters in the dialogues of Plato. For this part, we have chosen buddhas who are primordial, of this age, or of a future age. Their status as authors might not be obvious. However, the word “author” is etymologically related to “authority” (they are both derived from auctoritas), and in Buddhist traditions, buddhas are precisely those who have the most authority. Such authority is grounded in their long moral training and profound wisdom, their collection of merit, and their vision of reality as it truly is. While they themselves are not authors composing texts, their speech is recorded by others who attest to its accuracy and authenticity on the basis of their authority and thus make the buddhas authors as sources of the teaching. The words of these buddhas are embedded in narrative and doctrinal tracts full of rich philosophical content. As such, these buddhas – both those for whom we have seeming historical evidence and those whose identities come into greater focus only in literary or ritual milieus – can be considered foundational authors whose works generate and are justified by philosophical thinking. We start with Gotama Buddha, the buddha of the early discourses, oral texts that were eventually written down and that today form important root sources for the Theravāda tradition. Bhikkhu Anālayo, in his chapter, pieces together philological evidence from these discourses in Pāli and their parallels preserved in other languages to paint a picture that, as he writes, “takes us as close as possible to the historical Buddha” while acknowledging that this picture depends not on historical facts but rather on representations of the teacher Gotama Buddha as found in the surviving texts. Philosophically, this buddha appears most interested in how to attain the “deathless,” a term that for Gotama indicates not a state of eternal life but rather DOI: 10.4324/9781351030908-3

9

Buddhas as Philosophers

“a supreme condition of freedom of the mind such that even the terror of mortality has completely lost its sting.” The picture painted here is of a philosopher of deep pragmatic concern, whose teachings therefore involve explicit instructions for undertaking the contemplative practices designed to lead one to this state of radical freedom. Next, we move to Siddhārtha Gautama, a figure who might be considered the same as Gotama Buddha but whose representations in the Sanskrit discourses associated with the Mahāyāna stream of Indian Buddhism yield a rather different personage. Richard F. Nance tackles the vexed question of the authorship of these Sanskrit discourses through a sustained analysis of their understanding of what counts as buddhavacana, the “word of the/a buddha.” In contrast to the previous chapter, which sought to bring us as close as possible to the historical buddha, this chapter urges us to look “beyond the historical figure” to recognize that buddhavacana may, in the end, have very little to do with the question of provenance and much more to do with the effects that such inspired speech may have on particular audiences. In a manner quite different from that of Bhikkhu Anālayo, then, Nance arrives at a similar conclusion in urging us to read Siddhārtha Gautama’s philosophy more in terms of its enactment in embodied practice than as a set of philosophical doctrines. The third buddha under consideration is the tantric Buddha, the quintessential “primordial buddha” (ādibuddha) who appears in innumerable peaceful and wrathful forms and with a great plethora of names and attributes within the tantric streams of Indian Buddhism. Taking us still further away from “historical fact” toward something transcendent, Vesna A. Wallace explains in her chapter that the buddhas who author the texts known as tantras are traditionally understood as “nothing other than the ultimate truth” which reveals itself “in the conventional and at times symbolic terms that constitute a tantric text.” The collapse of author and text is here complete, a situation that itself raises a complex of philosophical questions. Yet, these texts also contain explicitly philosophical passages, and Wallace helpfully guides us through some of them. The chapter reveals the tantric Buddha as engaging in philosophical disputation and offering arguments to refute non-Buddhists and Buddhists alike. Yet, it is also clear that this is but one small portion of a buddha’s activity. Finally, we turn to the future Buddha, Maitreya, whose role as an author is particularly complex due to his relationship with a human author, Asaṅga (who is also treated later in this volume). Although he is technically a bodhisattva, we include Maitreya in this first part due to his transcendent and celestial status. Maitreya is a figure of great optimism for many Buddhists, as it is foretold that he currently resides in a heavenly realm, Tus․ ita (“Joyful”), waiting for the moment when all traces of the previous Buddha’s teachings have disappeared from the earth. At that point, Maitreya will take birth in human form, attain nirvāṇa, as did Śākyamuni Buddha before him, and reintroduce the Dharma to the world. Meanwhile, though, it is possible for some advanced meditation practitioners to meet Maitreya  – as we find in the case of Asaṅga. In his chapter, Klaus-Dieter Mathes relates how Asaṅga is said to have received oral texts from Maitreya and then written them down. These texts include analyses of buddha-nature and buddhahood, as well as the path of transformation from ordinary mind to insight into the true nature of phenomena. While they have sometimes been interpreted as consistent with Madhyamaka philosophy, they generally present what came to be identified as a Yogācāra Buddhist view. Mathes carefully examines one popular corpus of five Maitreya texts to conclude that there is sufficient evidence to hold that these works were all composed “by a single hand.”

10

1 GOTAMA BUDDHA His Quest and His Teachings Bhikkhu Anālayo

Introduction The main purpose of the present chapter is to delineate the figure of Gotama Buddha as a seminal ancient Indian thinker and practitioner in the way he is depicted in the “early discourses,” an expression whose implications I will explore in more detail in the next section. The information that can be gathered from a comparative study of these textual materials takes us as close as possible to the historical Buddha. Nevertheless, it is necessary to keep in mind that this kind of evidence does not enable reconstructing with certainty historical facts about an individual who lived in ancient India.1 To do so is not possible, given the limitations of the sources at our disposal. For this reason, the reconstruction presented here is only a reflection of early layers of textual depictions of Gotama Buddha. References given in the present chapter to his thought or practice should for this reason invariably be understood to intend descriptions or representations of his thought or practice in the textual sources without any implicit claim to be reporting definite facts. At the same time, it also needs to be borne in mind that these textual depictions have had a lasting impact on later generations of Buddhists, and to that extent, they are themselves historical facts. Even though the actual person who would have walked on the soil of India some 2,500  years ago is beyond our reach, textual descriptions of Gotama Buddha, which have made him a source of lasting inspiration for Buddhists ancient and modern, are within our reach. Their historical-critical study enables us to discern layers in the development of such descriptions and thereby understand the growth of Buddhist thought and practice. In the end, this is perhaps even more historically significant than a quest for certainty about historical facts regarding the person who lived in ancient India. A convenient way of approaching the early textual depictions of the Buddha’s thought and practice is through his quest for, and eventual attainment of, awakening, examined from the viewpoint of how these events relate to key aspects of early Buddhist teachings. The received biography of the Buddha, however, is of rather limited use for this purpose, as it comes inexorably intertwined with later hagiography.2 Hence, in what follows, I will employ only material from the early discourses in order to position my presentation within the earliest period of Buddhism still accessible to us in the present time.

DOI: 10.4324/9781351030908-4

11

Bhikkhu Anālayo

The overall concern of these early discourses is what the Buddha taught rather than what he did as an individual. This type of textual material does not even offer a continuous biography but only biographical snippets here and there, used in order to illustrate or clarify some aspect of the doctrine. Thus, the material employed here is not really biographical and its main purpose in its original setting is to clarify philosophical points, which is precisely the way it will be used in the following pages. I begin with a brief introduction to the source material to be used in this chapter. Then I examine passages that have a bearing on the Buddha’s progress to awakening and hence the unfolding of his philosophical and soteriological contributions, in the way this is reported by the texts. Precise dating and historical accuracy remain perennial problems for ancient Indian texts. For this reason, what I present here is confined to an emic perspective, in the sense of gathering together what the early Buddhist texts considered to have been key aspects of the Buddha’s realization, and in what ways, according to the same texts, these key aspects differed from related conceptions held by other thinkers and practitioners of his time. My reference to the thought and practice of Gotama Buddha is deliberate. It is intended to signal that my exploration needs to do justice to the fact that the Buddha’s thought, as a contribution to ancient Indian philosophy, is inseparably connected to a corresponding mode of living and practicing. It is, in a way, neither strictly philosophical nor strictly religious. Instead of fitting smoothly into one of these two categories, it can perhaps best be seen to combine, in a pragmatic manner, dimensions of both under the overarching aim of providing a viable answer to the human predicament. From the emic perspective, such an answer needs to be both realistically practicable and philosophically coherent. As a starting point for exploring the Buddha’s quest, I consider his motivation to go forth from lay life to become a renunciant, followed by examining depictions of his temporary overcoming of mental defilements in order to gain deep concentration experiences. Then I turn to his pursuit of ascetic practices. Next, I take up three knowledges he reportedly attained on the night of his awakening and examine the textual records of his decision to teach, together with exploring the content of what tradition regards as his first teaching. The passages from the early discourses examined in this way bring out central teachings of early Buddhist philosophy-cum-soteriology. These include a redefinition of the “deathless”; an analytical attitude toward deep concentration experiences; a re-evaluation of the notion of ­Nirvana and of the pleasure/pain paradigm by introducing the crucial ethical distinction between what is wholesome/skillful and its opposite; insight into conditionality and not self (placed within the context of rebirth); the importance of mindfulness and the awakening factors for meditation practice; the role of compassion in relation to the teaching activity of a Buddha; and the apparent adoption of a medical scheme of diagnosis to formulate the core teaching of the four noble truths. In conjunction, these various aspects converge on the role of the Buddha as a remarkable teacher who has remained a source of inspiration for generations upon generations of Buddhists from ancient India to contemporary times.

Source Material The “early discourses” that are the source material in the ensuing exploration are the final results of centuries of oral transmission of texts believed to have been spoken originally by the Buddha and his chief disciples in their various encounters with followers, visitors, and opponents in the ancient setting. These orally transmitted discourses eventually reached Sri Lanka, where they were subsequently committed to writing in an Indian language called Pāli. According to the traditional report in the Ceylonese (Sri Lankan) chronicles, the Pāli discourses were written down shortly before 12

Gotama Buddha

the beginning of the Common Era (Adikaram 1946/1994, 79). By then, the main philosophical content of the Pāli discourses appears to have been fairly closed, leaving aside the minor variants that tend to occur naturally in written transmission. Hence, these materials can be reckoned as testimonies for the developments in Buddhist thought up to that time (Anālayo 2012a). Other Buddhist oral transmission lineages in India also preserved their records of the Buddha’s teachings (Salomon 2018, 52f), and the texts passed on orally by these lineages were also written down around the same time as the writing down of the Pāli discourses in Sri Lanka or soon thereafter. Due to climatic conditions in most of the Indian subcontinent, written records of the teachings, made on such fragile material as palm leaves, are in constant need of being recopied. With the eventual disappearance of Buddhism from India, the process of copying manuscripts came to an end, and much of the early material that had been committed to writing there was lost. In Sri Lanka, however, the Pāli manuscript tradition continued to be maintained. Early Buddhist discourse material has also been preserved in manuscripts that mostly come from Central Asia. Due to the dry climatic conditions in Central Asia, texts written on material like birch bark endure much longer than palm-leaf manuscripts in Sri Lanka or India. As far as we know, the discourse collections also reached Tibet, but the resultant translations into Tibetan appear to have been lost during a period in which Buddhism declined. Hence, the Buddhist canon preserved in Tibetan translation no longer has counterparts to the four main discourse collections, but only to some selected discourses preserved either on their own or else as citations in other works. Collections of discourses from different Buddhist oral transmission lineages had also been brought to China and translated into Chinese. The large body of Buddhist texts translated into Chinese contains counterparts to each of the four Pāli discourse collections. These are the Chinese Āgamas (Anālayo 2015a), each of which stems from one of the various oral transmission lineages that attempted to preserve the teaching of the Buddha and that served as the basis for the canonical textual collections of Buddhist monastic traditions. In the case of discourses extant in the Chinese Āgamas, there is, of course, the problem of translation errors. Rendering an Indic text into Chinese involves working with two languages that are substantially different from each other, making it a rather demanding task to transpose content from one language to the other without a change or even a loss of the meaning. Nevertheless, the discourses in the Chinese Āgamas are not themselves products of Chinese culture but are instead testimonies of early Indian Buddhism. Recourse to the Chinese Āgamas enables comparing different versions of a particular discourse extant in Indic language(s) and thereby identifying and potentially rectifying transmission errors. Such comparison of sources provides the main foundation for my exploration of the thought and practice of Gotama Buddha.

In Quest of the Deathless A background to the future Buddha Gotama’s decision to go forth in quest of the deathless emerges from a reflection attributed to him in a Pāli discourse and its Chinese Āgama parallel (AN 3.38 and MĀ 117; trans. in Bodhi 2012, 240 and Anālayo 2017d, 6, respectively).3 The two separately transmitted versions of this discourse agree in reporting that Gotama was profoundly affected by seeing others experiencing sickness, old age, and death: these sights made him realize that he was subject to the same predicament himself. The texts explicitly present this as a type of reaction distinct from the usual response by ordinary people, who tend to turn away in disgust when encountering others being sick, growing old, and passing away, thereby overlooking their own situation. 13

Bhikkhu Anālayo

The different outlook that emerges in this way is a departure not so much from the ancient Indian philosophical perspective but from ordinary ways of thinking. For this reason, it remains quite relevant to any time and society, including our own. The challenge of facing disease, old age, and death is a perennial one for humanity and has motivated a range of responses in philosophy and religion. The distinct early Buddhist perspective on the matter, in the way it emerges from the passage under discussion, is the need to face squarely these predicaments of human life in full recognition that they apply to all of us (Anālayo 2016b). The need to address the existential question posed by these fundamental aspects of human life recurs in another Pāli discourse and its Chinese Āgama parallel. According to this passage, the Buddha described to his disciples what had motivated him to go forth: the realization that – being himself subject to disease, old age, and death – it was not appropriate to be spending his life seeking what is also subject to these. For this reason, he decided to dedicate his life to the quest for that which leads beyond disease, old age, and death (MN 26 and MĀ 204; trans. in Ñāṇamoli 1995/2005, 256 and Anālayo 2012b, 25, respectively). Both texts mark out that Gotama was still in the prime of his youth at the time of taking this decision and setting out on such a quest. The two texts in question also agree in reporting that Gotama went forth even though this made his parents weep. This serves to clarify an element in later hagiography, according to which his going forth took the form of a secret departure at night. This tale is the outcome of later narrative developments and is at odds with the testimony of the early discourses. It might be the result of a mode of storytelling in ancient India that involved the employment of a kind of canvas to illustrate key aspects of a narrative during oral performance. In such a setting, a pictorial depiction of Gotama’s decision to go forth could have been the starting point for these hagiographic developments, in the sense that the graphic illustration was taken literally and then influenced textual accounts. The two early discourses just mentioned, which record the Buddha describing to his disciples what had motivated him to go forth, also report that, upon attaining awakening, he proclaimed to have attained the “deathless” (amata/amṛta). The usage of this expression to denote the successful completion of his quest is significant in view of its connotations in the ancient Indian setting as a referent to a form of immortality or eternal existence (Collins 1982, 43; Vetter 1995, 217). The attainment of the deathless in early Buddhist thought is not the achievement of a state of eternal life. Instead, it signifies a liberating insight that results in the complete conquest of any fear of death. With full awakening reached, the Buddha and those of his disciples who had also reached complete liberation from defilements of the mind are considered to have reached a condition of the mind that will remain completely composed in the face of their own passing away (or that of others). In this way, their conquest of death neither avoids the passing away of their physical body nor leads to gaining a state of immortality in a heavenly realm. Instead, it involves a supreme condition of freedom of the mind such that even the terror of mortality has completely lost its sting.

Analysis of Absorption Returning to the account of the future Buddha Gotama’s quest for the deathless, the relevant material for reconstructing an early record of his progress to awakening occurs in various early discourses that unfortunately do not stand in an explicit chronological order to one another. For this reason, at times it is uncertain whether a particular episode should be placed earlier or later in relation to other episodes. This holds for a description of Gotama’s struggle with a range of mental obstructions that prevent the attainment of deep concentration in the form of 14

Gotama Buddha

absorption ( jhāna/dhyāna). Although the commentarial tradition considers this description to report events that took place only during the night of the Buddha’s awakening, the relevant discourses contradict this directly: they describe a mode of practice involving several days and nights, which cannot be confined to what happened in a single night (Anālayo 2011, 741 n. 270). This makes it probable that the episode in question should be placed at an earlier time, well before the actual event of awakening. The episode itself takes its occasion from a visit paid by the Buddha to a group of seriously meditating monastics. During their practice, these monastics had been encountering inner experiences of light, which subsequently disappeared again. The Buddha is on record for informing them that the same had happened to him during the time before his awakening. At that time in the past, he had identified various mental obstructions that needed to be overcome in order to prevent the inner light disappearing. Overcoming these obstructions eventually led to his successful attainment of absorption (MN 128 and MĀ 72; trans. in Ñāṇamoli 1995/2005, 1015 and Anālayo 2017d, 35, respectively). The description of the attainment of absorption, given in the Pāli and Chinese versions of the discourse in question, does not employ the standard pericope for such description found elsewhere among the early discourses. Instead of listing the four levels of absorption mentioned in the standard pericope, it presents three. The resultant two modes of description, enumerating either three or four levels of absorption, both involve an analysis of the mental factors present during absorption attainment.4 The main concern behind the threefold model found in the description of Gotama’s own pre-awakening struggle with obstacles to the gaining of deep concentration appears to be the degree of effort required for attaining absorption and remaining in it. This degree of effort gradually diminishes as concentrative absorption deepens. At first, more effort is required for directing the mind and keeping it in its collected condition. Then directing becomes effortless and the mind only needs to be kept where it is. Eventually, even that is no longer necessary. The standard pericope instead presents a fourfold analysis of absorption. The first level involves an experience of joy and happiness born of seclusion from sensual distraction and unwholesome states of mind. The second level concerns an experience of joy and happiness born of concentration proper. With the third level of absorption, joy is left behind and only happiness remains. Happiness is then also transcended with the fourth level of absorption, an experience that is characterized by deep equanimity. The differences between these two modalities of analysis, which survey what in actual practice would be the same process of gradually deepening concentration, appear to reflect different concerns: whereas the threefold analysis places more emphasis on the gradual reduction of effort, the fourfold analysis gives more room to the hedonic dimension of absorption experience. Now, an emphasis on the degree of effort required to enter deep concentration would have been a natural way of analyzing absorption attainment as part of Gotama’s progress to awakening, particularly in view of his own experience of needing to overcome a range of mental obstructions. Several of these, in fact, require a fine-tuning of effort such that the mind is neither too lax nor too tense. Giving priority to the type of joy and happiness present or absent during different levels of absorption instead seems to relate to a crucial understanding gained by Gotama subsequently, after he had given up an attempt to reach awakening through the pursuit of ascetic practices (a topic to which I will return in what follows). It seems that the realization of the important contribution offered by wholesome types of joy and happiness to progress to awakening may have informed the mode of presentation found regularly in teachings given on other occasions, which take the form of the fourfold analysis. 15

Bhikkhu Anālayo

Both schemes mark a departure from the way absorption attainment appears to have been conceived by some meditators in the ancient Indian setting. Judging from the testimony of other discourses, experiences of absorption led some non-Buddhist practitioners to a range of speculative views and metaphysical conclusions about the nature of the individual and the world (DN 1 and DĀ 21, as well as other parallels; trans. in Walshe 1987, 73–87 and Anālayo 2017b, 122–70, respectively). In contrast, the two modes of analyzing absorption under discussion eschew any metaphysical conclusions and instead focus squarely on the condition of the mind itself. This condition is to be examined closely in order to discern the principal mental factors responsible for absorption to occur. Such analysis offers clear directives on how this experience can be repeated and also points directly to its conditioned nature. In this way, it can stimulate the cultivation of two meditative qualities that in early Buddhist thought collaborate (Anālayo 2003, 88–91): tranquility (samatha/śamatha) and insight (vipassanā/vipaśyanā).

Immaterial Attainments and Nirvana Presumably subsequent to having successfully attained absorption in the way described in the episode concerning a range of mental obstructions to inner experiences of light, Gotama continued to deepen his cultivation of the mind under the tutelage of two contemporary teachers, Āḷāra Kālāma and Uddaka Rāmaputta, whose teaching involved the meditative experiences of “nothingness” and of “neither-perception-nor-non-perception.” The relevant Pāli and Chinese discourses report that he was soon able to reach a personal experience of the meditative attainments on which the teachings of Āḷāra Kālāma and Uddaka Rāmaputta were based (MN 26 and MĀ 204; trans. in Ñāṇamoli 1995/2005, 257 and Anālayo 2012b, 27, respectively). In the standard accounts of meditation practice in other discourses, these meditative attainments find a placing as the third and fourth of four “immaterial” (arūpa) experiences. The whole set of four immaterial attainments requires leaving behind any perceptions related to the material world. They take as their respective objects the perceptual experience of infinite space, infinite consciousness, nothingness, and neither-perception-nor-non-perception. Their cultivation requires the concentrative strength of the fourth absorption (in the fourfold scheme discussed previously). The teaching of Āḷāra Kālāma was concerned with the immaterial attainment of nothingness; the teaching of Uddaka Rāmaputta was concerned with neitherperception-nor-non-perception (an attainment of such subtlety that one is neither really perceptive nor actually impercipient). The discourses report that both teachers acknowledged Gotama’s success to the extent of inviting him to take on a teacher’s role himself. Gotama is on record as declining the offer, as he did not consider these attainments to provide the answer to his existential quest. This thereby marks another departure from the ancient Indian setting, in as much as the notion of the final goal is concerned. Although the respective attainments must at first have seemed promising enough to motivate Gotama to choose Āḷāra Kālāma and Uddaka Rāmaputta as his teachers, their actual experience failed to provide him with what he was seeking: a definitive solution to the human predicament of being subject to old age, disease, and death. The solution he eventually found, on the night of his awakening, involves a reappraisal of the notion of Nirvana. The survey of metaphysical interpretations of absorption experiences, mentioned previously, covers some relevant notions, presumably held by other practitioners in the ancient Indian setting. These are on record for identifying sensory enjoyment or absorption experiences as forms of realizing Nirvana (DN 1 and DĀ 21, as well as other parallels; trans. in Walshe 1987, 85–86 and Anālayo 2017b, 168–70, respectively). Another passage even reports that bodily health was identified with Nirvana (Anālayo 2011, 410). 16

Gotama Buddha

The implications of the same term in early Buddhist discourse differ, as the central idea here is a complete transcendence of all modes of being and existing. The realization of Nirvana attained with full awakening, besides being a realization of the deathless, can also be seen as a realization of the “birthless,” so to say. This is meant in the sense that it leads beyond the prospect of any future rebirth in saṃsāra, the endless cycle of births and deaths that – according to ancient Indian thought – holds in bondage ordinary humans and other living beings. The novelty of the idea of a middle-way approach by way of a complete transcendence of saṃsāra that neither involves an alternative form of existence nor is mere annihilation forms a recurrent topic in the discourses. Time and again, these discourses show the Buddha and his disciples in discussion with contemporary practitioners unable to understand what the final goal of early Buddhist practice was about. The attainments of nothingness and neither-perception-nor-non-perception, in spite of their profound nature, were for this reason insufficient to provide an answer to Gotama’s quest. They only led to rebirth in a corresponding heavenly realm and hence failed to be the “birthless” solution that Gotama was seeking.

Ascetic Practices and the Pleasure/Pain Paradigm Whereas the chronology of the account of Gotama’s absorption practice in relation to other episodes relevant to his progress to awakening is open to different interpretations, the early discourses provide clear indications regarding what happened after he left his two teachers, Āḷāra Kālāma and Uddaka Rāmaputta. In his continued quest for what leads beyond old age, disease, and death, Gotama is on record for engaging in some of the ascetic practices that were in vogue in ancient India. Closer inspection of relevant textual material brings to light that some of these descriptions appear to refer to experiences he had in a previous life, spent as an ascetic (Anālayo 2011, 116). Hence, of direct relevance to the account of his progress to awakening in his final life, after having left his two teachers, appear to be only three ascetic practices: forceful control of the mind, breath retention, and fasting. An account of these three modalities of asceticism can be found in a Pāli discourse and a parallel extant in Sanskrit fragments (MN 36 and Liu 2010; trans. in Ñāṇamoli 1995/2005, 337 and Anālayo 2017d, 56, respectively). The attempt to control the mind by sheer force takes the form of firmly clenching the teeth, pressing the tongue against the roof of the mouth, and then trying to beat down and constrain the mind. The idea could be that, since even the profound meditative experiences of nothingness and neither-perception-nor-non-perception had not led to liberating the mind from the bonds of saṃsāra, the alternative approach was to face things head on by trying to force the mind into becoming liberated from any bondage. Once this approach had not brought tangible results, a natural extension of the same attitude was breath retention. In fact, clenching the teeth and pressing the tongue against the roof of the mouth already come fairly close to such a mode of practice. The sources agree that Gotama tried out different modalities of breath retention, yet none of these led him to the goal of his aspiration. From attempting to reduce his intake of air to a minimum, the next step was an attempt to reduce his intake of food to an absolute minimum. This resulted in a serious weakening of his body to the extent of bringing him close to death, yet it did not bring about the desired liberation. The failure of all three ascetic practices to result in progress to awakening led Gotama to reconsider his approach. This made him realize that wholesome types of pleasure need not be 17

Bhikkhu Anālayo

shunned. The resultant shift of understanding motivated him to abandon asceticism and paved the way for his attainment of awakening, once his body had regained strength. The decisive shift of perspective at this juncture reflects a crucial undercurrent of early Buddhist thought and practice, which is the foundational ethical distinction between what is wholesome or skillful (kusala/kuśala) and what is unwholesome or unskillful (akusala/ akuśala). The new viewpoint requires stepping out of the ordinary pleasure/pain paradigm. Based on a thoroughgoing emphasis on what happens in the mind, this novel outlook gives importance to the ethical repercussions of any experience over its affective quality. The ordinary untrained mind is prone to pursue what is pleasant and avoid what is painful. In recognition of the bondage of sensual indulgence that ensues from this, ascetic ideology tends to advocate the pursuit of self-inflicted pain and the concomitant avoidance of pleasure. Leaving behind both attitudes as two extremes to be avoided, the solution proposed in early Buddhist thought takes the form of a middle path that emphasizes the ethical quality of any experience. This ethical quality relates to the intention that stands behind that experience. The resultant perspective reveals that some agreeable experiences are obstacles to progress to awakening, such as sensual indulgence, but other pleasant experiences can support such progress, such as the pleasure and joy of deep concentration. The same distinction holds for painful or neutral experiences, some of which are detrimental whereas others are beneficial. In this way, the evaluation of experience is decoupled from the pleasure/pain paradigm and instead oriented along the distinction between what is wholesome/skillful and what is unwholesome/unskillful (MN 70 and MĀ 195; trans. in Ñāṇamoli 1995/2005, 579 and Anālayo 2017d, 75, respectively). What is wholesome or skillful should be pursued, regardless of whether it is pleasant or unpleasant. Conversely, what is unwholesome or unskillful should be avoided, no matter how pleasant (or unpleasant) it might be. Moreover, the emphasis shifts from the adoption of certain modes of external conduct and behavior to the mental state that informs them. Of crucial importance here is the question of intention (cetanā). This is where the distinction between wholesome/skillful and unwholesome/unskillful really matters. Intention is what sets the course for what is then undertaken by speech and action. The same ethical distinction can be seen at work in relation to the three modalities of asceticism described previously. Even though Gotama stopped his ascetic practices and considered them as not conducive, on their own, to liberation, he nevertheless retained specific aspects of each mode of conduct in his instructions to his followers. The idea of overpowering the mind by sheer force has found a place in series of strategies for overcoming unwholesome thoughts, where it comes as a last resort when all else has failed (MN 20 and MĀ 101; trans. in Ñāṇamoli 1995/2005, 213 and Anālayo 2013, 153, respectively). In this way, from being the sole approach to gaining control over the mind, the use of force is relegated to being an emergency measure for preventing the occurrence of unwholesome actions, rather than being completely discarded. The potential of attending to the process of breathing takes the form of detailed instructions on the cultivation of mindfulness of inhalations and exhalations, which differ from breath retention by eschewing control and instead giving prominence to non-interfering mindful observation (MN 118 and SĀ 803; trans. in Ñāṇamoli 1995/2005, 943 and Anālayo 2013, 228, respectively). The undertaking of fasting has continuity in instructions to monastic disciples on the observance of intermittent fasting, in the sense of taking only a single meal per day (MN 65, MĀ 194, and EĀ 49.7; the first trans. in Ñāṇamoli 1995/2005, 542 and the last in Anālayo 2016a, 90).

18

Gotama Buddha

In this way, the shift of perspective evident in the account of Gotama’s progress to awakening spells out in a middle path that is able to accommodate aspects of the three types of asceticism to the extent that these can facilitate progress to awakening. What counts – here and elsewhere in early Buddhist thought – are the ethical repercussions, in terms of fostering what is wholesome/skillful and avoiding its opposite.

Recollection of Past Lives and the Construction of Personal Identity With strength regained, after having given up the practice of fasting, Gotama is on record for having cultivated a recollection of his own past lives during the first part of the night of his awakening (MN 4 and EĀ 31.3; trans. in Ñāṇamoli 1995/2005, 105 and Anālayo 2016a, 21, respectively). In the narrative context of Gotama’s progress to awakening, such turning to the past stands in continuity with his earlier review of the previous practices he had undertaken in this life in order to gain awakening. This retrospection had led him to the realization that wholesome forms of joy and happiness need not be shunned. The benefits derived from looking back at his own past experiences might have inspired his decision to take the same approach further by way of recollecting his previous lives. The ability to recollect previous lives, a meditative practice based on mindfulness, features in the early discourses as being accessible to those who have reached mastery of the four absorptions. It involves being able to recall one’s former name and living circumstances, the food eaten and the pleasures and pains experienced in a past life, as well as remembering the transition from one life to another. The inclusion of such knowledge among what, according to tradition, was one of the three key insights that brought about the Buddha’s awakening is only one in a range of passages that show rebirth to have been an integral part of early Buddhist thought (Anālayo 2018a). The description of such recollection of former lives does not spell out how this type of knowledge contributed to Gotama’s eventual awakening. In view of the centrality of the teaching on not self in early Buddhist doctrine, however, it seems fair to propose that the direct witnessing of one past life after another could have been a starting point for Gotama to discern the constructed nature of the sense of identity. In any of these former lives, which he apparently recollected vividly, a particular name would have been experienced as “me,” a certain family and living circumstances been taken to be “mine,” and specific food as well as pleasure and pain been perceived as “my” experiences. Yet, with a change of scene, so to say, from one life to the next, a different name would have become “me,” a different family “mine,” and so on. The notion of a continuity from one life to the next provides a needed perspective for a proper appreciation of the early Buddhist teaching on not self (anattā/anātman). This doctrine does not imply a denial of the reality of subjective experience. It only denies that there is a permanent and unchanging entity involved in such subjective experience. Instead, subjective experience is seen as nothing but a flow, the product of specific conditions and bereft of anything permanent. Hence, continuity from one life to another does not involve the reincarnation of a permanent entity or soul. Instead, it is considered to be simply a continuity of the flow of causes and conditions operative during life. A difference is that, at the time of death, these proceed independently of the physical body. An illustrative example is a flame which, with the support of wind, may ignite even what is at a distance (Anālayo 2018a, 22). The flame is just a process of combustion and does not require any permanent entity to remain afire. It continues as a changing process of burning as long as the required conditions are in place, and it is even able to ignite something not immediately contiguous to its fuel, as long as there are supportive conditions for that.

19

Bhikkhu Anālayo

The doctrine of not self is perhaps the most substantial departure of early Buddhist thought from its ancient Indian setting. In one discourse, this finds expression in an analysis of four types of clinging, which are clinging to sensual indulgence, to views, to rules and observances, and to the notion of a self (MN 11, MĀ 103, and EĀ 27.2; the first trans. in Ñāṇamoli 1995/2005, 161 and the last in Anālayo 2016a, 42). Whereas the first three modalities of clinging were according to this discourse within the purview of contemporary Indian practitioners and thinkers, the problem posed by clinging to a sense of self is seen as a unique contribution of the Buddha. The doctrine of not self is what makes his teaching exceptional for its comprehensive approach to the problem of clinging (see also Anālayo 2022).

The Divine Eye and Conditionality Recollection of one’s own past lives has its counterpart in the ability to witness the passing away and being reborn of other sentient beings, called the “divine eye.” These two knowledges stand in close relationship to each other, just as the doctrine of not self stands in a close relationship to the early Buddhist teaching on dependent arising ( paṭiccasamuppāda/pratītyasamutpāda). It is only for ease of presentation that I have taken up the notion of not self in relation to recollection of past lives and now turn to dependent arising. This is not meant to imply that these two doctrines are separate and can be matched precisely with the first and second knowledge gained according to the traditional report by the Buddha on the night of his awakening. Instead, both are best seen as concomitant dimensions of a gradually dawning insight that found its completion in the third knowledge, the realization of Nirvana as the actual event of awakening. With this caveat in place, however, it does seem meaningful to relate the divine eye to a deeper discernment of conditionality, since the textual sources explicitly connect the vision of other sentient beings passing away and being reborn to insight into the operative causes behind this process. The account of the Buddha’s own development of this type of knowledge conforms to this pattern by relating ethical misconduct to bad rebirths after passing away and, conversely, moral restraint to rebirth in pleasant circumstances (MN 4 and EĀ 31.3; trans. in Ñāṇamoli 1995/2005, 106 and Anālayo 2016a, 21, respectively). The basic principle of conditionality, evident in this way, corresponds to a key doctrine of early Buddhism that can be considered the other side of the coin of not self: dependent arising. The standard presentation of this principle takes the form of twelve links, which lead from the root cause of ignorance to old age and death. Besides the twelve-link formula, the teaching on dependent arising can take a variety of different forms in shorter formulations, all of which serve as expressions of the same principle. This principle constitutes another significant departure from the ancient Indian philosophical setting. In line with the implications of not self, the doctrine of dependent arising proposes an interplay of a series of causes and conditions without any permanent entity or agent involved. Although a novel conception, the frequent formulation of dependent arising in the form of twelve links appears to be in dialogue with a Vedic myth depicting the creation of the world (Jurewicz 2000). In contrast to this myth of creation, the Buddhist formulation describes the conditioned genesis of the human predicament: old age and death. Moreover, it complements this by pointing to the way out of this predicament, which requires the eradication of ignorance and hence the interruption of the chain of conditions that lead from ignorance to old age and death. In this way, a novel conception appears to have been couched in terminology familiar to the ancient Indian audience, so as to be more easily assimilated.

20

Gotama Buddha

Awakening Based on recollection of his own past lives (and hence probably a dawning insight into the constructed nature of the sense of personal identity) and witnessing the passing away and being reborn of others (and thus witnessing the cause and effect relationship that stands behind the notion of karma), the third knowledge on the night of awakening marks the breakthrough by which Gotama became a Buddha (MN 4 and EĀ 31.3; trans. in Ñāṇamoli 1995/2005, 106 and Anālayo 2016a, 22, respectively). The textual sources consider this breakthrough to correspond to an eradication of all unwholesome influences (āsava/āsrava) in the mind. As a result, the Buddha (and similarly, those of his disciples who subsequently reached the same level of mental liberation by becoming arahants/arhats) was forever free from mental defilements like sensual desire, anger, etc. With all craving and attachments removed, fully awakened ones are considered to have gone beyond the confines of the predicament of saṃsāra. In other words, they have attained the deathless. The way to bring about such awakening (bodhi), sometimes translated as “enlightenment” (Anālayo 2021), involves seven mental qualities known as the “factors of awakening” (bojjhaṅga/bodhyaṅga). These are considered a specific discovery of the Buddha and hence a novel contribution, although they were apparently soon adopted by other practitioners in the ancient Indian setting (Gethin 1992, 177–83). The first of these factors of awakening is mindfulness (sati/smṛti), a quality that in recent times has found widespread employment for secular purposes in the clinical setting, education, and so forth. Its cultivation as a means for progress to liberation takes the form of four establishments of mindfulness (satipaṭṭhāna/ smṛtyupasthāna), which the discourses present as the direct path to the realization of awakening (Anālayo 2003, 2013, 2018b). Building on mindfulness as a mental quality that is required at all times, the remaining six factors of awakening fall into two groups. One set of three serves to energize the mind and is for this reason commendable when some degree of sluggishness manifests: investigationof-states, energy, and joy. The other set of three awakening factors calms the mind and is therefore appropriate when some degree of excitement manifests: tranquility, concentration, and equanimity. The cultivation of these seven awakening factors is a central teaching on meditation in early Buddhist thought.

Teaching and Compassion With awakening attained, the Buddha is on record for having hesitated to teach (MN 26; trans. in Ñāṇamoli 1995/2005, 260). The relevant Pāli discourse reports that this hesitation prompted the Indian creator god Brahmā to intervene and request that the Buddha share his discovery of the path to awakening. Notably, this whole episode is absent from the relevant Chinese Āgama parallel (MĀ 204; trans. in Anālayo 2012b, 32). This version proceeds directly from the Buddha’s awakening to his reflection on whom he should teach first. Nevertheless, the Buddha’s hesitation is a recurrent element in other discourses and later texts (Anālayo 2011, 178–79). The rather substantial difference that emerges here can be interpreted in two ways: either the episode was added to the relevant Pāli discourse (and also influenced other texts), or else it was lost (or even removed intentionally) from the Chinese Āgama version (Anālayo 2012b, 34–37). Whatever may be the final word on these two interpretations, what remains certain is that in the textual material common to parallel versions of the early discourses, the Buddha was not considered to have intentionally prepared himself throughout many past lives for the

21

Bhikkhu Anālayo

task of becoming a compassionate teacher. The idea of having made a corresponding decision upon meeting a previous Buddha by the name of Dīpaṃkara is only attested in later texts (Anālayo 2010, 87). From the viewpoint of the early discourses, the Buddha was simply an exceptionally gifted person who, on realizing the challenge of old age and death, decided to dedicate his life to finding a solution for this predicament. Hence, the question of whether he would subsequently share the solution he had found, by way of teaching others, appears to have been, at least at the outset, not central to his quest. Needless to say, his actual teaching activity remains an expression of his outstanding compassion. The point is only that his overall aim, in the way this is recorded in the early textual sources, had been to liberate himself, rather than to become a compassionate teacher. The bodhisattva ideal, the aspiration to become a Buddha in a future life, is clearly a later development. Its beginnings can be traced through comparative study of the early discourses, which show how relevant elements and notions gradually appear in the course of time (Anālayo 2010, 2017a). Compassion itself, however, appears to have entered the stage only at a comparatively later time. In other words, the original development of the bodhisattva ideal seems to have been stimulated predominantly by the wish to become equal to the Buddha. This holds particularly for wanting to acquire omniscience, which in the meantime had come to be attributed to the Buddha, and for desiring to acquire the marks of his bodily perfection. The compassionate motivation to deliver sentient beings is not evident in what appear to be the first stages in the development of the bodhisattva path. Once such motivation had come to be combined with the bodhisattva ideal, however, it swiftly gained traction. Hence, from the viewpoint of the contribution made by Gotama to thought and practice in the ancient Indian setting, the compassionate motivation to deliver sentient beings by undergoing a long course of gradual preparation over a series of lifetimes should be accredited to later generations.

The Four Noble Truths With or without being prompted by Brahmā, the Buddha is on record for eventually deciding to teach. The relevant discourse reports that he first wanted to share his discovery with his former teachers, Āḷāra Kālāma and Uddaka Rāmaputta, only to find that they had recently passed away (MN 26 and MĀ 204; trans. in Ñāṇamoli 1995/2005, 262 and Anālayo 2012b, 32, respectively). He then decided to teach instead five companions from the time of his ascetic practices. These five had left him upon seeing that he had given up asceticism, believing that he had become lax and was no longer worthy of their support and companionship. In order to convey to them that he had not just reverted to a life of sensuality, the Buddha reportedly began teaching them by proposing a middle path apart from the two extremes of sensual indulgence and ascetic self-mortification (SN 56.11 and EĀ 19.2, as well as other parallels; trans. in Bodhi 2000, 1844 and Anālayo 2016a, 268, respectively; see also Anālayo 2015b, 347–88). The notion of such a middle path is a recurrent motif in early and later Buddhist traditions and can be considered another significant contribution to the ancient Indian setting. The same holds for the actual teaching reportedly given on this occasion, which concerns the four noble truths. This teaching appears to have been modeled on an ancient Indian medical diagnosis (Anālayo 2015b, 25–40), which involves four aspects: the disease’s diagnosis, its etiology, the prognosis, and the cure. Presumably, the Buddha found his realization to be

22

Gotama Buddha

so distinct that he decided to avoid taking up any of the philosophical notions known in the ancient Indian setting. Instead, he preferred to rely on a medical precedent to formulate in a pragmatic manner the key aspects of his realization. In line with its medical precedent, the first noble truth identifies the human predicament. The term employed here is dukkha/duḥkha, whose import is inadequately rendered as “suffering.” Given that dukkha/duḥkha applies to all conditioned phenomena, a translation has to be found that does not result in making the teaching itself become incongruous, such as when pleasant experiences are flatly considered to be “suffering” or else when mountains and rivers are qualified as “suffering.” They obviously are not in themselves suffering, but still they fall within the range of applicability of the term dukkha/duḥkha. Pleasant experiences are pleasant, but they fail to give lasting satisfaction. Mountains and rivers can be perceived in a variety of ways, depending on the perspective taken, but they also are incapable of providing lasting satisfaction. Hence dukkha/duḥkha, when employed in a general sense applicable to all conditioned phenomena, is best rendered as “unsatisfactory.” In contrast, “suffering” is not a quality shared by all conditioned phenomena, but a reaction of the untrained mind. The second noble truth identifies the pathogen responsible for the manifestation of dukkha/ duḥkha, which is craving (taṇhā/tṛṣṇā), in the sense of an incessant thirst for the satisfaction of desires. The same can be stated in more detail by recourse to the whole chain of dependent arising, starting with ignorance. This chain leads via several interim links to the experience of feeling tones (vedanā), which can be pleasant, unpleasant, or neutral. Whereas such feelings tones are a given of experience, the reaction to them by way of craving as the next link in the chain is not. Hence, the presence of mindfulness at the juncture of feeling tones can enable the prevention of reactions impelled by craving. The third noble truth presents freedom from dukkha/duḥkha as the final goal. Such freedom requires the eradication of both craving and ignorance, two complementary perspectives on the root cause of the human predicament. Eradicating ignorance corresponds to the cessation mode of the chain of dependent arising and thereby to the eventual cessation of old age and death. Needless to say, even fully awakened ones, including the Buddha, still grow old and pass away. But they no longer suffer because of that, as their minds are free of craving and ignorance. Nor will they be reborn and experience old age and death in a future life. It is in these two interrelated senses that the cessation of ignorance has brought about for them the cessation of old age and death. The fourth noble truth presents the path that leads to freedom from dukkha/duḥkha, which serves as the conduit to the deathless (and birthless) realization of Nirvana. This path is qualified as being eightfold, as it comprises eight dimensions of practice. The first of these requires establishing the appropriate view (diṭṭhi/dṛṣṭi), in the sense of seeing things in the right perspective, namely, in such a way as to lead towards awakening. This first path factor is based on an initial understanding of the four noble truths as a directive for practice, an understanding that gradually matures until, with the breakthrough to awakening, the four noble truths are fully understood and have become a matter of personal realization. Informed by the directive set by the appropriate view, the next path factor concerns intentions. These should involve the willingness to give precedence to renunciation over sensuality, the absence of anger over irritation, and the presence of harmlessness over cruelty. The intentional disposition established in this way then informs speech, action, and livelihood, which need to be streamlined in such a way that they accord with these basic principles. Here and elsewhere, ethics are a central concern of early Buddhist thought.

23

Bhikkhu Anālayo

The remaining three factors of the eightfold path are properly directed effort, mindfulness, and concentration. These are interrelated aspects of mental culture, reflecting the importance given to cultivation of the mind in the early Buddhist path to deliverance.

Conclusion The chief contributions made by the Buddha Gotama to the ancient Indian setting, in the way these are reflected in the early discourses, appear to involve the following aspects: the notion of the deathless, bereft of connotations of immortality or permanent existence in a higher realm, serves to express freedom from the terror of mortality as a result of reaching awakening. Deep concentration experiences are divested of any metaphysical connotations and seen as merely products of certain conditions, whose cultivation leads to the corresponding experience. The final goal of Nirvana comes to signify a complete transcendence of saṃsāra such that neither continuity of existence nor its annihilation fits the case, a notion that from ancient India until today has not been easily understood by those who have not experienced it. Progress toward this goal involves a stepping out of the pleasure/pain paradigm, in that an evaluation of experience according to its hedonic quality is superseded by the ethical distinction between what is wholesome/skillful and what is unwholesome/unskillful as the proper criterion. Moreover, an assessment along these lines focuses on the mind, in that intention needs to be monitored, first of all, as it in turn influences actual deeds. The teaching on the absence of a permanent self of any type, a radical departure from the ancient setting, has its correlate in the doctrine of dependent arising. Both together explain continuity of life beyond death without positing any form of a permanent entity, relying entirely on the notion of a process of causes and conditions. From the viewpoint of mental cultivation, a key innovation is the teaching of seven mental qualities as factors of awakening, in the sense of their propensity to effect a waking up to the true nature of reality, thereby issuing in liberation of the mind. The first of these is mindfulness. The central teaching of four noble truths that forms the framework for early Buddhist soteriology adopts a medical diagnostic scheme to pinpoint the causes for the human predicament, together with identifying a viable solution and the practical path to be undertaken for that end. In this way, central teachings of early Buddhist thought – which involve distinct contributions to Indian philosophy and soteriology – can be related to the Buddha’s own progress to – and eventual gaining of – awakening.

Abbreviations AN DĀ DN EĀ MĀ MN SĀ SN T Vin

Aṅguttara-nikāya Dīrgha-āgama (T 1) Dīgha-nikāya Ekottarika-āgama (T 125) Madhyama-āgama (T 26) Majjhima-nikāya Saṃyukta-āgama (T 99) Saṃyutta-nikāya Taishō edition Vinaya 24

Gotama Buddha

Notes 1 Although Drewes (2017) is right in pointing out that up to now, Buddhist studies have failed to reconstruct with certainty “the historical Buddha,” he tends to overstate his case. His claim that “it is not clear that the tradition itself envisioned the Buddha as an actual person” is not correct, as is the case for his allegation that the “early texts, such as the suttas of the Pali canon, say hardly anything about the Buddha’s life” (Drewes 2017, 16). The early discourses contain a range of details about the Buddha’s life; otherwise, it would hardly have been possible for Ñāṇamoli (1972/1992) and myself (Anālayo 2017d) to write whole books on this topic based just on the early texts. For critical replies to Drewes (2017), see also Wynne (2019), Levman (2019), and von Hinüber (2019), with corrections to the latter two in Anālayo (2019, 93f) and (2020, 21–25). 2 Laumakis (2013) is right in principle that it is problematic to employ the “received” biography of the Buddha for reconstructing his thought, but in the course of making his case, he tends to overstate the situation – comparable in this respect to the case of Drewes (2017), discussed in the previous note. Although of course “the Buddha never wrote anything,” (Laumakis 2013, 13), we do have orally transmitted records that can be studied by comparing parallel versions. Hence, the situation is more nuanced than the simplistic assumption that “his supposed teachings were compiled anywhere from a hundred to a few hundred years after his death.” It is not quite the case that it remains doubtful whether anything at all “can be said with any degree of certainty with respect to what the man who became the Buddha actually thought or taught.” In fact, a whole range of later developments of Buddhist thought can with certainty be identified as such, making it possible to come to the definite conclusion that these do not belong to what “the man who became the Buddha actually thought or taught.” Although I agree with Drewes (2017) and Laumakis (2013) that it is important to recognize the limitations of the source material at our disposition, it seems to me that excessive skepticism by overstating the situation, however fashionable it may be, risks obfuscating the potential of the material to which we do have access. 3 Bhikkhu Bodhi (2012) departs from the standard way of reference by using the numbering of the Pali Text Society editions. As a result, in the present case, he refers to the relevant discourse as 3.39. 4 The implications of the first of these mental factors, vitakka/vitarka, are debated, making it difficult to translate them without implicitly taking a position. Briefly stated, the question at issue is whether it refers to plain “thought” or not. It does in other contexts, but a perusal of relevant material make it in my view preferable to understand this term to convey the sense of an “application of the mind” (Anālayo 2017c, 123–28).

References Adikaram, E. W. 1946/1994. Early History of Buddhism in Ceylon, or ‘State of Buddhism in Ceylon as Revealed by the Pāli Commentaries of the 5th Century A.D.’ Dehiwala: Buddhist Cultural Centre. Anālayo, Bhikkhu. 2003. Satipaṭṭhāna, the Direct Path to Realization. Birmingham: Windhorse Publications. ———. 2010. The Genesis of the Bodhisattva Ideal. Hamburg: Hamburg University Press. ———. 2011. A Comparative Study of the Majjhima-nikāya. Taipei: Dharma Drum Publishing Corporation. ———. 2012a. “The Historical Value of the Pāli Discourses.” Indo-Iranian Journal 55: 223–53. ———. 2012b. Madhyama-āgama Studies. Taipei: Dharma Drum Publishing Corporation. ———. 2013. Perspectives on Satipaṭṭhāna. Cambridge: Windhorse Publications. ———. 2015a. “Āgama/Nikāya.” In Brill’s Encyclopedia of Buddhism, edited by J. Silk, O. von Hinüber, and V. Eltschinger, vol. 1, 50–59. Leiden: Brill. ———. 2015b. Saṃyukta-āgama Studies. Taipei: Dharma Drum Publishing Corporation. ———. 2016a. Ekottarika-āgama Studies. Taipei: Dharma Drum Publishing Corporation. ———. 2016b. Mindfully Facing Disease and Death, Compassionate Advice from Early Buddhist Texts. Cambridge: Windhorse Publications. ———. 2017a. Buddhapada and the Bodhisattva Path. Bochum: Projektverlag. ———. 2017b. Dīrgha-āgama Studies. Taipei: Dharma Drum Publishing Corporation. ———. 2017c. Early Buddhist Meditation Studies. Barre, MA: Barre Center for Buddhist Studies. ———. 2017d. A Meditator’s Life of the Buddha, Based on the Early Discourses. Cambridge: Windhorse Publications.

25

Bhikkhu Anālayo ———. 2018a. Rebirth in Early Buddhism and Contemporary Research. Somerville, MA: Wisdom Publications. ———. 2018b. Satipaṭṭhāna Meditation: A Practice Guide. Cambridge: Windhorse Publications. ———. 2019. Mindfully Facing Climate Change. Barre, MA: Barre Center for Buddhist Studies. ———. 2020. “The Tevijjavacchagotta-sutta and the Anupada-sutta in Relation to the Emergence of Abhidharma Thought.” Journal of Buddhist Studies 17: 21–33. ———. 2021. “Awakening or Enlightenment? On the Significance of Bodhi.” Mindfulness 12.7: 1653–58. ———. 2022. “Situating Mindfulness, Part 2: Early Buddhist Soteriology.” Mindfulness 13.4: 855–62. Bodhi, Bhikkhu. 2000. The Connected Discourses of the Buddha: A New Translation of the Saṃyutta Nikāya. Somerville, MA: Wisdom Publications. ———. 2012. The Numerical Discourses of the Buddha: A Translation of the Aṅguttara Nikāya. Somerville, MA: Wisdom Publications. Collins, Steven. 1982. Selfless Persons: Imagery and Thought in Theravāda Buddhism. Cambridge: Cambridge University Press. Drewes, David. 2017. “The Idea of the Historical Buddha.” Journal of the International Association of Buddhist Studies 40: 1–25. Gethin, Rupert. 1992. The Buddhist Path to Awakening: A Study of the Bodhi-Pakkhiyā Dhammā. Leiden: E. J. Brill. Jurewicz, J. 2000. “Playing with Fire: The Pratītyasamutpāda from the Perspective of Vedic Thought.” Journal of the Pali Text Society 26: 77–103. Laumakis, Stephen J. 2013. “The Philosophical Context of Gotama’s Thought.” In A Companion to Buddhist Philosophy, edited by Steven M. Emmanuel, 13–25. Chichester: Wiley-Blackwell. Levman, Bryan. 2019. “The Historical Buddha: Response to Drewes.” Canadian Journal of Buddhist Studies 14: 25–56. Liu, Zhen. 2010. Dhyānāni tapaś ca, 禅定与苦修. Shanghai: 古籍出版社. Ñāṇamoli, Bhikkhu. 1972/1992. The Life of the Buddha According to the Pali Canon. Kandy: Buddhist Publication Society. ———. 1995/2005. The Middle Length Discourses of the Buddha: A Translation of the Majjhima Nikāya. Edited by Bhikkhu Bodhi. Somerville, MA: Wisdom Publications. Salomon, Richard. 2018. The Buddhist Literature of Ancient Gandhāra: An Introduction with Translations. Somerville, MA: Wisdom Publications. Vetter, Tilmann. 1995. “Bei Lebzeiten das Todlose Erreichen, Zum Begriff Amata im Alten Buddhismus.” In Im Tod gewinnt der Mensch sein Selbst, Das Phänomen des Todes in asiatischer und abendländischer Religionstradition, edited by Gerhard Oberhammer, 43–74. Wien: Österreichische Akademie der Wissenschaften. von Hinüber, Oskar. 2019. “The Buddha as a Historical Person.” Journal of the International Association of Buddhist Studies 42: 231–64. Walshe, Maurice. 1987. Thus Have I Heard: The Long Discourses of the Buddha. London: Wisdom. Wynne, Alexander. 2019. “Did the Buddha Exist?” Journal of the Oxford Center for Buddhist Studies 16: 98–148.

26

2 SIDDHĀRTHA GAUTAMA Beyond the Historical Figure Richard F. Nance

The Limits of Provenance Each chapter in this volume bears a proper name as its title. The name marks an author who has made a distinctive contribution to Buddhist philosophy in India, and the chapters are meant to provide readers with a sense of who these thinkers were as authors. What were the topics they tended to favor? What was their own sense of the work they were doing, as thinkers, and what should our sense of that work be? In each case, the name of the author serves as a center of gravity, around which orbit topical, methodological, and stylistic concerns. The name of the author indexes forms of consistency across texts, enshrining “relationships of homogeneity, filiation, reciprocal explanation, authentification, or of common utilization” (Foucault 1977 [1969], 123).1 In cases when one finds it difficult to establish such relationships, one tends to think there must be different authors at work. Contemporary scholars are not alone in assigning texts to authors and authors to texts; traditional Buddhists have long undertaken this work, as well. Traditional attributions are sometimes trustworthy and sometimes troubling. When we are troubled by a traditional attribution, we may have the sense that the text before us is not wholly consistent with what we think its putative author would be doctrinally, stylistically, or historically disposed to say.2 Our judgments of consistency are shaped by a complex set of assumptions as to what is possible and probable, and these assumptions do not always match those of Mahāyāna Buddhist traditions. So, for example, whereas Mahāyāna traditions have historically described buddhas as bearing prophetic capabilities and superhuman attributes, we contemporary scholars have tended to dismiss the veracity of such claims. Whereas the traditions have insisted that these buddhas are multiple – that they manifest in countless miraculous ways in the past, present, and future – we have tended to focus on a “founding figure” of the past: the figure of Siddhārtha Gautama (alt. Śākyamuni Buddha). And whereas Mahāyāna traditions appear to attribute a vast and internally diverse corpus of texts to this figure – often speaking of these sūtras as “the word of the Buddha” or “the word of a buddha” (buddhavacana) – we have tended to dismiss such attributions. The grounds for our dismissal have been various: doctrinal inconsistency, stylistic variance, historical anachronism. From a purely philosophical perspective, the texts diverge wildly, with some appearing to endorse ontological realism; others, nominalism or idealism; still others, skepticism. We take such divergences to warrant our judgments DOI: 10.4324/9781351030908-5

27

Richard F. Nance

regarding provenance: the texts strike us as simply too different from one another to constitute the teachings of a single figure. Mahāyāna Buddhist traditions, for their part, have taken a rather different approach to speaking of the/a buddha’s word. Rather than denying the obvious heterogeneity of texts counted as buddhavacana, they have embraced and thematized it, often making reference to nine, twelve, or 84,000 divisions of Buddhist teaching.3 They have also tended to resist the conclusion that textual heterogeneity need imply multiple authorial hands (or mouths) at work – differences between texts could be read as pointing not to a diversity of authors, but rather to a diversity of audiences. After all, great teachers “read the room” and adjust their message accordingly – a capacity that the tradition marks by speaking of a buddha’s unfailing “skillful means” (upāyakauśalya) – and Siddhārtha Gautama was nothing if not a great teacher. Emphasizing the profound malleability of Śākyamuni’s speech in this way, Mahāyāna traditions manage to transform the very ground for contemporary scholarship’s attributional denials into testimony for the Buddha’s pedagogical prowess. Mahāyāna traditions have not, however, spoken with a single voice on this issue. Across the centuries, questions over the provenance of texts traditionally counted as the word of the/a buddha have been pressed not only by scholars of Buddhist studies, but also by Buddhists – and these questions have left traces in the texts themselves. Consider, for example, the following passage, drawn from the Questions of Rāṣṭrapāla (Rāṣṭrapālaparipṛcchā), which likely reflects attributional anxieties. The speaker is purportedly the Buddha Śākyamuni himself, and what he says is rather surprising: When they hear this tranquil dharma, these ignoble ones, who are always hostile to the dharma, who offend against the teaching, and who are devoid of virtues, declare: “This was not spoken by the Victor.” My teacher was an ocean of knowledge, very learned, the best of expounders of the dharma. And yet this sūtra was forbidden by him: “It is by no means the word of the Buddha.” Moreover, he too had an aged teacher, also possessed of an unlimited abundance of virtues, and this sūtra was also not accepted by him: “Do not apply yourself to it; it is false.” They charge:  .  .  .  “invented, imagined by the evil-minded and by those who think like heretical teachers are the Mahāyāna teachings. The Victor could never have said these words, which are a rebuke against monks.”4 Śākyamuni here situates the teaching of “this sūtra” – the very sūtra he is teaching – as having been subjected to repeated dismissal in the past, suggesting that he is somehow at one and the same time both the source and student of the Questions of Rāṣṭrapāla. This framing is, at best, awkward; it is very hard to see how a text that records its own rejection by previous generations of teachers could itself be the very same text rejected by those teachers.5 Problems of reflexivity can be avoided, however, if we read this passage as presenting the incursion of a later editorial voice – one concerned to recount moments in the text’s contested history of transmission. The text may then be read as speaking figuratively of the circumstances of its own journey – a very turbulent journey – through multiple generations of Buddhist teachers who question its status as buddhavacana (Boucher 2008, 72). But if what we are encountering here is a later editorial hand at work, then the conclusion cannot be avoided that Śākyamuni Buddha did not teach the Mahāyāna sūtras – at least some of them, or some portions of them. It is tempting to see in this passage a squabble over authorship between non-Mahāyāna Buddhists and their Mahāyāna brethren. But the sūtra leaves room for the possibility that disagreements over provenance also erupted between Mahāyānists themselves. Daniel Boucher 28

Siddhārtha Gautama

has called attention to an irreducible diversity of viewpoints that characterized the nascent Mahāyāna: what we see in the early Mahāyāna texts, he suggests, is not consensus, but rather a variety of perspectives, indicating that “Mahāyāna sūtras must have been produced by individuals situated in vastly different social experiences” (Boucher 2018, 97). Such differences likely affected attributional claims, as well: it is difficult to imagine that all Mahāyāna practitioners agreed about the provenance of all Mahāyāna sūtras. Perhaps, therefore, it was not only non-Mahāyāna Buddhists, but also Mahāyāna practitioners (sometimes referred to as “bodhisattvas”), who wished to reject attributional claims regarding Mahāyāna texts. The following passage, from Asaṅga’s fourth-century Bodhisattva Stage (Bodhisattvabhūmi), suggests as much: a bodhisattva who, after hearing the profound subjects contained in the bodhisattva collection, fails to develop proper devotion toward those supremely profound subjects that pertain to the nature of reality or to the powers of the buddhas and bodhisattvas and denigrates them by saying, “These teachings do not possess any meaning, they do not possess any genuine dharma, they were not spoken by the Tathāgata, they do not provide benefit and happiness to sentient beings,” commits an offense, a transgression; he or she acquires an afflicted offense (kliṣṭam āpattim), regardless of whether he or she denigrates them because of his or her own improper manner of thinking or because of his or her desire to comply with the views of others.6 Asaṅga seems to be bothered here by those who find themselves tempted to issue attributional challenges to texts he favors. Precisely which texts are at issue is not entirely clear. As Ulrich Pagel has noted (Pagel 1995, 17–36), the term “bodhisattva collection” (bodhisattvapiṭaka) is used in various ways, some wider than others. It is, however, clear that the term is meant to pick out one or more Mahāyāna texts; and it is no less clear that the attributional challenge discussed here is mounted from within the Mahāyāna fold(s), for Asaṅga uses the term bodhisattva to label the person who mounts it. Battles over provenance sometimes pitted Mahāyānists against the mainstream, but they also seem to have at times pitted them against one another. It is wrong, Asaṅga suggests, for a Mahāyāna Buddhist to say that Mahāyāna sūtras were not “spoken by the Tathāgata” (an epithet traditionally used for buddhas); to say this constitutes a transgression against one’s bodhisattva vows. The assumptions of provenance endorsed by the Buddhist tradition thus appear to be at odds with the assumptions of provenance made by scholarship. Clearly, this presents a problematic situation for a would-be scholar-practitioner. When she takes a historical perspective, the conclusion seems to her ineluctable: the Mahāyāna sūtras cannot be tied back to the time of Śākyamuni – and that means, for her, that they cannot be properly classed as buddhavacana, much as Mahāyāna traditions might claim otherwise. That seems a responsible scholarly position to take. But she also is a practitioner, and so she may be brought up short when she finds statements in śāstric texts such as the Bodhisattvabhūmi that unambiguously suggest that to challenge the status of the Mahāyāna sūtras as buddhavacana is to commit an offense. The would-be scholar-practitioner of Mahāyāna traditions today would thus seem to be torn between the demands of her profession and the demands of her religion. While she might reject the traditional attribution of the Mahāyāna sūtras to the Buddha and preserve her academic reputation, she would, in so doing, fail to act as a bodhisattva. Alternatively, she might endorse the traditional attribution, thereby preserving her religious commitments but at considerable cost to her reputation as an academic. The conflict between these demands cannot be ameliorated by attempting to compartmentalize one’s obligations. When one takes a bodhisattva vow, 29

Richard F. Nance

after all, one does not promise to work tirelessly to heal the sorrows of the world for as long as the earth and sky shall last except when academic conferences are in session. Of course, the problem sketched in the preceding paragraph need not trouble a scholar who is not also a Buddhist. But by insisting that Mahāyāna traditions fail a historical test, she risks committing a different sort of offense: the scholarly offense of failing to attend adequately to the complexity of her sources. For Buddhist traditions affirm at least two separable approaches to settling questions of provenance, only one of which looks to the past and sees teachings as inheritances that have come down over centuries. A second account, by contrast, would tie the transmission of Buddhist teachings not back to an awakened figure of the distant past, but rather across to an awakened figure (or figures, or representatives of such figures) of the present. This account manifests in various ways. We are told, for example, that there exist forms of meditative concentration that afford practitioners the possibility of encountering myriad “buddhas of the present” ( pratyutpannabuddha); or we are told that the real nature of buddhas is atemporal and unchanging, and that, for this reason, a buddha can teach at any time via a mutable “emanation body” (nirmāṇakāya).7 It is easy to dismiss these formulations as forms of magical thinking, but they raise real interpretative and philosophical issues. To take such points seriously is to imagine our way into a perspective from which the time of the/a Buddha is not only then, but also now – a perspective that is clearly different from the one that tends to inform contemporary scholarship on Buddhism. This difference affects how one is predisposed to read the whole issue of provenance. For if we grant the possibility of “tying across,” then whether a particular utterance is properly to be counted buddhavacana is not simply a historical matter. To see this more clearly, consider the relation between the following two questions. 1. Was p taught by an awakened being who lived a few centuries prior to the Common Era? 2. Should p properly be classed as buddhavacana? Contemporary scholars tend to collapse these two questions into one. Historians of Buddhist traditions are, after all, champions of “tying back,” and they are likely to view any attempt to distinguish between (1) and (2) as ideologically suspect: end-runs around history are not afforded by appealing to emanation bodies. And while philosophers may find traditional claims of “tying across” to be metaphysically significant and intriguing, they are, in the end, also likely to find those claims to be unwarranted. However we understand the relation between (1) and (2), then – and whether we happen to think of ourselves as historians or philosophers – we are likely to see traditional attempts to assign the status of buddhavacana to Mahāyāna Buddhist sūtras as so much wishful thinking. That might be wrong. In the sections to follow, I will make a case for a different way of reading the relation between (1) and (2). In the reading I will propose, talk of “the word of the/a buddha” need not be unpacked either in terms of “tying back” or in terms of “tying across” (or, for that matter, in some uneasy combination of the two). As we will see, such talk need not concern provenance at all. This is a possibility that, I will suggest, has been hiding in plain sight in Mahāyāna Buddhist texts. If we are to see it clearly, though, we will need not only to reread those texts, but also to learn from them how to read them better. Doing this promises to advance us toward a more historically adequate and – to my mind – more philosophically interesting view of what Indian Mahāyāna Buddhists of the first millennium CE may have been up to in their talk of “the word of the/a buddha.” In order to bring this view into focus, we need first to clarify how

30

Siddhārtha Gautama

we ourselves presently understand what it means to describe a text in this way. This is the task of the next section.

Word: Phrasing and Meaning What do we take buddhavacana to be? The way in which the Sanskrit compound has been translated here – as “the word of the/a Buddha” – is meant to point to the fact that the compound does not completely settle questions of provenance, for the term buddha may or may not index Siddhārtha Gautama. I want to leave that point aside for a moment to explore another ambiguity present in the compound: one that besets the term vacana, here translated as “word.”8 In order not to overwhelm the reader with too many interpretive possibilities at once, I will, for the time being, assume that the buddha in buddhavacana does index – at least – Siddhārtha Gautama, and render the compound as “the word of the Buddha” throughout the remainder of this section. How, then, do we understand the vacana in buddhavacana? What do we mean in speaking of the Buddha’s “word”? Two interpretive paths immediately suggest themselves. On the one hand, “word” could refer to the specific words and phrases ( pada, vyañjana) uttered by Śākyamuni on some occasion of teaching. When the phrase is taken this way, then when we deny that the Mahāyāna sūtras are the word of the Buddha, what we are denying is that the words and phrases presented in the Mahāyāna sūtras as they exist for us today are the precise words and phrases spoken by Śākyamuni. On the other hand, “word” could refer in a more general way to what the Buddha taught: less to specific words and phrases than to the sense (artha) conveyed thereby. When the phrase is taken this way, then when we deny that the Mahāyāna sūtras are the word of the Buddha, what we are denying is that the doctrinal content of the Mahāyāna sūtras corresponds to the doctrinal content of teachings that strike us as more reasonably attributable to Śākyamuni. These two ways of interpreting our denials are clearly different. The first interpretation construes what counts as the word of the Buddha as a collection of words: these words are the very words spoken by the Buddha, verbatim. I will call this a “locked-to-lexemes” view. In a locked-to-lexemes view, the word of the Buddha is strictly unparaphrasable and untranslatable: although buddhavacana might be memorized or copied out, it cannot be summarized or glossed without ceasing thereby to be buddhavacana.9 This is an extreme view, with extreme consequences. It calls into question not only the status of the Mahāyāna sūtras, but also the status of any extant sūtra text, whether it be in Sanskrit or Pāli. For although we do not know for certain which language(s) were adopted by Siddhārtha Gautama for the purpose of teaching, it is generally accepted today that whatever linguistic choices he made are not captured precisely in any version of the teachings that has come down to us.10 Even so, a locked-to-lexemes view may initially appear to square with what we know of the reception history of texts counted as buddhavacana within Buddhist traditions. These texts were clearly memorized (√dhṛ) and collectively recited (saṃ + √gai) within early Buddhist communities. Collective recitation would seem to require agreement on a specific set of words, recited in a specific order. This can tempt us into thinking that these communities championed something resembling a locked-to-lexemes view. But we need to be cautious here, for it is possible to hold both that one must agree on a specific sequence of words to be memorized for the purpose of recitation and that certain paraphrases of what one has memorized might likewise count as buddhavacana. Indeed, there is evidence to suggest a normative traditional understanding of buddhavacana according to which the meaning of what a buddha teaches is to be granted equal, or even greater, importance than its lexical form (Lamotte 1988 [1949]).11

31

Richard F. Nance

This brings us to the second view – one that focuses on meaning, conceived as floating free of specific turns of phrase. I will call this a “mere-meaning” view. When a person who holds a mere-meaning view denies that the Mahāyāna sūtras comprise the word of the Buddha, she is raising a worry not about the words of the Mahāyāna sūtras, but rather about their content, here understood as something that could be paraphrased in more than one way, using more than one succession of words and phrases. The worry is that this content fails to conform to anything that Śākyamuni could possibly have meant. A mere-meaning view may at first seem a more promising route for thinking about issues of sūtra provenance. For one thing, it seems to hew more closely to what contemporary scholars likely have in mind when they – we – issue their attributional challenges. What we have in mind in denying Mahāyāna claims of provenance has to be more a matter of content than phrasing. As already noted, most contemporary scholars are prepared to accept that the languages in which the Buddhist teachings have come down to us do not conform in every respect to the language(s) in which the Buddha is likely to have taught. Even so, many of us would like to be able to say something about what the Buddha did (or did not) teach – and that suggests that we see Buddhist teaching as amounting to (translatable, paraphrasable) messages, imperfectly recoverable though those messages may be.12 A mere-meaning view may also look to be more readily compatible than its rival with claims about the teaching that are made in Buddhist texts. Consider, for example, the fact that the Mahāyāna sūtras routinely advise those who endeavor to understand the teaching of buddhas to “rely on meaning (artha, don), not on phrasing (vyañjana, tshig ’bru).”13 A strong reading of this advice would see it as suggesting that attention to phrasing can be eschewed in favor of attention to meaning – precisely the sort of counsel we would expect from a Buddhist community that had adopted a version of the mere-meaning view. Those seeking support for this strong reading might look to an analogy voiced in the Sūtra which Incites Resolve (Adhyāśayasañcodanasūtra), wherein the teaching is likened to sugarcane: The sugar-cane bark has no juice at all; The taste-giving delight is all found inside; The man who chews only on the bark Cannot find the sugar’s sweet taste. Talking is like the husk of the cane; But thinking on meaning is like the taste. So renounce this joy in talking, And be ever attentive and think on meaning.14 The practical appeal of the mere-meaning view can, however, lead us to ignore the real problems it presents. For one thing, it does not square at all well with the traditional emphasis  – noted previously  – on the memorization and recitation of specific words and phrases. Moreover, a close reading of the sugarcane analogy suggests that we may misunderstand it in taking it to endorse a mere-meaning view – for the “husks” of talk to be discarded are not the Buddha’s own words, but rather the words of those who would content themselves with talking about matters they do not understand. The point of the analogy may thus not be that one need not attend to the Buddha’s words. It may rather be something like the opposite: in lieu of trumpeting one’s own superficially acquired knowledge, one should learn to read more carefully. I would not wish to be read as claiming that no Buddhist communities ever embraced a mere-meaning view  – or, for that matter, a locked-to-lexemes view. Perhaps some did. I do think, however, that the texts of the tradition suggest that neither of these views adequately captures what is likely to have been a more common traditional understanding of 32

Siddhārtha Gautama

buddhavacana – and that we would do better to take our cue from Buddhist traditions on this score. Unfortunately, however, we are not very good at bracketing our own assumptions long enough to get a clear view of the data. When we are asked who is traditionally understood to be the source of texts counted as buddhavacana, for example, the answer may seem patently obvious: “the Buddha,” “a buddha,” or “buddhas.” Yet even the most cursory reading of Buddhist sūtra texts – the paradigmatic texts of buddhavacana – suffices to show that this way of answering the question cannot be correct, for the voices raised in those texts are not always those of buddhas. This is a point that is at once obvious and routinely overlooked: a sūtra text may, and typically does, feature multiple voices: the voice of the/a buddha, to be sure, but also the voice of Ānanda (in the words “thus have I heard” that open many sūtras) and other characters, including kings, brahmins, commoners, monks, nuns, male and female laypersons, competing teachers – even the Buddha’s nemesis, Māra.15 We will come back to this point. For now, it is enough to note that Buddhist traditions appear to be quite comfortable using the term buddhavacana to label texts in which the voices of non-buddhas – and at times, the voice of an anti-buddha – can intermittently be heard.16 This suggests that we may have gone awry in thinking that what is at stake in claims to buddhavacana can be adequately captured in talk of provenance. At the very least, it would seem that, on a traditional view, the word of a buddha extends well beyond the word of a buddha.

The Well Spoken What I have just suggested is that to call something buddhavacana may not be to call it the word of a buddha. That claim, on its face, looks wildly implausible – for a straightforward translation of the compound just is “the word of the/a buddha” (we will now return to the translation with which we began).17 How, then, are we to explain the fact that Buddhist traditions apply the term to texts in which multiple voices can be heard? One explanation would be to count such cases as instances of synecdoche (what in Sanskrit might be termed upalakṣaṇa). Buddhist traditions, on this reading, are using a term that properly pertains to some part(s) of a text – i.e., those parts in which the/a buddha speaks – for the purpose of naming the whole. Reading the usage in this way enables us to hold on to the idea that buddhavacana must pertain to provenance, but at the cost of finding the tradition habitually imprecise – even sloppy – in how it marks texts. That cost strikes me as unreasonably high. Moreover, making this move raises new questions that are not easily answered, for it is not the case that all texts in which the voice of a buddha figures are traditionally counted as buddhavacana. Buddhist philosophical treatises, for example, often contain lengthy passages in which statements of the/a buddha are quoted – but these treatises are not ipso facto “the word of the/a buddha.” Why would Buddhist traditions hold synecdochic usage to be acceptable in certain cases, but not others? There is an alternative explanation: one that would see the practice of labeling sūtra texts as “the word of the/a Buddha” as amounting to a form of idiomatic usage – what might, in Sanskrit, be termed lākṣaṇikaśabda. To my knowledge, the compound buddhavacana is never explicitly marked as an instance of lākṣaṇikaśabda by commentators.18 While this might seem enough to rule out the possibility of idiomatic usage, it must be remembered that we do not always mark our idioms as such. The word “idiom” need not be in the vocabulary of an English speaker who uses “beat around the bush” to mean “fail to get to the point.” What idiomatic usage requires of a speaker is not the knowledge of metalinguistic terms, but knowledge of certain conventions that bear upon the use of a term or phrase. What, then, were the operative conventions governing the use of buddhavacana in traditional contexts? A key to what they may have been could lie in an oft-quoted characterization 33

Richard F. Nance

of Buddhist teaching that echoes across various sūtra texts, according to which the teaching of a buddha is “excellent in meaning and phrasing” (svarthaṃ suvyañjanam, Tib. don bzang po tshig ’bru bzang po).19 What if this characterization captured something like a sufficient condition for counting an utterance or text as the word of the/a buddha? Read in that way, a use of the term buddhavacana would bear less on provenance than on assigning to a text certain qualities traditionally held to characterize the speech of buddhas. That is close to assigning provenance; it requires communal preunderstanding of how a buddha normatively speaks, so worries over provenance cannot drop out completely. But it is importantly distinct, for it suggests that “p is buddhavacana” need not imply “a buddha said p.” Rather, “p is buddhavacana” may imply that p is good in the beginning, middle, and end; that it comprises excellent meaning and phrasing; that it is conducive to the attainment of one’s aims; that it has the capacity to pacify suffering; and so on.20 In short, to call a text buddhavacana may amount to a Buddhist way of saying that the text is well spoken (subhāṣita). There is evidence to suggest that this interpretation is less farfetched than it may first appear. Steven Collins has called attention to the slippage that occurs between an early claim, found on an Aśokan inscription, that “everything spoken by a buddha is well spoken,” and a claim that we find voiced in a text that we have already had occasion to cite: the Sūtra which Incites Resolve, where we read that “whatever is well spoken is spoken by a buddha” (Collins 2013 [1990]).21 These are two very different claims, with very different consequences. According to the first, the term subhāṣita applies to everything that a buddha happens to say: buddhas say only what is well said. Note that this claim leaves room for the possibility that the speech of those who are not (yet) buddhas might also be well spoken. The Sūtra which Incites Resolve, by contrast, appears to shut down that possibility – for it looks to be committed to the claim that only buddhas say what is well said. If that is right, then when we identify any particular utterance as well spoken, we have identified its speaker as a buddha. Before we assume that we have adequately understood what is going on in the Sūtra which Incites Resolve, however, we need to explore the relevant passage in context. To do so is to see a rather different reading emerge. As David Snellgrove has noted, the claim that whatever is well spoken is spoken by a buddha is invoked multiple times in the text. Fortunately, one of these invocations is extant in Sanskrit; it is cited in Śāntideva’s Anthology of Training (Śikṣāsamuccaya) and Prajñākaramati’s Commentary on the Entry into the Practice of Awakening (Bodhicaryāvatārapañjikā). The relevant passage reads: Furthermore, O Maitreya, inspired speech ( pratibhāna) should be understood to be spoken by a buddha, for four reasons. Which four? Here, O Maitreya, inspired speech is meaningful, not meaningless; it accords with the teaching and does not contradict it; it destroys afflictions and does not exacerbate them; it indicates the virtues and advantages of nirvāṇa and does not exacerbate the faults of saṃsāra. All inspired speech endowed with these four qualities should be understood to be spoken by a buddha. When the inspired speech, Maitreya, of any monk or nun, layman or laywoman manifests with these four qualities now or in the future, faithful sons and daughters will there be induced to conceive an idea of the Buddha (buddhasaṃjñā). When an idea of the Teacher has been produced, the dharma is bound to be heard (śāstṛsaṃjñāṃ kṛtvā sa dharmaḥ śrotavyaḥ). Why? Because, O Maitreya, whatever is well spoken is spoken by a buddha. So, Maitreya, whoever would reject these four forms of inspired speech and show no respect for them out of personal contempt ( pudgalavidveṣena), saying “These are not words spoken by a buddha,” would

34

Siddhārtha Gautama

thereby reject all of the inspired speech spoken by buddhas. Rejecting the teaching, he acts in a way that leads to the teaching’s ruin, and so he goes to hell.22 This passage integrates the remark we are working to understand into a discussion of utterances characterized as “inspired speech” ( pratibhāna, Tib. spobs pa).23 Although the remark itself does not refer to inspired speech, the framing unambiguously suggests that the epithet “well-spoken” is to be taken as bearing only on speech that is considered inspired. Despite appearances, therefore, we are not here faced with a universal claim about well-spoken speech. To see this more clearly, consider a parallel case. Imagine that you are in a faculty meeting, and a proposal is floated to which you strongly object. You are worried, however, that your objections will not be preserved accurately in the meeting minutes. A colleague might attempt to reassure you by saying, “Whatever is said is entered into the record.” That claim is perfectly straightforward within the relevant frame. Get rid of the frame, however, and the remark invites puzzlement. How could the record of a faculty meeting contain everything that has ever been said? Obviously, it cannot – and, importantly, no one has suggested that it could. The contextual frame matters, in as much as it conditions what the words one has uttered can be plausibly construed to say. When we attend to the claim from the Sūtra which Incites Resolve within the contextual frame provided by the sūtra, we can see more clearly that despite appearances to the contrary, the text is not claiming that anything well spoken is the speech of a buddha. Instead, it is making two rather different claims, both of which bear on inspired speech. It is claiming, first, that inspired speech is well spoken. And it is claiming, second, that inspired speech is the speech of a buddha.24 The sūtra does not foreclose the possibility that an utterance might be well spoken but not inspired – and thus it leaves room for the very same possibility that one finds in the Aśokan inscription: even non-buddhas might sometimes speak well. That said, there are still two claims made in this sūtra that look to be on a collision course – and the two claims look specifically to concern provenance. One is that inspired speech arises from buddhas; the other is that inspired speech may issue forth from men or women, monastics or non-monastics. How is one to square these claims? One way to do so is by reading the passage as claiming that any of these different sorts of being might attain buddhahood and then go on to teach in inspired ways. That reading reduces tension, but also goes well beyond what the text explicitly says. We may be helped in our efforts here, however, by attending to another passage in the same sūtra that invokes the phrase “whatever is well spoken is spoken by a buddha.” In this second passage, we read that “bodhisattvas lacking in skillful means” (byang chub sems dpa’ thabs la mi mkhas pa) may at times accuse others of “teaching the dharma according to their own inspiration” (rang gi spobs pa chos ston par byed) . . . “in order to please themselves” (rang dgar byas pa), rather than “teaching the doctrine which has its origin in the sūtra and vinaya.” Such accusations, the sūtra suggests, can be dismissed. They are mounted by “fools who do not know that whatever has been well spoken, all is the word of a buddha” (Snellgrove 1958, 623).25 As before, the phrase is invoked in the service of dismissing an objection. The grounds for this dismissal are, however, left underdetermined. On the one hand, the sūtra could be read as accepting the distinction that the objector presumes – accepting, that is, that a bright line can be drawn between teachings that derive from one’s own inspiration and teachings that derive from buddhas – but encouraging the objector (and us) to widen the latter category, allowing prima facie problematic teachings to find a new home on the far side of the line. On the other hand, the sūtra could be read as challenging the objector’s attempt to draw that line, and as encouraging him (and us) to re-think the assumptions that look to warrant the attempt – one of which is attachment to

35

Richard F. Nance

the notion that the word of buddhas ought to be constrained to “that which has its origin in the sūtra and vinaya.” For the sūtra has already suggested the shape of an alternative account: one that would locate buddhavacana in inspired speech, and see its power to lie less in its perfect recapitulation of something already said by a figure of the past than in its ability to induce what the text calls an “idea of the Buddha” (buddhasaṃjñā) or an “idea of the Teacher” (śāstṛsaṃjñā). What are we to make of these two compounds? As Peter Skilling has noted (Skilling 2009, 76), śāstṛsaṃjñā is most plausibly construed as “perception of the Teacher” or “perception as the Teacher” (the term “Teacher” is generally held to refer to Śākyamuni – thus the capitalization). Roughly the same might be said about the compound buddhasaṃjñā: “perception of the Buddha” or “perception as the Buddha.” Beyond these observations, however, we enter speculative territory, for the Mahāyāna sūtra texts that put the compounds to work tend not to unpack them. To go further, we must look beyond these texts to see how the terms are handled in Buddhist philosophical and narrative literature – keeping in mind that whatever picture of the meaning of these terms we end up developing may not conform to how the terms have been understood across the various Buddhist communities through which the Mahāyāna sūtras have historically passed. What is crucial for our purposes is to clarify the term saṃjñā (Pāli saññā, Tib. ’du shes) – a term that can also be rendered into English as “apperception,” “recognition,” “idea,” or “conception.” I will adopt the last of these translations in the discussion to follow. The term saṃjñā, commonly used to index one of the five aggregates invoked in classical Buddhist discussions of experience, is succinctly defined in Vasubandhu’s Treatise on the Five Aggregates (Pañcaskandhaka). It is, Vasubandhu tells us, “the grasping of an object’s sign” (viṣayanimittodgrahaṇam) (Li and Steinkellner 2008, 4).26 Sthiramati’s commentary elaborates: “Object” (viṣaya) means a mental object (ālambana). “Sign” (nimitta) means that quality of the object (viṣayaviśeṣa) that serves as a cause for determining that it is blue or yellow, etc. Thus, to grasp a sign means to make the determination “This is blue” or “This is yellow,” and the like. . . . While all saṃjñās have the same nature, they should be understood to be different owing to their basis and mental object.27 Sthiramati goes on to expand on some of the terms that he has used to unpack Vasubandhu’s definition. He notes that the term “basis” should be understood to refer to the six sense-bases (of the eye, ear, nose, tongue, body, and mind), and that the term “mental object” ranges over four types of mental objects, only two of which need concern us here.28 The first of these is an object that is “associated with signs” (sanimitta), where “sign” picks out “that which is ascribed to an object’s nature on the basis of words” (śabdadvāreṇa yad artharūpam adhyāropitaṃ). The second is an object that is “without signs” (animitta), lacking such an ascription (Kramer 2013, Vol. 1, 30). It is not clear into which of these categories Sthiramati would place an instance of buddhasaṃjñā or śāstṛsaṃjñā, but whether or not the object of a saṃjñā is associated with signs, its arising would seem to put one in a position to identify (what we would call) a certain intentional content as occurrent. On this understanding, a buddhasaṃjñā would then be – at least – a mental episode in which a buddha-shaped content, identifiable as such, is presented. When would such a saṃjñā arise? The most obvious occasion would be when one is in the presence of a buddha. We find just such a circumstance recounted in the voluminous Pāli Legends of the Buddhist Saints (Apadāna), when a saint named Raṃsisaññaka offers the following memory report of an encounter that results in a conception of the Buddha (buddhasaññā): [I saw] the superb Tiger-Bull, Well-Born-One, on a mountainside, 36

Siddhārtha Gautama

like the rising hundred-rayed [sun], like the sun [when its] rays are cool. The majesty of the Buddha was shining on the mountainside. [My] heart pleased in the rays; for an aeon I sported in heaven. In [all] the subsequent aeons, goodness was completed by me because of that mental pleasure, and remembering the Buddha. In the thirty thousand aeons since I obtained that conception then, I’ve come to know no bad rebirth: The fruit of conceiving buddhas.29 The conception of the Buddha noted here clearly possesses significant apotropaic properties. For thirty thousand eons, Raṃsisaññaka tells us, he has experienced no negative rebirth. Protection from harm, then, is one fruit of conceiving buddhas (buddhasaññāyidaṃ phalaṃ) – and protection is a quality repeatedly associated with buddhasaññā, even as other stories from the same collection suggest that the circumstances for its arising may vary. In one of these stories, a saint named Saddasaññaka reminisces over an occasion on which he too came into the presence of the Buddha. Saddasaññaka does not see the Buddha but instead hears his voice – enough, the text suggests, for the arising of a conception of the Buddha that Saddasaññaka later calls up when it comes time for him to die: When the time of my death arrived, I called to mind that conception of the Buddha. In the ninety-four aeons since I obtained that conception then, I’ve come to know no bad rebirth. Both of these stories suggest that conceptions of the buddha can be generated in the presence of buddhas, whether that presence is registered visually or aurally. And both stories suggest that conceptions of the buddha, once generated, can later be remembered, and thus afford access to that which is not immediately present. We might still wonder, however, about the circumstances that must be in place for an apotropaic conception of the buddha to be initially generated. For we might reason that although countless people have brought to mind a conception of the buddha, the practice of doing so does not seem to have unfailingly protected them. Is the immediate presence of a buddha required for that? Perhaps not. A further story from Legends of the Buddhist Saints tells of a master of the three Vedas named Buddhasaññaka, who hears his students report that a buddha has arisen in the world. Wishing to see him, Buddhasaññaka sets off from his hermitage. On the cusp of completing his journey, however, he falls gravely ill: When a league and a half was left To go, illness arose in me. Calling to mind the best Buddha, I passed away right on the spot. 37

Richard F. Nance

In the ninety-four aeons since I obtained that conception then, I’ve come to know no bad rebirth: The fruit of conceiving buddhas (buddhasaññāyidaṃ phalaṃ). Note that Buddhasaññaka has heard from his students about the Buddha; he has not yet had the good fortune to encounter the Buddha directly. Even so, he comes into possession of a conception of the Buddha that works just a well as a conception prompted by direct encounter. An effective buddhasaññā can, it seems, arise on the basis of hearsay; thus have I heard.30 Let us return to the question that prompted this digression on the conception of the Buddha: the question of whether “the word of the/a buddha” (buddhavacana) might be understood as an idiomatic Buddhist way of branding certain utterances as “well-spoken.” As we have seen, this interpretation is on the right track, but needs refining: “the word of the/a buddha” is better construed as an epithet for inspired speech – which is to say, speech that has the power to induce a buddhasaṃjñā.31 The story of Buddhasaññaka helps us to see that speech of this kind need not be issued by a buddha. Paradoxical as it may sound, therefore, the word of the/a buddha need not be a buddha’s word. It may even be the word of an anti-buddha. In a passage at the close of one of the sūtras of the Mahāyāna Heap of Jewels (Ratnakūṭa) corpus – the Sūtra of Instruction on the Inconceivable Sphere of Buddhas (Acintyabuddhaviṣayanirdeśasūtra, Tib. sangs rgyas kyi yul bsam gyis mi khyab pa bstan pa’i mdo) – we find Māra making the following promise (the translations offered in the following are slightly modified from Chang et al. 1983, 35–36): I, Pāpīyān the evil one, have always sought opportunities to oppose buddhas and to create turmoil among sentient beings. Now I vow that, from this day on, I will never go nearer than one hundred leagues away from the place where this doctrine prevails, or where people have faith in, understand, cherish, receive, read, recite, and teach it. Māra immediately goes on to note that his promise cannot, by itself, end all the problems that the tradition labels “deeds of Māra.” Although Māra is now personally inclined to keep his distance, he warns the Buddha that “some of my kindred are determined to distract the devotees’ minds so as to destroy the teaching of buddhas.” That worry is, however, swiftly solved – and by none other than Māra himself, who says: I will chant the following spell so that devotees can vanquish these demons. If good men or good women read, write, and recite this incantation, or teach it to others, the celestial demons will benefit and will, in return, cause the teachers of this teaching to feel joyful in body and mind, to practice vigorously, to possess unimpeded inspired speech and spells, and not to lack services, food and drink, clothing, bedding, or medicine. The mantra is then provided. The Buddha applauds Māra for his efforts and notes that Mañjuśrī should be credited for Māra’s eloquence: “ ‘Wonderful, Wonderful! You should know that all your inspired speech (khyod kyi spobs pa) is a manifestation of Mañjuśrī’s awesome power.’ ”32 We see here something like the same move as we have encountered in the Sūtra which Incites Resolve: Māra’s inspired speech, though marked as his own (it is, the Buddha notes, “yours” (khyod kyi), also manifests a power associated with another: the bodhisattva of wisdom. And while the text does not go so far as to brand Māra’s utterance an instance of buddhavacana, 38

Siddhārtha Gautama

the text from which this passage is derived is unproblematically counted as such by Mahāyāna Buddhists more generally.33

Beyond Provenance: Untying We can now return to a worry that we first confronted at the conclusion of the first section of this chapter: the worry about whether claims regarding provenance that are made by contemporary scholars are actually at odds with the ways in which Buddhists have historically applied the notion of “the word of the/a buddha.” When a contemporary scholar denies that the specific teachings that Mahāyāna Buddhists brand buddhavacana were taught by a figure who lived during the fifth–fourth century BCE, the scholar is clearly making a claim about provenance. When, on the other hand, Asaṅga decries the notion that a particular teaching is not the speech of the Tathāgata (na tathāgatabhāṣitā), what he is likely objecting to is the idea that that text does not comprise buddhavacana. But the latter category, as we have seen, may not carry any implication regarding provenance. To call an utterance (or text) buddhavacana may amount to no more than branding it an instance of inspired speech ( pratibhāna). Inspired speech is not restricted to buddhas – much less to a single figure who purportedly lived during the fifth–fourth century BCE. It may, according to the Sūtra which Incites Resolve, be spoken by “any monk or nun, layman or laywoman.” What Asaṅga deems an offense may thus not be the move familiar to us from contemporary scholarship, but something else; what he is worried over may be less a denial of provenance than a denial of liberating potency. And that sort of denial is not the sort of denial that the contemporary scholar is generally concerned to make. This will likely come as a relief to the would-be scholar-practitioner.34 But I hope that it will also provoke those who are not practitioners to re-think their own assumptions about what Buddhists of the past may have been up to in speaking of buddhavacana. What, then, are the more general takeaway points of this, concerning the figure whose name graces this chapter as its title? If we think of “Siddhārtha Gautama” as naming a historically placed individual who lived during the fifth–fourth centuries BCE, then much of what I have been arguing is likely to seem irrelevant – for we may seem to have learned little as to how to conceptualize the unique bearer of that proper name. To focus on the unique bearer of the name, however, risks missing what the tradition itself suggests to be worth pursuing – and I have tried to take my bearings from tradition on this score, downplaying notions of uniqueness in favor of thinking more broadly about buddhahood, texts, and what is at stake when questions of provenance are raised. Buddhist traditions, and particularly Mahāyāna traditions, are not terribly concerned to demarcate the Buddha as a single individual: the name Siddhārtha Gautama applies not to the, but a, buddha. Buddhas are legion, and the stories spun about them are stereotyped to a degree that suggests their deep similarity, if not interchangeability. “Among buddhas,” the Prophecy for the Magician Bhadra (Bhadramāyākāravyākaraṇasūtra) reminds us, “there is no distinction whatsoever” (D. dkon brtsegs ca 26b6: sangs rgyas rnams la tha dad bya ba’ang med). Recall the two questions that were initially raised in the first section of this chapter. 1. Was p taught by an awakened being who lived a few centuries prior to the Common Era? 2. Should p properly be classed as buddhavacana? Where p = the corpus of Mahāyāna sūtras, a responsible scholarly answer to the first question must be negative, but a responsible scholarly answer to the second question need not be, if one is capable of hearing the question in a certain way. The second question may be asking whether 39

Richard F. Nance

a teaching, regardless of its source, should be counted as inspired speech ( pratibhāna) – i.e., speech contributing to the production of an idea of the buddha (buddhasaṃjñā). Such an “idea of the buddha” could, of course, be an idea bearing on provenance (“this teaching is properly attributable to Siddhārtha Gautama”). But as we have seen, it need not be. Alternative ideas of the buddha could be, e.g., “The being now before me is a buddha”; “I once encountered a buddha”; “I would like to come into the presence of a buddha.” These ideas might arise from the inspired speech of various speakers, none of whom need actually be a buddha. If we understand buddhavacana to be another name for inspired speech, then the way is open for us to hold that it is like emptiness and buddha nature: not restricted to buddhas, and still less to Siddhārtha Gautama alone. This way of interpreting “the word of the/a buddha” is less a form of “tying back” or “tying across” than it is a form of untying – one meant to free us from attachment to the thought that buddhas, whether past or present, can only be elsewhere. In this chapter, I have tried to show that scholars today, whether Buddhist or non-Buddhist, may have much to learn from Buddhist traditions about how to understand what can at first appear to be straightforward attributional claims. I do not wish to suggest that the phrase “the word of the/a buddha” cannot be used to speak of provenance; that is, after all, the most natural and obvious way to read it. But it is not the only way. We can learn from the tradition how to read it otherwise. Moreover, the work of doing so is philosophically valuable, in as much as it encourages us to attend to, and perhaps to reconsider, our own commitments, whether those commitments be scholarly or existential. It is possible, of course, that the vision of buddhavacana proposed here will turn out, on further reflection, to be no more than “an illusion, a mirage . . . a dream . . . an apparition . . . a magical creation.”35 If so, then we can at least take solace in the fact that, in a Buddhist view, that vision will find itself in good company.

Notes 1 The bibliography on concepts of authorship is, of course, immense. For orientation to some central issues, see Burke (1995); Chartier (1992); Griffin (1999); Haynes (2005); Jaffe (2007); Rosenbaum (2006). Studies that take up conceptions of authorship as they bear specifically on Indian Buddhist texts include Cabezón (2001); Cantwell et al. (2013/14) [2015]; Davidson (1990); Gold (2015, Ch. 1); Kapstein (1989, 2018); Kramer (2016). 2 This way of putting the point makes our assessment of a text’s content and aims primary: it is the assessment of content that puts us in a position to make judgments as to the veracity of traditional attributions. But it is also true that what we assume about a putative author may itself influence what we take a text’s content and aims to be. Consider, as a case in point, the Hymn in Praise of Reality (Dharmadhātustava), a work traditionally attributed to Nāgārjuna. The authorship of this text has been the subject of scholarly debate (see Liu 2015, ix), and where we stand on this debate may shape how we read the text. The content and aims of Hymn in Praise of Reality-authored-by-Nāgārjuna are arguably not the same as the content and aims of Hymn in Praise of Reality-not-authored-byNāgārjuna: We are far more tempted to read the former as deploying a series of similes borrowed from buddha nature (tathāgatagarbha) discourse in order to affirm the emptiness of the real, and the latter as invoking ideas associated with the perfection of wisdom literature in order to affirm the eternality of the real. 3 This move is not restricted to the Mahāyāna. For summary remarks, see Heim (2018, 35–38). 4 Translation slightly modified from Boucher (2008, 137). In keeping with the conventions of this volume, Boucher’s use of square brackets has been eliminated, and capitalization has been adjusted. Relevant Sanskrit may be obtained by consulting the Göttingen Register of Electronic Texts in Indian Languages (GRETIL), at http://gretil.sub.uni-goettingen.de/. 5 We might note, in passing, that attempting to answer worries of the sort identified here by appealing to a buddha’s unimpeded knowledge across the three times – i.e., by appealing to prophecy – would come at the cost of introducing new and equally intractable interpretive difficulties. For if I, as a teacher, were to disparage a sūtra only to discover, upon reading it carefully, that it has precisely

40

Siddhārtha Gautama anticipated the terms of my disparagement, I would likely no longer disparage it. I would, however, then need to find a way of explaining how the text fails to anticipate my own change of heart with regards to it. (Perhaps noticing this failure would itself suffice to convince me once again that the text is rightly to be disparaged. Should its prophecy then be counted as fulfilled?) 6 Translation slightly modified from Engle (2016, 291). For the Sanskrit, see Dutt (1966, 119). 7 See Harrison (1990, 2003); Griffiths (1994); Makransky (1997). 8 The reader may believe that the discussion to follow is merely tangential to questions of provenance, in as much as the same string of words can be uttered, or the same content conveyed, by more than one person. This means that getting clear on what is said is not likely to advance one very far toward answering the question of who does the saying. In the case of Buddhist sūtras, however, questions of sourcing cannot be neatly separated from questions of phrasing and meaning, in as much as who does the saying can come into view for us only on the basis of what is said. How we understand “word” in “word of the Buddha” thus matters to settling issues of provenance, in as much as it shapes our judgments of divergence. If we are prone to construe “word” in terms of lexical form, we are liable to see differences between texts that we might fail to see if we construe “word” in terms of semantic content – and vice-versa (for just as divergence in form need not imply divergence in content, divergence in content need not imply divergence in form). 9 In a famous passage from the vinaya literature, the Buddha instructs his interlocutors that his words may be rendered “with one’s own [alt. my own] expression” (sakāya niruttiyā). How this phrase is properly to be understood has, however, been the subject of considerable scholarly debate in the wake of Brough (1980). For a useful summary, see Levman (2008–2009). 10 For discussion of the debate, see Levman (2016). For more recent salvos, see Gombrich (2018); Karpik (2019). 11 Pace Levman, who writes (Levman 2008–2009, 40): It has long been appreciated that Buddhist monks were expected to memorize and recite buddhavacana as an integral part of their practice . . . [y]et the fact that this contradicts the conventional understanding of Cullavagga v. 3 – that each monk was to learn, and presumably recite the dharma in his own dialect – has not been noted. For if everyone were reciting something different, then there is no root text to memorize. The final claim here is decidedly peculiar. Consider a situation in which different Buddhist communities memorize and collectively recite slightly different root texts. Within one such community, there may be collective consensus on what the words of the root text are; within another such community, there might equally be collective consensus on what the words of the root text are. Under such circumstances, the claim that there is “no root text to memorize” would strike the members of any of these communities as peculiar – for there is a root text in and for that community. When members of the community have memorized this root text, the practice of collective recitation can proceed smoothly; the fact of inter-communal divergence is of no consequence. Things might, of course, strike us – those of us in a position to note inter-communal divergence – rather differently. We might be troubled by inter-communal divergence in a way that the Buddhist communities we study were not. It should be noted, though, that Indian Buddhists were not blind to such divergence; it is, for example, explicitly acknowledged by Vasubandhu in the opening pages of the Logic of Explication (Vyākhyāyukti), when he notes that reciting “good in meaning, good in phrasing” (svarthaṃ suvyañjanam) is to be preferred over reciting “possessed of meaning, possessed of phrasing” (sārthaṃ savyañjanam) in a well-known pericope pertaining to the teaching of buddhas. (We will come back to this pericope later; for a fuller translation of the context for this passage, see Nance [2012, 131]). Even so, it seems a stretch to think of the situation sketched here as one in which there is no root text to memorize. If anything, there are multiple root texts in play. 12 Representative here is Gombrich (2009, 3): a successful interpretation of the Buddha will make clear not only the ideas he expressed but also how those ideas lent themselves to the various interpretations which are in fact historically attested. The Buddha will thus stand as the source for a successful history of Buddha ideas – even though to compose such a history, even in outline, may be beyond the powers of any single scholar. 13 This advice – one of the so-called “four reliances” (catuḥpratisaraṇa) – is attributed to a buddha in numerous Mahāyāna sūtras, and it is repeated in śāstric texts that span centuries, from the Śrutamayībhūmi section of the Yogācārabhūmi (c. fourth century) to Ratnākāraśānti’s Ratnāloka (eleventh century).

41

Richard F. Nance 14 Translation slightly modified from Sherburne (2000, 199). This passage is cited in the fifth chapter of Atiśa’s Commentary on the Lamp for the Bodhisattva Path (Bodhimārgapradīpapañjikā) (Sherburne 2000, 196). The relevant Tibetan (corrected from Sherburne 2000, 196, 198) here reads: bu ram shing shun snying po ci yang med / dga’ bar bya ba’i ro ni nang na ’dug / shun pa zos pa mis ni bu ram ro / zhim po rnyed par nus pa ma yin no / shing gi shun pa de bzhin smra ba ste / ro lta bu ni ’di ni don sems yin / de ltas bas na smra la dga’ spangs te / rtag tu bag yod byos la don sems shig. Cf. also the *Mahāyānapathasādhanavarṇasaṃgraha (D, dbu ma khi, 299a5–302b6), which offers a very similar verse. 15 This point regarding the presence of multiple voices could, of course, be expanded by attending to voices that surface in various ways across the text’s history of transmission, i.e., the voices of interpolators, compilers, redactors, scribes, orators, teachers, etc., responsible for working with texts over time. 16 This multiplicity of voices does not seem to be a late introduction to the scriptural corpus. That is, we have no reason to suppose that the use of the term buddhavacana is an atavistic holdover from a time when Buddhist sūtra texts were taken to comprise nothing more than soliloquies of the Buddha (or, for that matter, Ānanda). As far as we are able to tell, there was no such time. 17 While explicit analyses – vigrahas – of the compound buddhavacana that survive in Indic languages are hard to come by, the phrase buddhasya vacanaṃ is attested more than once – and the Pāli canon is riddled with occurrences of the comparable buddhassa sāsanam (“the teaching of the/a buddha”). The relevant Sanskrit shows up in the Legend of Aśoka (Aśokāvadāna) and elsewhere, and variants of the Pāli phrase buddhassa sāsanam show up nearly two hundred times across the canonical and postcanonical materials assembled in the Chaṭṭha Saṅgāyana collection, searchable at https://tipitaka.org/ search. The compound veritably begs to be construed as a straightforward subjective genitive (ṣaṣṭhī tatpuruṣa) construction: to call something buddhavacana is simply to say that it is a buddha’s speech. 18 A global account of language as constitutively metaphoric (Tzohar 2018) would arguably have little trouble accommodating this idea. That said, the interpretation of buddhavacana I am forwarding here does not require assent to such a theory. 19 The Pāli canonical variant is sātthaṃ sabyañjanam. Literally, this means “possessed of meaning and phrasing,” but commentators in the Pāli tradition take pains to note that what is meant by this is that both the meaning and the phrasing of a buddha’s teaching are possessed of magnificence or excellence (sampatti) – and it is this idea that is captured more literally in the Sanskrit version, which is widely attested across Mahāyāna and non-Mahāyāna texts. 20 These characterizations are paraphrased from an oft-cited pericope describing the teaching of buddhas, discussed at the opening of Vasubandhu’s Vyākhyāyukti. For a translation, see Nance (2012, 130). 21 The first claim occurs in the Bhabrā inscription of Aśoka (Collins 2013 [1990], 7): e keci bhaṃte bhagavatā budhena bhaṣita save se subhāsite vā. The second claim will be discussed at length in what follows; a parallel in Pāli occurs at AN 4:164. 22 Translation modified from Snellgrove (1958). Snellgrove offers a Sanskrit text that corrects Bendall’s edition of the Śikṣāsamuccaya on the basis of the Narthang edition of the Tibetan. As already noted, much of the relevant Sanskrit shows up also in Prajñākaramati’s commentary to verse 9.42 of the Entry into the Practice of Awakening (not consulted by Snellgrove), but the Sanskrit that survives in Prajñākaramati’s text does not entirely match that which survives in Śāntideva’s text – and both diverge in certain respects from the Tibetan translation of the sūtra (which nevertheless presupposes an underlying text closer to Prajñākaramati than to Śāntideva). 23 See MacQueen (1981, 1982); Braarvig (1985); Nance (2012, Chapter 2). 24 Again, a parallel case might be instructive. I tell you that all ravens are black; I tell you that all ravens are birds. I have not licensed you to conclude that I hold everything black to be a bird. 25 Cf. a parallel Pāli passage at AN 4.164, where the remark that whatever is well-spoken is spoken by a buddha ( yaṃ kiñci subhāsitaṃ sabbaṃ taṃ tassa bhagavato vacanaṃ arahato sammāsambuddhassa) is made in response to much the same worry – i.e., a worry over whether a teaching that has not been heard directly from the mouth of the Buddha should be thought to “derive from [the speaker’s] own inspiration” (sakapaṭibhānaṃ upādāya) or whether it instead should be thought to present “the word of the Blessed One, the Arahant, the Perfectly Awakened One” (bhagavato vacanaṃ arahato sammāsambuddhassā). In the Pāli text, this worry is answered with a simile: we are asked to imagine the teaching of the Blessed One as akin to a great heap of grain from which others can draw sustenance. We are further asked to imagine a situation in which some persons who have retrieved a portion of this grain are stopped and asked about its source: “Where did you get this grain?” The correct answer to this question is, the sutta tells us, “We got it from that great heap of grain.” It is in

42

Siddhārtha Gautama precisely the same way, the sutta goes on to note, that we should understand all that is well-spoken to be the word of the Buddha, and that “I myself and others derive our good words from him” (tato upādāyupādāya mayaṃ caññe ca bhaṇāmā ti). One might hear in the analogy an echo of the Vedic likening of human knowledge to handfuls of dust (Taittirīya Brāhmaṇa 3.10.11.3–5). Precisely how we are to trace out its implications is not clear; what is clear, though, is that the analogy does not suggest that one who speaks well must for that reason alone be understood to be a buddha. 26 For more on the shifting conception of saṃjñā in Buddhist thought, see Kumagai (2019). 27 Translation slightly modified from Engle (2009, 271). For relevant Sanskrit, see Kramer (2013, Vol. 1, 29). As Engle notes, the opening portion of the passage cited here is quite similar to a discussion of saṃjñā found in the Triṃśikabhāṣya (Buescher 2007, 56). 28 The remaining two are “those that are immeasurable” (tshad med pa, apramāṇa) and “those that are associated with the Sphere of Nothingness” (ci yang med pa’i skye mched, ākiṃcanyāyatanam). For our purposes here, these may be set aside, as they occur only in advanced meditative states. 29 All translations from Legends of the Buddhist Saints have been slightly modified from Walters (2018). 30 The well-known tale of an encounter between Upagupta and Māra that is related in the Legend of Aśoka (Aśokāvadāna) is also relevant here. In this story, which has been translated and discussed at some length by John Strong, Māra appears in the guise of a buddha (Strong 1992, 105–9, 192–96). At the sight of Māra in this form, Upagupta’s mind is overwhelmed by a buddhasaṃjñā, and he falls prostrate, “like a tree cut off at the root.” Upagupta’s buddhasaṃjñā is clearly prompted by an encounter not with a buddha, but rather with Māra. His falling prostrate thus looks to be something of a gaffe, requiring him to explain that his reverential act should be understood as directed not toward Māra himself, but to the being whose form Māra had so skillfully assumed. If Buddhasaññaka’s conception of the buddha is prompted by hearsay, Upagupta’s is prompted by something like “seesay” – that is, he sees the form of a buddha, though no buddha is in fact before him. And whereas the outcome of Buddhasaññaka’s conceiving of a buddha is unambiguously positive, the outcome (or, at least, one outcome) of Upagupta’s conceiving of a buddha is a moment of awkwardness. Despite these differences, however, both stories suggest that a conception of a buddha can be generated even when no encounter with a buddha has occurred. Moreover, neither story occurs in a Mahāyāna source, suggesting that the view of buddhasaṃjñā the stories imply may not be a Mahāyāna innovation. 31 An etic way of putting this point, using terms borrowed from Austin (and quite alien to Buddhist tradition), is to say that buddhavacana can be understood to label a species of perlocution. See Austin (1975 [1962], 101–4). 32 The relevant Tibetan may be found at D. dkon brtsegs ca 275b3–276a5. 33 The text’s Tibetan translation is catalogued among the texts of the Bka’ ’gyur (“Translation of the word”) – a collection traditionally held to comprise the discourses of buddhas. Moreover, the most well-known English translation of texts drawn from the Heap of Jewels corpus (Chang et al. 1983) bears the title A Treasury of Mahāyāna Sūtras (and, as previously noted, sūtras are paradigm cases of buddhavacana). 34 Alas, the scholar-practitioner is not out of the woods yet. She will need to think carefully about the category of inspired speech – and more specifically, what that category excludes. The standing threat here is that once we begin to think of buddhavacana as speech that prompts buddhasamjñā, the latter category will expand willy-nilly, and take the former category with it. Consider, for example, the claim that “all buddhas are manifestations of Māra.” By traditional lights, that is a mistaken claim. But it is also a claim about buddhas, and so could be understood as prompting an idea of the Buddha. Why it should not be so understood is something that would take some showing. But the example suggests that if we go wrong in seeing traditional invocations of buddhavacana as uniformly gesturing toward provenance, we go wrong also in seeing them as uniformly gesturing toward a particular content. The point here is not to shift the way we translate the compound buddhavacana from “speech of a Buddha” to “speech about a Buddha.” What is needed for a buddhasaṃjñā to arise clearly must be more than mere mention of Buddhas. 35 The reference here is to a famous pericope from the Diamond Cutting Transcendent Wisdom (Vajracchedikā Prajñāpāramitā) describing the status of all conditioned phenomena.

Bibliography Austin, John Langshaw. 1975 [1962]. How to Do Things with Words: The William James Lectures Delivered at Harvard University in 1955, edited by J.O. Urmson and Marina Sbisà. Oxford: Oxford University Press.

43

Richard F. Nance Beyer, Stephan. 1978. “Notes on the Vision Quest in Early Mahāyāna.” In Prajñāpāramitā and Related Systems: Studies in Honor of Edward Conze, edited by Lewis Lancaster and Luis O. Gomez, 329–40. Berkeley, CA: Berkeley Buddhist Studies Series. Boucher, Daniel. 2008. Bodhisattvas of the Forest and the Formation of the Mahāyāna. Honolulu: University of Hawai’i Press. ———. 2018. “Recruitment and Retention in Early Bodhisattva Sodalities.” In Setting Out on the Great Way: Essays on Early Mahāyāna Buddhism, edited by Paul Harrison, 95–118. Sheffield: Equinox Publications. Braarvig, Jens. 1985. “Dhāraṇī and Pratibhāna: Memory and Eloquence of the Bodhisattvas.” Journal of the International Association of Buddhist Studies 8 (1): 17–29. Brough, John. 1980. “Sakāya Niruttiyā: Cauld kale het.” In The Language of the Earliest Buddhist Tradition, edited by Heinz Bechert, 35–42. Göttingen: Vandenhoeck & Ruprecht. Buescher, Hartmut. 2007. Sthiramati’s Triṃśikāvijñaptibhāṣya: critical editions of the Sanskrit text and its Tibetan translation. Vienna: Austrian Academy of Sciences Press. Burke, Seán. 1995. Authorship: From Plato to Postmodernism. Edinburgh: Edinburgh University Press. Cabezón, José. 2001. “Authorship and Literary Production in Classical Buddhist Tibet.” In Changing Minds: Contributions to the Study of Buddhism and Tibet in Honor of Jeffrey Hopkins, edited by Guy Newland, 233–63. Ithaca, NY: Snow Lion Publications. Cantwell, Cathy, Jowita Kramer, Robert Mayer, and Stefano Zachetti, eds. 2013/14 [2015]. “Authors and Editors in the Literary Traditions of Asian Buddhism.” Journal of the International Association of Buddhist Studies 36/37: 195–561. Chang, Garma et al., trans. 1983. A Treasury of Mahāyāna Sūtras: Selections from the Mahāratnakūṭa Sūtra. University Park, PA: Pennsylvania State University Press. Chartier, Roger. 1992. The Order of Books. Stanford, CA: Stanford University Press. Collins, Steven. 2013 [1990]. “On the Very Idea of the Pali Canon.” In Self and Society: Essays in Pali Literature and Social Theory, 1988–2010, 3–30. Chiang Mai: Silkworm Books. Davidson, Ronald. 1990. “An Introduction to the Standards of Scriptural Authenticity in Indian Buddhism.” In Chinese Buddhist Apocrypha, edited by Robert E. Buswell, 291–325. Honolulu: University of Hawai’i Press. Dutt, Nalinaksha, ed. 1966. Bodhisattvabhumi. Tibetan Sanskrit Works Series, vol. 7. Patna: K. P. Jayaswal Research Institute. Engle, Artemus, trans. 2009. The Inner Science of Buddhist Practice: Vasubandhu’s Summary of the Five Heaps with Commentary by Sthiramati. Ithaca, NY: Snow Lion Publications. ———. 2016. The Bodhisattva Path to Unsurpassed Enlightenment: A Complete Translation of the Bodhisattvabhūmi. Ithaca, NY: Snow Lion Publications. Foucault, Michel. 1977. “What Is an Author?” In Language, Counter-Memory, Practice: Selected Essays and Interviews, 113–38. Ithaca, NY: Cornell University Press. Gold, Jonathan. 2015. Paving the Great Way: Vasubandhu’s Unifying Buddhist Philosophy. New York: Columbia University Press. Gombrich, Richard. 2009. What the Buddha Thought. London, UK: Equinox. ———. 2018. Buddhism and Pāli. Oxford: Mud Pie Books. Griffin, Robert. 1999. “Anonymity and Authorship.” New Literary History 30 (4): 877–95. Griffiths, Paul J. 1994. On being Buddha: The Classical Doctrine of Buddhahood. Albany, NY: State University of New York Press. Gyatso, Janet. 2017. Being Human in a Buddhist World: An Intellectual History of Medicine in Early Modern Tibet. New York: Columbia University Press. Harrison, Paul. 1990. The Samādhi of Direct Encounter with the Buddhas of the Present. Studia Philologica Buddhica 5. Tokyo: International Institute for Buddhist Studies. ———. 2003. “Mediums and Messages: Reflections on the Production of Mahāyāna Sūtras.” Eastern Buddhist 35 (1–2): 115–51. Haynes, Christine. 2005. “Reassessing ‘Genius’ in Studies of Authorship: The State of the Discipline.” Book History 8: 287–320. Heim, Maria. 2018. Voice of the Buddha: Buddhaghosa on the Immeasurable Words. Oxford: Oxford University Press. Jaffee, Martin S. 2007. “Rabbinic Authorship as a Collective Enterprise.” In The Cambridge Companion to the Talmud and Rabbinic Literature, edited by Charlotte E. Fonrobert and Martin S. Jaffee, 17–37. Cambridge, UK: Cambridge University Press.

44

Siddhārtha Gautama Kapstein, Matthew. 1989. “The Purificatory Gem and Its Cleansing: A Late Tibetan Polemical Discussion of Apocryphal Texts.” History of Religions 28 (3): 217–44. ———. 2018. “Who Wrote the Trisvabhāvanirdeśa? Reflections on an Enigmatic Text and Its Place in the History of Buddhist Philosophy.” Journal of Indian Philosophy 46 (1): 1–30. Karpik, Stefan. 2019. “The Buddha Taught in Pali: A Working Hypothesis.” Journal of the Oxford Centre for Buddhist Studies 16: 10–86. Kramer, Jowita. 2012. “Descriptions of ‘Feeling’ (vedanā), ‘Ideation’ (saṃjñā), and ‘the Unconditioned’ (asaṃskṛta) in Vasubandhu’s Pañcaskandhaka and Sthiramati’s Pañcaskandhakavibhāṣā.” Rocznik Orientalistyczny 65 (1): 120–39. ———. 2013. Sthiramati’s Pañcaskandhakavibhāṣā Part I: Critical Edition; Part II: Diplomatic edition. Vienna: Austrian Academy of Sciences Press. ———. 2016. “Some Remarks on Sthiramati and His Putative Authorship of the Madhyāntavibhāgaṭīkā, the *Sūtrālamkāravṛttibhāṣya and the Trimśikāvijnaptibhāṣya.” Buddhist Studies Review 33 (1–2): 47–63. Kumagai, S. 2019. “Classifications of Perception (saṃjñā) in Buddhist Āgama and Ābhidharma Treatises.” In Reasons and Lives in Buddhist Traditions: Studies in Honor of Matthew Kapstein, edited by Dan Arnold, Cécile Ducher, and Pierre-Julien Harter, 257–64. Somerville, MA: Wisdom Publications. Lamotte, Étienne. 1988 [1949]. “Assessment of Textual Interpretation in Buddhism.” Translated by Sara Boin-Webb. In Buddhist Hermeneutics, edited by Donald S. Lopez, 11–27. Honolulu: University of Hawai’i Press. Levman, Bryan. 2008–2009. “Sakāya Niruttiyā revisited.” Bulletin d’Études Indiennes 26–27: 33–51. ———. 2016. “The Language of Early Buddhism.” Journal of South Asian Languages and Linguistics 3 (1): 1–41. Li, Xuezhu, and Ernst Steinkellner, eds. 2008. Vasubandhu’s Pañcaskandhaka. Vienna: China Tibetology Publishing House, Austrian Academy of Sciences Press. Liu, Zhen. 2015. The Dharmadhātustava: A Critical Edition of the Sanskrit with the Tibetan and Chinese Translations, a Diplomatic Transliteration of the Manuscript and Notes. Vienna: Austrian Academy of Sciences. Lopez, Donald S. 1988. “On the Interpretation of the Mahāyāna Sūtras.” In Buddhist Hermeneutics, edited by Donald S. Lopez, 47–70. Honolulu: University of Hawai’i Press. MacQueen, Graeme. 1981. “Inspired Speech in Early Mahāyāna I.” Religion 11: 303–19. ———. 1982. “Inspired Speech in Early Mahāyāna II.” Religion 12: 49–65. Makransky, John. 1997. Buddhahood Embodied: Sources of Controversy in India and Tibet. Albany, NY: State University of New York Press. Nance, Richard. 2012. Speaking for Buddhas: Scriptural Commentary in Indian Buddhism. New York: Columbia University Press. ———. 2013. “The Voice of Another: Speech, Responsiveness, and Buddhist Philosophy.” In A Companion to Buddhist Philosophy, edited by S. Emmanuel, 366–76. Hoboken, NJ: Wiley-Blackwell. Pagel, Ulrich. 1995. The Bodhisattvapiṭaka (Buddhica Britannica, Series continua V). Tring: The Institute of Buddhist Studies. Rosenbaum, Ron. 2006. The Shakespeare Wars: Clashing Scholars, Public Fiascoes, Palace Coups. New York: Random House. Sherburne, Richard. 2000. The Complete Works of Atīśa Śrī Dīpaṃkara Jñāna, Jo-bo-rje. New Delhi: Aditya Prakashan. Skilling, Peter. 2009. “Seeing the Preacher as the Teacher: A Note on Śāstṛsaṃjñā.” Annual Report of the International Research Institute for Advanced Buddhology at Soka University XII: 73–100. Snellgrove, David. 1958. “Note on the Adhyāśayasaṃcodanasūtra.” Bulletin of the School of Oriental and African Studies, University of London 21 (1): 620–23. Strong, John. 1992. The Legend and Cult of Upagupta: Sanskrit Buddhism in North India and Southeast Asia. Princeton, NJ: Princeton University Press. Tatz, Mark. 1994. The Skill in Means (Upāyakauśalya) Sūtra. New Delhi: Motilal Banarsidass. Tzohar, Roy. 2018. A Yogācāra Buddhist Theory of Metaphor. Oxford: Oxford University Press. Vaidya, Parashuram Lakshman. 1961. Mahāyāna-Sūtra-Saṃgraha. Darbhanga: The Mithila Institute of Post-Graduate Studies in Sanskrit Learning. Walters, Jonathan, trans. 2018. Legends of the Buddhist Saints (Apadānapāli). http://apadanatranslation.org/.

45

3 THE TANTRIC BUDDHA Primordial Buddhas as Philosophical Authors Vesna A. Wallace

A Buddha as an Author in Buddhist Tantric Sources In Indian Buddhist tantras, the buddha, as a teacher of Buddhist tantric tenets, speaks on a variety of topics to a variety of audiences in various realms for different purposes, establishing himself as the ultimate reality appearing in multifaceted forms to lead others to the highest bliss of awakening. Thus, establishing himself as a compassionate and omniscient authority, he teaches tantras in his various manifestations of enjoyment bodies (saṃbhogakāya) and sometimes even in his ultimate Vajradhara form. According to the Stainless Light Commentary (Vimalaprabhāṭīkā) on the Wheel of Time Tantra (Kālacakratantra), the buddha teaches by means of limitless terms, and the redactors wrote down such terms employing the terminology of scholastic treatises (śāstras), vernacular terms, mantric terms, and single-syllabic terms. A tantric practitioner is expected to understand all of these terms in both their provisional (neyārtha) and definitive meanings (nītārtha) (Stainless Light Commentary, vol. 1, 1986, 31–34, 47). While in some tantras, the buddha is primarily concerned with teaching tantric practices, both mundane and those deemed indispensable for awakening; in other tantras, he equally delves into the doctrinal and practical aspects of tantric systems. In the higher classes of tantras, such as the yoga and unexcelled yoga tantras (niruttarayoga-tantra), his authoritative teachings seem to be most focused on those aspects of tantric discourses that are systematically pointing out the nondual, blissful, and empty nature of ultimate reality present in all sentient beings and attainable within a single lifetime only by means of a non-conceptual mind of bliss. Although many of his expositions on emptiness and on the innate, luminous nature of sentient beings’ minds correspond to those expounded in some Mahāyāna sūtras, it is this persistent focus on bliss and preservation of the body in tantric teachings that differentiates the tantric buddha from the buddha as an authoritative figure in the sūtras. Regarded as omniscient, he stands as the highest authority on all systems of thought, Buddhist and non-Buddhist, and as a single, all-pervading mind of the world, he has the entire world as his body. Thus, there is nothing that stands outside of him, including the deities worshipped by various non-Buddhist groups. Everything that can be experienced by the senses and is cognizable by the ordinary mind is inseparable from his provisional forms that manifest to minds still veiled by mental obscurations. Thus, he is established in Buddhist tantras as a sole authority. As the ultimate, perfectly enlightened author of tantras, he invariably manifests as male, although, as we are told in the Wheel of Time Tantra, by nature he is neither male nor female. 46

DOI: 10.4324/9781351030908-6

The Tantric Buddha

To understand a buddha as an author of a tantra, it may be useful to turn to a common definition of tantra given in Buddhist tantric literature as a “connected discourse” ( prabandha) that consists of the cause, result, and method of achieving buddhahood. Fundamental to these three aspects of a tantra is a buddha’s blissful gnosis, present in all sentient beings and expressing itself by means of the provisional and definitive statements under the names of Kālacakra, Hevajra, Heruka, and so on. Thus, in Indian Buddhist tantras, the buddha appears as a teacher in diverse peaceful and fierce forms including Hevajra, Cakrasaṃvara, and Kālacakra, and is referred to by a variety of appellations including the Omniscient One (sarvajña, sarvavid), the Exalted One (bhagavān), Vajrī, Vajrasattva, and Vajradhara.1 He is presented as superior to all Hindu gods such as Brahmā, Viṣṇu, and Śiva, all of whom take refuge in him and worship at his feet. In the Hevajratantra, he is called Bhagavān because he is endowed with six excellent qualities in their entirety, which are mentioned in the Supreme Primordial Buddha Tantra (Paramādibuddhatantra) as supremacy, completeness, gnosis, splendor, glory, and beauty.2 In the Hevajratantra and in the Abridged Wheel of Time Tantra (Laghukālacakratantra), he is given this name because he crushes the inner negative forces, the so-called māras embedded in an ordinary human mind-body complex. He is also referred to as “ultimate reality” (tattva) because he is true bliss. In the Hevajratantra, we read that [T]he Buddha has neither existence nor is he of the nature of nonexistence as a Dharma body; he has a form with face, arms, and legs in his enjoyment and emanation bodies, and yet he is without form due to his supreme bliss ( paramasaukhya). (Tripathi and Negi 2001, 147) While in many tantras, the buddha appears as a single teacher, in others, such as the Secret Assembly Tantra (Guhyasamājatantra), multiple buddhas make their individual declaration with regard to specific doctrinal tenets, and yet at other times, an instruction is collectively given by the entire assembly of the buddhas. In tantras in which the opening line, “Thus I have heard” (which marks texts within the tradition as the authentic words of the buddha), is absent, the Bhagavān as the teacher (śāstṛ) engages in a dialogue with either bodhisattvas or other buddhas, making references to his own various enlightened manifestations in the third person. He does so as if saying, “do not identify me with these forms”; and yet, these manifestations are often spoken of in tantras as the heart, essence, or secret of the body, speech, and mind of all buddhas. In Buddhist tantras, the buddha as ultimate reality ( paramārtha, tattva) is also known as a gnosis (  jñāna) of the sublime bliss of all the buddhas, as the progenitor of the best of victors (  jina). Likewise, as stated in the Guhyasamājatantra, he is known as “the mind of awakening (bodhicitta), which is without beginning and end, peaceful, beyond existence and nonexistence, and omnipresent due to the equality (samatā) of phenomena and as nonduality of emptiness and compassion” (Guhyasamājatantra; Matsunaga 1978, ch. 2, 10). As taught by the buddha Hevajra in the Hevajratantra: He alone is a life ( prāṇa) of living beings, and he alone is the supreme indestructible ( paramākṣara). He pervades everything and dwells in the body of all beings. (Tripathi and Negi 2001, 114)

The Buddha as Philosopher and Doxographer in the Wheel of Time Tantra According to the Wheel of Time tantric tradition, the original author of the extensive and now lost Primordial Buddha Tantra (Ādibuddhatantra) was the historical buddha Śākyamuni, who 47

Vesna A. Wallace

prophesized that the author of its abridged version, the Abridged Wheel of Time Tantra, will be the Bhagavān Mañjuśrī Yaśas, who will redact it for the Brahmanic sages residing in Kalāpa, the capital of the mythical land of Śambhala. In the Stainless Light Commentary, the most authoritative commentary on this tantra, which appeared in India in the early eleventh century at the same time as the tantra itself, we are told that the author of the tantra and its subject matter are one and the same. The author, who is nothing other than the ultimate truth, is revealing himself in the conventional and at times symbolic terms that constitute a tantric text. Seated on the lion throne in the Kālacakra maṇḍala, having been worshipped by the best of gods, he teaches the tantra at the request of Sucandra, the king of Śambhala, who wants to know the arrangement, dimensions, and constitution of the cosmos, the ways in which the cosmos is present in a human body, and a tantric method that leads to enlightenment. Prompted by Sucandra’s inquiry, the buddha expounds these topics in the five chapters of the tantra, dealing with the cosmos, the individual, tantric initiation, meditational, visualization practice of mentally generating oneself as the central deity and its multilayered maṇḍala, and the stage of completion, involving a tantric yoga by means of which the practitioner actualizes his buddhahood in the form of the deity that one previously visualized and with whom one imaginatively identified himself. Throughout the tantra, it becomes increasingly clear that the buddha, being both the author of the tantra and its subject matter, speaks of himself as being indivisible from the cosmos, a person’s mind-body complex, and the path and result of a tantric practice. All these are his multifarious appearances perceived differently by different sentient beings, according to their karmic obscurations. In the section called “The Great Exposition of Our Own and Others’ Philosophical Views, Logic, and Analysis,” contained in the second chapter of the Abridged Wheel of Time Tantra, the buddha Śākyamuni, in his manifestation as Kālacakra, expounds the views of different Buddhist and non-Buddhist systems of thoughts in nineteen verses in order to refute them, making this tantra the only unexcelled yoga tantra in which a doxography is expounded. He begins his doxographical exposition by first listing some of the tenets of the Kālacakra tantric system, namely: the identitylessness of persons and phenomena, maturation of karma, interdependent origination, the four bodies of the buddha, conscious emptiness (ajaḍaśūnyatā), the eighteen unique qualities of a buddha, the Four Noble Truths, and so on (Śrīkālacakratantrarāja, ch. 2, 161; Upadhyaya 1986, 255). His counterarguments to the claims made by the adherents of non-Buddhist and Buddhist systems of thought  – including Vaibhāṣikas, Sautrāntikas, and Yogācārins – are consistent with those given in the treatises of Mādhyamikas. Thus, the buddha appears to give primacy to Madhyamaka thought. After introducing his own views, he proceeds to point out some of the tenets of other systems in order to later rebut them. He begins first with a refutation of the Brahmanical assertion of the Vedas as self-arisen and uncreated (akṛtaka) like space and of the word “Veda” as inherently true: That Veda is not self-arisen. In the abode of the three worlds, the word Veda signifies the referent (artha). By the word Veda, Brahmā with his four mouths disclosed the referent long ago. But if a referent is not different from a word, then why does the uttered word “fire” not burn the mouth? Hence, indeed, there is a teacher (i.e., Brahmā) with regard to the unknown topic and with regard to the past and future meaning. The Veda is not similar to space because it is created, uttered with the mouths from different places. In terms of its use, it is local, recited by the mouths of the twice-born (brāhmaṇas); it is present everywhere, and others read it. Since the low

48

The Tantric Buddha

castes such as śūdras and others read and write it, this Veda is present everywhere. Therefore, that Veda is not a standard of knowledgeable and learned men (Śrīkālacakratantrarājasya; Upadhyaya, 1986, 260–61, vv. 166–167). Puṇḍarīka, the author of the Clear Light Commentary on the tantra, explicates these two verses of the critique of the Brahmanical tradition’s claim that the Veda is self-arisen, eternal, and inherent in its referent by first asserting that since the word “Veda” signifies a referent (artha) and is produced by means of a verbal articulation, it cannot be logically established as selfarisen. Furthermore, the word Veda cannot be logically asserted as eternal or as invariably inherent and nondifferentiated from its referent, because if a word and its referent are one and the same, then each time when the word “fire” is uttered, one’s mouth would burn from fire. Thus, it has been established that the word and referent are two separate things. Likewise, since god Brahmā has been established as the initial teacher of Vedic text, Veda is logically established as created and not as self-arisen. Moreover, since it is local in usage, recited not only by learned brāhmaṇas but also by lower castes, it is not a standard of a learned men (Śrīkālacakratantrarājasya; Upadhyaya, 1986, 160–62). The next set of rebuttals in the tantra pertains to both Śaivas’ view of God as the independent creator and Vaiṣṇavas’ position on Īśvara, who is said to incite others to action, though he himself is not considered as the agent of action. The buddha refutes these tenets from the viewpoint of interdependent origination in order to demonstrate that nothing in the world has a single cause or autonomous existence: Īśvara cannot be a creator or an independent instigator of action if he depends on others to perform actions, and if he were the one who instigates others to action, he would ultimately be the one to experience the results of karma and not someone else. Therefore, apart from the karma of living beings, there can be no creator who bestows upon a person the results of good or bad actions. In order to explain that everything, including Vaiṣṇavas’ creator Īśvara, arises due to contact with something else and is not from itself but due to the agglomeration of atomic particles, the buddha states: Here, if the atoms of the earth, water, fire, and wind in the sky originally do not belong to the creator, then in the absence of material substance (dravya), and devoid of the sense objects, the creator of all does not create anything. There is neither a perceptual ( pratyakṣa) nor an inferential ( parokṣa) means of valid cognition ( pramāṇa) for that creator who is devoid of the sense objects. Thus, everything comes into existence due to the agglomeration of atoms, O king; and the nature of karma is not due to the creator’s will. Water appears even from a moonstone and the reflection of an object in a mirror due to contact. Due to acidity, there is saliva on the tongue. Due to one’s own voice there is another voice (i.e., echo). A sprout is due to a pure seed. Due to a lodestone, there is the rolling of an iron stick. None of these occurs by the will of the creator. That efficacy of phenomena is not caused by anyone within the abode of the three worlds. (Śrīkālacakratantrarājasya; Upadhyaya, 1986, 262–63, vv. 169–170) With those two verses, Īśvara is negated as an independent creator on the grounds that in the absence of material substances, like atoms, Īśvara is unable to create anything since Vaiṣṇavas themselves claim that atoms originally do not belong to Īśvara. This refutation is grounded

49

Vesna A. Wallace

in a Buddhist view that not a single phenomenon arises from itself or by the will of a creator as a single cause, but instead arises in dependence on the sets of causes and conditions. The efficacy of a cause, or the causality of the cause, is conceived only with regard to the effect and the effect with regard to the cause. Since cause and effect arise in dependence on each other, all causes – including Īśvara – are nondetermined. In the following verses, continuing this line of argumentation from the standpoint of interdependent origination, or personal and phenomenal identitylessness, the buddha further refutes the notion of the omnipresent and permanent Self (ātman), asserting that if dependently arisen phenomena arise after they cease, then they cannot have the permanent, unchanging, autonomous Self because the ceased phenomenon no longer exists. Hence, one can speak only of the collection of phenomena in cessation but not of someone who departs to liberation or is in karmic bondage. Thus, he argues: If the Self is omnipresent, why does it experience suffering due to separation from relatives? And if that Self is permanent, why does it fall in love when hit by the arrow of love? If that Self was active in the waking state, why does it become torpid at the time of sleep? Thus indeed, that Self is neither omnipresent, nor the Lord, nor was it formerly active. There is no Self. There is an origination of interdependent phenomena, and there are pleasant and unpleasant results, which are devoid of a single agent. There is no one who departs, but there is a complete departure for liberation. There is bondage but there is not a bound one. There is existence, and there is non-existence of interdependently arisen phenomena. The world is devoid of inherent existence and ultimately free from momentariness. These true words of mine are not destroyed by the words of gods and nāgas, which are with demons of conceptualizations. (Śrīkālacakratantrarājasya; Upadhyaya, 1986, 262–65, vv. 171–172) According to the Stainless Light Commentary on these verses, the buddha is arguing that if the Self were omnipresent, then the suffering of all beings would be a suffering of one sentient being. But if one asserts the existence of many Selves, then one cannot assert the omnipresence of many Selves. The Self cannot be permanent because it is subject to change; therefore, the Self is neither omnipresent, not permanent, nor some eternal Lord. The Self of phenomena in cessation is non-existent because of the non-existence of that to which the Self would belong. Thus, there is only a collection of phenomena that are in cessation and not someone who departs. The practical relevance of this refutation of the permanent Self as taught in Hindu systems of thought is a way in which the buddha expounds the personal identitylessness ( pudgalanairātmya) to his original Brahmanic audience and to those who will study and practice the tantra in the future. He intentionally discusses the identitylessness in this second chapter of tantra, which precedes the chapters dealing with a tantric practice. This is because without understanding the identitylessness, or the absence of the permanent, autonomous Self, one cannot effectively engage in the practice of self-generation as a deity (deity-yoga). After rebutting the aforementioned views of various Hindu schools, he argues against the views of some Buddhist schools, such as Vaibhāṣikas’ view of the reality of the person ( pudgala), certain tenets of Sautrāntikas or Arthavādins, as well the Cittamātrins’ view of the reality of consciousness, and so on. As shown in what follows, in his arguments against these Buddhist schools, he does not take Nāgārjuṇa’s exposition of the four extremes (catuṣkoṭi) as a standard Madhyamaka exposition on freedom from any dogmatic position ( pakṣa), as one might expect. Instead, he focuses 50

The Tantric Buddha

on the assertion of the nonduality of emptiness and compassion. It is the author of the Stainless Light Commentary who resorts here to Madhyamaka’s exposition of the four extremes in his commentary on this verse. One (Vaibhāṣika) who asserts the reality called pudgala to be present in the body has failed due to implying the inherent nature of that pudgala. One (Sautrāntika) who asserts the object by means of conventional truth has failed, because he considers the unknown ultimate truth to be nonexistent. Likewise, one (Cittamātrin) who asserts the reality of consciousness has failed, because he considers everything within the three worlds to be consciousness. One who asserts the non-duality of emptiness and compassion is free from dogmatic positions and has not failed. (Śrīkālacakratantrarājasya; Upadhyaya, 1986, 265–66, v. 173) The author of the Stainless Light Commentary explains why the one who asserts the nonduality of emptiness and compassion is free from dogmatic positions, indirectly pointing to the buddha himself and asserting that such a one is endowed with compassion without an object and with emptiness that is free from conceptualization, as well as with the best of all aspects (sarvākāra) (Śrīkālacakratantrarājasya; Upadhyaya, 1986, 267). The next target of the buddha’s refutation is a view of the accumulation of the present karma as a result of the previously performed karma, which he claims is upheld by Tājiks (Tāyins), descendants of Iranian people and followers of a particular Islamic group whose members believe that a deceased person experiences either suffering in hell or pleasure in heaven through another human form due to pleasing Rahman, or Bismilla. It is worth mentioning here that this is the only tantra in which the author speaks explicitly about this particular Islamic group and its theology. He refutes their tenets in this manner: Tāyins hold that a person experiences his previously performed deeds in this life through another birth. If it is so, then a removal of people’s karma from one birth to another could not take place and there would be no escape from the cycle of existence, nor would there be the entrance into liberation over the course of immeasurable lives. This, indeed, is the doctrine of Tāyins. But it is discarded by other lineages. (Śrīkālacakratantrarājasya; Upadhyaya. 1986, 268, v. 174) By means of the arguments given in the following verse, he refutes the doctrine of Materialists (Cārvākas, Lokāyata), the followers of Amaraguru and Bṛhaspati who denied the efficacy of karma, a god as the creator, and the notion of an afterlife, instead propagating the efficacy of the agglomerated elements of the earth and so on in the production of living beings. This view was rejected by Buddhists on the basis that it does not explain why inanimate objects, like trees and the like, do not have consciousness. If the witnessing mind (sākṣicitta), like the power of intoxicating drink produced through the combination of different ingredients, arises due to the elements that are integrated into one, then would not trees have consciousness due to the agglomeration of the earth, water, fire, air, and space? But if those trees do not have the power of living beings, then the power of the agglomeration of the elements is utterly useless. That assertion of the Materialists does not bring agreeable results and has destroyed the path of liberation for people. (Śrīkālacakratantrarājasya; Upadhyaya, 1986, 268, v. 175) 51

Vesna A. Wallace

His last refutation relates to doctrinal tenets held by Jains, who are committed to the existence of the eternal, unchanging soul (  jīva), whose size corresponds to the size of the body, and to the permanence of atoms, and so on. The buddha argues against the Jains in this way: If the soul has the body as its measure, does it perish after the arms and legs are cut off? If an atom is permanent through the body, then why does it attain a gross state after its subtle state? Jains hold that the one who is free from karma departs from the cycle of existence to the state of bliss, which is located on the top of the world. The counterargument is: The three worlds, which are made of atoms, never remain permanent due to the efficacy of time. (Śrīkālacakratantrarājasya; Upadhyaya, 1986, 268, v. 176) The buddha concludes these discussed refutations of the tenets of different systems of thought with the assertion that his refutation of the aforementioned tenets of different Hindu and Buddhist systems of thought is given in accordance with the peoples’ dull, sharp, and other – meaning low – mental faculties, which are due to the habitual propensities (vāsanā) of people’s karma. His reason for refuting those tenets is to impart knowledge to those who hold them. At the same time, he warns his audience not to criticize any Dharma, whether it is one’s own or of another lineage on the grounds that the mind, like a crystal, becomes colored by colors of the object in proximity. It is not clear whether this statement expresses his concern that by studying the tenets of others in order to demolish them, one’s own beliefs may be influenced by them, or his concern that the spirit of criticism (dūṣaṇa) may negatively influence one’s mind. The Stainless Light Commentary is silent on this point, and does not give us clues. Just when one expects the buddha to end his rebuttals, he continues with his warning against the sacrificial, ritual violence authorized by the Vedas, as he advises the king Sucandra: The Dharma free from mundane concerns is a service to sentient beings, and the non-Dharma is harm to them. Violence, which has the Vedas as authority, does not yield pleasant rewards and always gives suffering. In the words of a fool, a true love, attached toward all sentient beings, produces supreme bliss. Therefore, for the sake of sentient beings cultivate with your mind only the lack of inherent existence, O king. (Śrīkālacakratantrarājasya; Upadhyaya, 1986, 271, v. 178) Finally, in the conclusion of this doxographical exposition, the buddha affirms his universal spiritual authority, stating: I am Indra, the guru of thirteen men in heaven, a cakravartī on earth, the king of nāgas in the underworld, revered by serpents. I am omnipresent and foremost. I am gnosis, a Buddha, the Lord of Sages, the indestructible and supreme sovereign, the yogīs’ indestructible union, the Veda, self-awareness, and the purifier ( pavitra). King, take refuge in me with all your being. (Śrīkālacakratantrarājasya; Upadhyaya, 1986, 271, v. 179) The criticism of the Brahmanic interpretation of the Vedas and the sentence in which the Buddha is identified with the Veda may perplex the reader of this tantra. It may be even more perplexing in lieu of his statement given in another section of the Wheel of Time Tantra, where he presents himself to the Brahmanic sages of Kalāpa, whom he seeks to convert to the Wheel of Time Tantra, as the ultimate author of the four Vedas. The buddha allows himself to make 52

The Tantric Buddha

such claims on the basis of his self-description as a single, unitary, and all-pervasive ultimate reality, as a source of all phenomena and all teachings, understood and interpreted by different sentient beings in accordance with their own mental dispositions.

Tantric Buddhas’ Teaching on Emptiness Just as in the Wheel of Time Tantra, in other tantras, including those already mentioned, all buddhas teach the doctrine of emptiness (śūnyatā), or identitylessness, each in his own presentational way in accordance with the contexts of the individual tantric systems. A reader familiar with Mahāyāna canonical sources can easily notice that some of the most foundational doctrinal tenets taught by a buddha echo those taught in Mahāyāna Sūtras. It is for this reason that many Indian commentators on Buddhist tantras saw Vajrayāna as inseparable from Mahāyāna, calling it a “Profound (gambhīra) Mahāyāna,” “Mantra-Mahāyāna,” “Great Mahāyāna” (mahāmahāyāna), and so on, differentiating it from the system of perfections ( pāramitānaya) only in its method and in the amount of time it takes the practitioner to achieve awakening.3 In Buddhist tantras, where the system of perfections and the system of mantras (mantranaya) is synthesized, the buddha propounds the doctrine of emptiness (śūnyatā) in ways similar to those expressed in the Perfection of Wisdom literature and in the works of Indian Mādhyamikas. Emptiness is one of the most central doctrinal tenets espoused in all Buddhist tantras. In the Vajra Crest Tantra (Vajroṣṇīṣatantra), the transcendence of wisdom ( prajñāpāramitā) that sees emptiness is defined as awakening (bodhi), foundation (ālaya), liberation (mukti), and a fulfiller of all hopes (cited in Dwivedī and Lāla 1990, 62). In the Hevajratantra, Hevajra Vajrī speaks of emptiness, or the absence of inherent existence, of one’s sense faculties and their corresponding sense objects as arising interdependently: Neither form nor that which sees (eye) it; neither sound nor that which hears (ear) it, neither aroma nor that which smells it, neither taste nor a taster, neither touch nor that which touches, neither the mind nor a thinker exists by its own nature (svarūpeṇa). (Tripathi and Negi 2001, 51) In another section of the same tantra, he illustrates emptiness in terms of the interdependent origination of all phenomena with these examples: For instance, a fire arises in dependence on sticks, on rubbing, and on the effort of a person’s hands. That fire dwells neither in sticks, in the rubbing, nor in the effort of a person’s hands. Although being searched for everywhere, it is not found anywhere. But that fire is neither a true thing nor is it false. Yoginīs should think of all phenomena in that way. (Hevajratantra 3.37; Tripathi and Negi 2001, 51) In the same tantra, Hevajra declares that the purity of all things is thusness (tathatā), which, due to being non-localized, is all-pervading. In the Secret Assembly Tantra, the five buddhas – who represent here the imperishable reality (tattvākṣara) of the body, speech, mind, secret, and Dharma of all buddhas – assert that the mind of tathāgatas, known as the mind of awakening (bodhicitta), is the ultimate nature of all phenomena; that is, emptiness, or identitylessness. A realization of each of the five buddhas is expressed in their individual states of meditative concentration (samādhi) that correspond to their particular realizations. Thus, among them, buddha Vairocanavajra, having entered the 53

Vesna A. Wallace

meditative concentration called “the vajra of the realization (abhisamaya) of all tathāgatas,” describes the mind of awakening in this way: One’s own mind, which has abandoned all phenomena and is devoid of the psychophysical aggregates (skandha) and sense bases (āyatana), or devoid of the apprehender (grāhaka) and the apprehended (grāhya), is primordially non-arisen and is of the nature of emptiness due to the equality (samatā) of the identitylessness of phenomena. (Guhyasamājatantra 2.3; Matsunaga 1978, 10) He thus asserts that the mind of all buddhas, which is empty by nature, when devoid of all psychophysical components forming the ordinary mind-body complex, which are the basis of dual perception, is all-pervasive, just like emptiness. Here also the buddha Akṣobhyavajra abiding in the meditative concentration called “the vajra of the indestructibility of all tathāgatas,” teaches that phenomena are ultimately non-arisen; hence, neither their characteristics nor their reality (dharmatā) can be found. On that basis, he likens identitylessness to space. The buddha Ratnaketuvajra, abiding in the meditative concentration called “the vajra of the identitylessness of all tathāgatas,” teaches that all phenomena are without becoming, they are devoid of the characteristics of phenomena and are arisen from the identitylessness of phenomena. The buddha Amitāyusvajra, while abiding in the meditative concentration called “the vajra of the lamp of the flames of gnosis of all tathāgatas,” declares that if phenomena are non-arisen, then there is neither becoming nor is there a meditation (bhāvanā) on phenomena. Finally, the buddha Amoghasiddhivajra declares from his meditative concentration called “the overpowering vajra of all tathāgatas” that phenomena are luminous ( prabhāsvara) by nature, utterly pure, and similar to space, that from this perspective, ultimately “there is neither awakening (bodhi) nor is there realization (abhisamaya)” (Guhyasamājatantra 2.4–7; Matsunaga 1978, 10). As in the Eight Thousand-Line Perfection of Wisdom Sūtra, in the Wheel of Time Tantra, the Vajrī maintains that ultimately there is neither a buddha, nor awakening, nor compassion that benefits others (Śrīkālacakratantrarāja Vol. 2, 4.6; Dwivedī and Bahulkar 1994a, 153). Moreover, like in the Eight Thousand-Line Perfection of Wisdom Sūtra, when speaking of the emptiness of all phenomena, he compares it to space and teaches meditation on space as a meditation on emptiness. We are told in the Secret Assembly Tantra that space is not conjoined with any phenomena, nor is it not non-conjoined. All phenomena are dreamlike appearances, as if arisen in a dreaming state. Space is formless, invisible, and inexplicable, and this is how all phenomena should be understood. The elements of body, speech, and mind and the space element are nondual. All phenomena abide in the space element, but the realm of space does not abide anywhere – neither in the desire realm, nor the form realm, nor in the formless realm. The realm of phenomena (dharmadhātu) does not abide within the triple world and is without origination. That which is without origination is not produced by any phenomenon. Therefore, all phenomena are devoid of inherent existence (niḥsvabhāva). The mind of awakening, which gives rise to the state of a Vajrī or to the one who has attained the vajra-like indestructible Perfect Awakening, which, in turns, brings forth the body, speech, mind, and gnosis of all tathāgatas, does not dwell in the body, speech, or mind. Whatever phenomenon does not dwell within the triple world, it does not have the origination. It does not occur to a sleep: “I should manifest the dream within the triple world.” Nor does it occur to the person: “I should see the sleeping state.” The activity of the triple world is similar to a dream and is arisen from sleep. As many buddhas and bodhisattvas there are and as many

54

The Tantric Buddha

sentient beings there are in all world systems (lokadhātu) within the ten directions, they all should be understood to be identityless and dreamlike. The non-origination and non-abiding of all phenomena are further illustrated with an example of the way in which one obtains gold or a gem, requesting it from the wish-fulfilling jewel that grants all jewels. Those who ask for gold or a jewel from the wish-fulfilling jewel are said to obtain it merely by mind, but the acquired jewel or gold does not dwell in the person’s mind, nor is it present in the wish-fulfilling jewel. It is in this way that phenomena should be understood (Guhyasamājatantra, ch. 15; Matsunaga 1978, 84). Similarly, all mantra-siddhis, which are ultimately non-arisen, are, nevertheless, present in one’s own body, speech, and mind – but those body, speech, and mind are not located in the desire realm, in the form realm, or in the formless realm. We are further instructed that the mind is not located in the body, and the body is not located in the mind, and that speech is not located in the mind, and the mind is not located in the speech, because, like space, they are pure by nature. One’s own body, speech, and mind abide in space, and space does not abide anywhere. Furthermore, in the same tantra, in the context of meditation on the recollection of non-origination of phenomena, the presiding lord (adhipati) of the vajras of the body, speech, and mind of all tathāgatas proclaims that everything is luminous by nature, signless (nirmitta), neither dual nor nondual, peaceful, similar to space, and entirely stainless (Guhyasamājatantra, ch. 7, vv. 34–35: Matsunaga 1978, 23).

A Buddha as a Teacher of the Nonduality of Sam  sāra and Nirvāna In the higher classes of tantras, the yoga and unexcelled yoga tantras, the ultimate nonduality of saṃsāra and nirvāṇa is expressed in different contexts and in a variety of ways. The buddhas’ teachings in these tantras on the nonduality of saṃsāra and nirvāṇa is actually a teaching on the very nondual, awakened mind of a buddha. In the Hevajratantra, for instance, The Vajrī asserts that apart from saṃsāra, there is not some other nirvāṇa. Saṃsāra is the form, sound, and so on; saṃsāra is the aggregate of feeling and so on; saṃsāra is the sense faculties, and saṃsāra is aversion and other mental afflictions. These phenomena are actually nirvāṇa but have the form of saṃsāra due to delusion. The non-deluded one, while transmigrating in saṃsāra, brings saṃsāra to cessation through purification. Nirvāṇa is the mind of awakening (bodhicitta), which has the ultimate and conventional forms. (Hevajratantra 9.2–3, Tripathi and Negi 2001, 104–5) In this tantric system, purity (śuddhi) is of the nature of self-awareness (svasaṃvedya); a self-aware bliss arises due to the pure nature of the sense objects. Although that innate purity is released by means of various purifications, it is not released by means of some other purity. On the basis that the sense objects and the sense faculties are pure, self-awareness is defined as the highest bliss. In the understanding of the eleventh-century Indian Buddhist scholastic and an abbot of Vikramaśīla monastery, Ratnākāraśānti, this statement implies that innate purity is not liberated through the purification of mental obscurations (āvaraṇa), which are accompanied by habitual propensities (vāsanā), because purification is attainable only by the path meditated upon for a long time. However, everything is pure by nature (Tripathi and Negi 2001, 105).

55

Vesna A. Wallace

In the Compendium of Tattvas King of Tantras (Tattvasaṃgrahatantrarāja), belonging to the class of yoga tantras, the buddha, the tathagata, speaks of the way in which the innate luminosity of the mind comes into manifestation through the systematic removal of mental defilements: O son of the noble lineage, this mind is luminous by nature, like the moon disc. Just as the moon disc is luminous by nature, so is gnosis. Just as the moon disc gradually becomes full, so also the jewel of the mind, which is luminous by nature, becomes full. Just as the moon disc is sequentially seen with incidental lunar digits due to the departure of the rays of the sun disc, so also the jewel of the mind, which is pure by nature, is seen full of the qualities of a Buddha due to the gradual departure of the spots of stains of mental afflictions. (cited in Dwivedī and Lāla 1990, 22) In response to various questions posed by Great Bodhisattvas, in the Secret Assembly Tantra, the assembly of tathāgatas declares that all phenomena are nondual (advaya) but characterized, or perceived, with a dual nature (Guhyasamājatantra 18.126a; Matsunaga 1978, 122). We find a similar statement in the Saṃpuṭatantra, the explanatory tantra belonging to the Hevajra cycle, where: The vajrī said that saṃsāra is a mind smeared by unbearable (durvāra) stains of attachment and the like. The luminous ( prabhāsvara) mind, free from ideation (kalpanā), has the smear of the stains of attachment and other mental afflictions removed. Due to the absence of grasping there is neither the apprehending subject (grāhaka) nor is there the apprehended object (grāhya). He said, this is the most excellent nirvāṇa. (Śrībuddhakapālatantrarāja cited in Śrībuddhakapālamahātantrarājaṭīkā of Abhayākāragupta 2009, ed. Dorje, 14, vv. 2b – 3) This twofold appearance of the mind is explained in the Vajra Ornament Tantra (Vajra­ maṇḍālaṃkāratantra) in terms of its ultimate and phenomenal aspects: The mind (cintā) that is free of defilements (anāśrava) and the like is stainless (nirmala), free from “mine” and “I.” It is empty, similar to space, and devoid of existence and non-existence. It is without beginning and end, peaceful, the master of all phenomena; it bears all forms and rests in the two truths. In terms of the ultimate truth, this supreme mind is known as the Dharma Body. In terms of conventional truth, it is consciousness (vijñāna), the mind (citta) that has all sense faculties at all times and everywhere. It is called “mental awareness” (manojñāna), wisdom, ( prajñā), intellect (buddhi), thought (mati), mindfulness (smṛti), and insight (vipaśyanā) because of the purification of a wherever-arisen gnosis. It is free from eternalism and nihilism and is devoid of the duality of the beginning and end. There are two conceptions ( prakalpanā) of it in terms of the conventional and ultimate truths. Just as a source of fire is established in dependence on kindling, or just as sesame oil is conceived to be in sesame seeds and sap in a sugar cane, fragrance in flowers,

56

The Tantric Buddha

and fresh butter in milk, so also the mahāmudrā, present in both the form and formlessness, dwells everywhere, in all beings: stationary and mobile. (Cited in Dwivedī and Lāla 1990, 58–59) In the Vajra Crest Tantra, the buddha replies to Vajrapāṇi’s question as to how is the ultimate reality ( paramārthasatya) is an abode (ālaya) of all phenomena by stating: Lord of Secrets, all phenomena that are of the nature of psychophysical aggregates and the sense-bases, and the inanimate and animate worlds are the support of the ultimate reality due to their impermanence. The ultimate reality is without the Self (anātmaka); it is unsurpassed and entirely empty; it neither increases nor diminishes; it is entirely pure, similar to space, and is of the stainless nature, imperishable, inexpressible, and devoid of the body, speech, and mind; it does not exist nor does it non-exist; it does not go anywhere nor does it not go; it is not far, nor it is near; it is neither empty nor not empty, nor it is middle in between the two. This is the ultimate reality. Therefore, this supreme, profound, and difficult to comprehend non-origination and non-cessation of all phenomena has conventional reality (saṃvṛtisatya) as its co-operating cause ( pratyaya), which itself is arisen from causes and conditions (hetupratyaya). (Cited in Lāla 2001, 65–66) Nirvāṇa is defined in the Wheel of Time Tantra as not having the beginning, middle, and end, as that which does not have the sense objects, the mind (citta), the material nature ( prakṛti), a person, a bondage, liberation, the agent, the seed, the time of manifestation, and as that which is not of the nature of suffering and happiness in the entire world. It is signless (nirnimitta), free of causes, and without qualities (Śrīkālacakratantrarāja Vol. 3, 1994b, Dwivedī and Bahulkar, 151, v. 244)

Tantric Buddhas’ Teachings on Gnosis In addition to emptiness, the nondual, self-aware gnosis (svasamvedya-jñāna) is a central doctrine taught by buddhas in the unexcelled yoga tantras, where the tantric paths to awakening are constructed around this concept. This gnosis is the supreme, immutable, and nondual mind of all the buddhas, connate to all sentient beings but not yet actualized or directly perceived by them. It is from that gnosis, which is indivisible from wisdom and compassion, that all tantras emerge in response to the requests posed by various, ten-stage bodhisattvas. In the Hevajratatantra, the Vajrī speaks of gnosis in these words: Gnosis is free from the notion of oneself and another due to self-awareness. It is similar to space, free from stains, and is empty. It is of the nature of existence and non-existence, and is supreme. It is merged with wisdom and compassion and with passion (rāga) and the absence of passion (virāga). (Hevajratantra 10.8b – 9, Tripathi and Negi 2001, 113–14) In this tantric system, gnosis is understood to be similar to space because it perceives itself and others as being of the same essence (samarasa), and it is merged with passion and the absence

57

Vesna A. Wallace

of passion because the nondual, innate gnosis is passion, or a sublime bliss (mahāsukha), and it is also the absence of ordinary passion (virāga) (Tripathi and Negi 2001, 114). The buddha also identifies gnosis as a self-knowledge (ātmavidyā) in the Wheel of Time Tantra, which is imperishable (akṣara), and from which the entire three worlds arise due to the experience of bliss and into which they dissolve at the time of death. That kind of gnosis, he claims, is not associated with scriptures of heretical groups or with the Vedas, Aṅgas, Smṛtis, Siddhāntas, and so forth. Owing to gnosis, one becomes a buddha, who is a unique teacher, possessing supernatural powers. A collection of the elements in one’s body, which is due to the consumption of the six tastes present in the food and drink, arises from space. Therefore, when one’s gnosis merges into space by means of a yogic, meditative practice, it becomes like space, eternal, indestructible, and of the same taste. The realized peace, or gnosis, is present in the elements within the triple world and should be known to be within one’s own body (Śrīkālacakratantrarāja, vol. 1, Upadhyaya, 1986, v. 3, 159, Ch. 2, 221, v. 96, 221; Dwivedi and Bahulkar, 1994a, Vol. 2, Ch. 3, v. 86, 89). In the Supreme Primordial Buddha Tantra, cited in the Stainless Light Commentary, the Primordial Buddha (Ādibuddha) describes the buddha’s gnosis with its five aspects in his discussion of meditation on the five psychophysical aggregates as ultimately being of the nature of the five types of gnosis. The mirror-like gnosis, which is the purified aggregate of form, knows that all phenomena in space are devoid of the nature of ideation (kalpanā) and are seen like an image in a virgin’s prognostic mirror. The gnosis of equality, or the aggregate of feeling, knows that after achieving union with all phenomena, there is only a single, indestructible phenomenon that abides; and arisen from the imperishable gnosis, that single phenomenon is without cessation and without eternity. The discriminating gnosis, or the purified aggregate of discernment, knows that in non-arisen phenomena, which are free from mental formations (saṃskāra), there is neither awakening nor buddhahood, nor a sentient being, nor living beings. The accomplishing gnosis, or the purified aggregate of mental formations, knows the phenomena that transcend the reality of consciousness (vijñānadharmatā) and that are purified in gnosis, untainted, and luminous by nature as the realm of phenomena (dharmadhātu). These five types of the gnosis characterizing the mind of the buddha are represented by five paradigmatic awakened figures, or buddhas: Akṣobhya, Ratnasaṃbhava, Amitābha, Amoghasiddhi, and Vairocana, who also represent the five types of awakened families. Thus, one could say that although the awakened author of the tantra is ultimately a single author, he speaks from the perspective of his five, mutually interconnected types of gnosis manifesting as the five types of buddhas.

Tantric Buddhas’ Teachings on the Naturally Luminous Mind and Its Obscurations Due to the all-pervasiveness of a buddha’s gnosis, the mind of sentient beings is ascertained in various tantras as inseparable from the nondual enlightened awareness. In the Wheel of Time Tantra, it is pointed out that just as space does not disappear from a jar when water is poured into the jar, “so also the sky-vajrī, who pervades the universe and is devoid of sense objects, is within the body” (Śrīkālacakratantrarājasya; Upadhyaya, 1986, Vol. 1, Ch. 2, 219, v. 91c–d). In the Hevajratantra, we are reminded that although the divine pervader (vyāpaka) dwells in the body of all things (vastu), he is not born from the body. When asked by Vajragarbha in the Hevajratantra why the body that is pure by nature and primordially devoid of inherent existence becomes the body composed of the elements, the Vajrī does not offer a fresh answer 58

The Tantric Buddha

to the question. Instead, to some degree, he reiterates what he already taught elsewhere in this tantra, pointing out that stillness (nistaraṅga), interpreted by the commentator Kāṇha to mean “the lack of conceptualizations,” is its nature that dwells in every body (Hevajratantram; Tripathi and Negi 2001, Part 1, Ch. 1, v. 12b, Ch. 10, vv. 35–36, 14, 123). As shown in what follows, we are further told in the Hevajratantra that although the six sense faculties, the five psychophysical aggregates, the six sense objects, and the five elements are pure by nature, they are obscured by ignorance and mental afflictions and thereby manifest as cyclic existence. The sense faculties remain unpurified for as long as the apprehending subject (grāhaka) and the apprehended object (grāhya) arise. However, by whatever internal or external sense object and the like one is bound, by that same sense object the bondage is released. The one who has become purified from mental obscurations sees all sense objects as pure by nature. For such a one there is no longer a scent, a sound, a form, and so on, because to that one the world is made of buddhas (Hevajratantram; Tripathi and Negi 2001, Part 1, Ch. 9, vv. 4–5, 19–20, 106). Kālacakra gives a similar teaching in the Wheel of Time Tantra, wherein the person, like a silkworm, binds himself with conceptualizations (vikalpa) and liberates himself with his mind when conceptualizations are no longer present. The mind has four different powers (śakti): the power of desire; the power of activity (kriyā); and the power of cognition – which respectively give rise to the perception of phenomena, the penetration into phenomena, and the understanding of the meaning of phenomena – and the nondual power, which does not grasp or relinquish anything. The person’s karma and the sense faculties are most intimately connected because karma is present in the sense objects, referred to in this tantra as the qualities (guṇas) of one’s own materiality ( prakṛti), which is the psychophysical aggregates, elements, and sense faculties. In saṃsāra, suffering and happiness are contained in those material constituents and are produced by karma, which can be of three types: gross, subtle, and peaceful. When the thought “I am an agent” arises, this should be known as a “wrong karma”; when the thought “Someone else or Paśupati is an agent” arises, it is called “karma.” The one who is separated from the constituents of the aforementioned materiality realizes, “Neither I nor someone else is an agent”; this is considered not to be karma. Therefore, neither the Self nor some creator gives or takes away the happiness and suffering of beings. It is the previously performed action that brings about its corresponding result. The mind of deluded persons alone is the agent of creation and destruction; it gives and takes away suffering and happiness. The rebirth that one enters bound by karmic fetters is due to one’s own mind. This world, then, is constructed by the karma of sentient beings (Śrīkālacakratantrarājasya; Upadhyaya 1986, Vol. 1, 85–91, 118–21).

Tantric Buddhas’ Teachings on the Body, Passion, and Bliss In the yoga and unexcelled yoga tantra classes of Indian Buddhist tantras, the generation of bliss is deemed indispensable for the actualization of gnosis (  jñāna) and for the attainment of the ultimate siddhi of awakening. These tantras repeatedly emphasize a soteriological significance of the body. In the Mundane and Supramundane Vajra Tantra (Laukikalokottaravajratantra), it is declared: The siddhi of awakening is obtained by means of bliss; the siddhi is not obtained by means of bodily mortifications because the buddhahood, arisen in meditative concentration is due to all bliss. Due to bodily mortification through severe fasting, thirsting, and the like a mental distraction (vikṣepa) arises; and due to distraction, there is no siddhi. Inferior beings 59

Vesna A. Wallace

(hīnasattva) who practice mortification do not succeed. Therefore, they are declared to be the ones who have a difficult task. (Cited in the Lupta Bauddha Vacana Saṃgraha, Dwivedī and Lāla 1990, Vol. 1 55) Similarly, in the Wheel of Time Tantra, one is advised to preserve the body for the sake of attaining the siddhi because in the absence of the body, one cannot utilize the body in tantric practices to generate sensual bliss and thus can attain neither a siddhi nor supreme bliss in this life (Śrīkālacakratantrarāja; Upadhyaya 1986, Upadhyaya, Vol. 1, Ch. 2, v. 107, 228). It is for this reason that only embodied human beings can practice tantra. The idea that a saṃsāric mind, which is invariably embodied in either a subtle or a gross material form, can achieve the supreme bliss only through the repeated generation of bodily bliss by means of tantric yogic techniques is unique to tantras. In the Hevajratantra, the Vajrī Hevajra makes a similar point by stating: How could there be bliss (saukhya) in the absence of the body? One cannot speak of bliss in the absence of the body. The world is permeated by bliss that is of the nature of the pervader (vyāpaka) and the pervaded (vyāpya). Just as a scent depends on a flower and is unobtainable in the absence of the flower, in the same way, due to the absence of form and other psychophysical aggregates, bliss is not obtained. (Hevajratantram; Tripathi and Negi 2001, Part 2, Ch. 2, vv. 35b–c and 36, 144) On the basis that awakening can be achieved in this very life, similar statements spoken by the buddha are scattered throughout many tantras. In the Hevajratantra, the Vajrī Hevajra also states: There can be no siddhi of awakening by means of all the Vedas and Siddhāntas nor by means of the multitude of rituals because there is still a rebirth in another world due to purification. Without that bliss there is no siddhi in this life or in the other world. (Hevajratantram; Tripathi and Negi 2001, Part 1, Ch. 8, vv. 54–55a, 102) According to the Ḍākinī Vajra Enclosure (Ḍākinīvajrapañjara), the world arises due to passion and reaches its destruction due to discarding passion. It is due to the complete knowledge of the vajra-passion that the mind becomes Vajrasattva (cited in the Lupta Bauddha Vacana Saṃgraha, Dwivedī and Lāla 1990, Vol. 1, 22). In the Hevajratantra, the Vajrī emphasizes that by passion, meaning, mundane passion, the world is bound, and it is liberated by passion (of innate bliss), like a person who knows the nature of poison and can destroy the poison from which all people die. The sublime, connate bliss is eternal and supreme, and it manifests as the five types of gnosis. Ultimately, the connate bliss is the entire world and its nature is nirvāṇa (Hevajratantram; Tripathi and Negi 2001, Part 2, Ch. 2, vv. 44, 46, 51, 59a, 148–50, 152). According to the Wheel of Time Tantra, to achieve buddhahood, a meditation on emptiness alone is insufficient. Just as grape does not come from a nimba tree, or nectar from poison, or lotus from a brahma tree, in the same way, the buddhahood and connate bliss are millions of eons away for the person who, devoid of immutable bliss, meditates on emptiness (Śrīkālacakratantrarājasya; Dwivedi and Bahulkar 1994a, Vol. 2, Ch. 5, 38, v. 72). 60

The Tantric Buddha

Tantric Buddhas on the Buddhahood of Sentient Beings The view of all sentient beings as being ultimately buddhas, expounded in the unexcelled yoga tantras, has already been either implicitly or explicitly pointed out in various statements cited in this chapter. This identification is laid out in accordance with the individual models of tantric yogic practices developed in different tantras. In the Discourse in One Hundred Thousand Stanzas (Lakṣābhidhāna), and the Saṃvara Root Tantra, Vajrasattva is also referred to as a sublime bliss, which dwells in the mind of all sentient beings (cited in the Stainless Light Commentary on the Kālacakratantra,; Upadhyaya 1986, Vol. 1, Ch. 1, 33). As indicated previously, in the Hevajratantra, Hevajra explicitly states that all sentient beings are buddhas, but their minds are obscured by incidental defilements (āgantukāsrava). The fact that these defilements are not essential to sentient beings and can be removed means that sentient beings are essentially buddhas but ignorant of their true nature. From the Vajrī’s perspective of sentient beings, The denizens of hell, pretas, animals, gods, humans, and asuras, even the worms on the excrement and such are always blissful by nature. No Buddha is found elsewhere, in some place within the world systems (lokadhātu). Since the mind (citta) alone is a perfect Buddha there is no Buddha elsewhere. (Hevajratantram; Tripathi and Negi 2001, Part 2, Ch. 4, vv. 64–65a, 73–75, 188–90) In the Wheel of Time Tantra, the buddha gives nearly the identical statement, saying: Sentient beings are buddhas, and there is not some other great Buddha in this world system. By serving them, the unlimited cyclic existence is cut off due to the absence of conceptualizations. (Śrīkālacakratantrarāja, vol. 3, Dwivedi and Bahulkar 1994b, 37, vv. 66a – b) In different unexcelled yoga tantras, a buddha gives different explanatory models of the association of the Four Bodies of the Buddha4 with living beings’ four bodily cakras and with the four states of the mind. Thus, in the Hevajratantra, the four bodies of Hevajra – the fourth being the body of sublime bliss (mahāsukhakāya) – are said to be below the navel, below the heart, in the throat, and on the head, respectively. An emanation body, considered here to be a material body (sthāvara), is in the womb cakra ( yoni-cakra) because this is where a sentient being arises and is fashioned (nirmīyate). On the basis of the view that the Dharma is of the nature of the mind, it is associated with the heart cakra. An enjoyment body, which is said to correspond to one’s enjoyments of the six tastes, is placed in the throat; and the cakra of sublime bliss is on the head (Hevajratantram; Tripathi and Negi 2001, Part 2, Ch. 4, vv. 51–55, 183–84). In the Wheel of Time Tantra, the locations of the buddha Kālacakra’s four bodies in the person’s bodily cakras differ from those given in the Hevajratantra. In this tantra, the fourth body, here called the Innate (sahaja), or Pure (śuddha) body, is associated with three cakras – the secret cakra, the crown cakra, and the navel cakra. While the locations of the Dharma body and enjoyment body correspond to those taught in the Hevajratantra, an emanation body is here associated with the navel cakra. Moreover, the four bodies of the buddha Kālacakra are identified with the waking state, dreaming, deep sleep, and the state of bliss, respectively. In the waking state, the mind experiences the world by means of the body and speech; in the 61

Vesna A. Wallace

dreaming state, due to the power of the habitual propensities of the sense objects, the mind reflects on objects invisible to others; in the deep sleep state, it is devoid of the sense faculties; and in the fourth state, it attains bliss (Śrīkālacakratantrarāja; Upadhyaya 1986, Vol. 1, Ch. 2, v. 27, 170; and Dwivedi and Bahulkar, 1994a, Vol. 2, Ch. 4, v. 107, 202).

Conclusion As author of a Buddhist tantra, a buddha thus takes a normative stance when teaching the doctrinal tenets that express his ontological and epistemological views regarding his ultimate nature and the nature of the world that is pervaded by him and ultimately indivisible from him. He repeatedly directs the attention of his audience to the fact that he is their true nature, and he does so by resorting to the principles and conventions of a tantric tradition. A tantra as a text, whether oral or written, is understood by tradition as an authoritative discourse; it is a linguistic expression of a buddha’s mind, given with the intent to bring others to understand his truths and to convert them, as seen in the Wheel of Time Tantra, whereby he engages in the work of a philosopher and a doxographer. Thus, the act of listening or reading a tantra is meant to be an event of interacting with the buddha’s mind, through which one can begin to know awakened mind and to transform one’s own mind.

Notes 1 Some of his other names mentioned in Buddhist tantras are: pledge being (samayasattva); sublime vajra (mahāvajra); reality (tattva), the holder of the body, speech, and mind vajras of all tathāgatas; sublime bliss (mahāsukha); imperishable bliss (akṣarasukha); sublime vajra (mahāvajra); All-Good (samantabhadra); the Self of all beings (sarvātma); the innate body (sahajakāya); natural body (svabhāvikakāya); gnostic body (  jñānakāya); pure body (śuddhakāya); vajra body (vajrakāya); gnosis-vajra (  jñānavajra); gnosis indivisible from wisdom and compassion; sublime prāṇa (mahāprāṇa); the vajrayoga; Primordial Buddha (ādibuddha); and so on. 2 A complete list of the six qualities is given in the Supreme Primordial Buddha Tantra, cited in the Sekoddeśaṭīkā 1941, 3. See Ratnākāraśānti’s Muktāvalī Pañjikā for a slightly different reading (2001, 57). 3 For different interpretations, see Wallace (2011, 95–111). 4 In Vajrayāna, the four bodies of the buddha are the four aspects of Awakened mind. The Gnosis body (  jñānakāya) is the blissful aspect of the gnosis of emptiness, and the Dharma body (Dharmakāya) is an empty aspect of the awakened mind. These two bodies are accessible only to buddhas. The enjoyment body (Sambhogakāya) manifests as luminous, divine form in which the awakened mind appears to those whose mental obscurations are diminished, teaches them, and blesses them; and the emanation body (nirmāṇakāya) is a form in which the awakened mind appears to ordinary sentient beings due to their mental obscurations. It is believed that a single buddha can simultaneously manifest in innumerable emanation bodies in countless world systems. Thus, buddhas manifest differently to different beings in accordance with their mental obscurations and propensities.

Bibliography Abhayākāragupta. 2009. Śrībuddhakapālamahātantrarājaṭīkā. Edited by Chog Dorje. Bibliotheca IndoTibetica Series, vol. 68. Sarnath, Varanasi: Central Institute of Higher Tibetan Studies. Bagchi, Sitansusekhar, ed. 1965. Guhyasamāja Tantra or Tathāgataguhyaka. Darbhanga: The Mithila Institute. Dwivedī, Vrajavallabha, and Banārasī Lāla, eds. 1990. Lupta Bauddha Vacana Saṃgraha, vol. 1. Rare Buddhist Texts Series, vol. 6. Sarnath, Varanasi: Central Institute of Higher Tibetan Studies. Dwivedī, Vrajavallabha, and S. S. Bahulkar, eds. 1994a. Śrīkālacakratantrarājasya Kalkinā Śrīpuṇḍarikeṇa Viracitā Ṭīkā Vimalaprabhā, vol. 2. Rare Buddhist Texts Series, vol. 12. Sarnath, Varanasi: Central Institute of Higher Tibetan Studies. ———. 1994b. Śrīkālacakratantrarājasya Kalkinā Śrīpuṇḍarikeṇa Viracitā Ṭīkā Vimalaprabhā, vol. 3. Rare Buddhist Texts Series, vol. 13. Sarnath, Varanasi: Central Institute of Higher Tibetan Studies.

62

The Tantric Buddha Lāla, Banārasī, ed. 2001. Luptabauddhavacanasaṃgraha, vol. 2. Rare Buddhist Text Series, vol. 25. Sarnath, Varanasi: Central Institute of Higher Tibetan Studies. Matsunaga, Yukei, ed. 1978. The Guhyasamāja Tantra. Osaka: Toho Shuppen, In Indo-Tibetica Series, vol. 48. Sarnath, Varanasi: Central Institute of Higher Tibetan Studies. Newman, John R. 1987. “The Outer Wheel of Time: Vajrayāna Buddhist Cosmology in the Kālacakra Tantra.” PhD diss., University of Wisconsin. ———. 1992. “Buddhist Siddhānta in the Kālacakra Tantra.” Wiener Zeitschrift für die Künde Südasiens 36: 226–34. Sekoddeśaṭīkā of Naḍapāda (Nāropā): Being a Commentary of the Sekoddeaśa Section of the Kālacakra Tantra. 1941. Sanskrit text edited by Mario E. Carelli. Gaekwad’s Oriental Series. Baroda: Oriental Institute. Shashni, Thinlay Ram, ed. 1997. Bauddha Tantra Kośa, vol. 2. Rare Buddhist Text Series, vol. 15. Sarnath, Varanasi: Central Institute of Higher Tibetan Studies. Snellgrove, D. L. 1959. The Hevajra Tantra: A Critical Study. Parts 1–2. London Oriental Series, vol. 6. London: Oxford University Press. Tripathi, Ram Shankar, and Thakus Sain Negi, eds. 2001. Hevajratantram Mahāpaṇḍitācāryaratnākā raśāntiviracita Hevajrapañjikā-Muktāvalī Saṃvalitam. Bibliotheca Indo-Buddhica Series. Sarnath, Varanasi: Central Institute of Higher Tibetan Studies. Upadhyaya, Jagannatha, ed. 1986. Śrīkālacakratantrarājasya Kalkinā Śrīpuṇḍarikeṇa Viracitā Ṭīkā Vimalaprabhā, vol. 1. Bibliotheca Indo-Tibetica Series, vol. 11. Sarnath, Varanasi: Central Institute of Higher Tibetan Studies. Wallace, Vesna A. 2001. The Inner Kālacakratantra: A Buddhist Tantric View of the Individual. New York: Oxford University Press. ———. 2004. The Kālacakratantra: The Chapter on the Individual together with the Vimalaprabhā. Tanjur Translation Initiative, Treasury of Buddhist Sciences Series. New York: American Institute of Buddhist Studies, co-published with Columbia University’s Center for Buddhist Studies and Tibet House (Annotated translation from Sanskrit, Tibetan, and Mongolian, together with a critical edition of the Mongolian text). ———. 2009a. “The Body as a Text and the Text as the Body: A View from the Kālacakratantra’s Perspective.” In As Long as Space Endures: Essays on the Kālacakra Tantra in Honor of H. H. The Dalai Lama, edited by Edward A. Arnold, 179–92. Ithaca, NY: Snow Lion Publications. ———. 2009b. “Medicine and Astrology in the Healing Arts of the Kālacakratantra.” In As Long as Space Endures: Essays on the Kālacakra Tantra in Honor of H. H. The Dalai Lama, edited by Edward A. Arnold, 277–300. Ithaca, NY: Snow Lion Publications. ———. 2010. The Kālacakratantra: The Chapter on Sādhana together with the Vimalaprabhā. Tanjur Translation Initiative, Treasury of Buddhist Sciences Series. New York: American Institute of Buddhist Studies at Columbia University, co-published with Columbia University’s Center for Buddhist Studies and Tibet House. ———. 2011. “A Brief Exploration of Late Indian Buddhist Exegeses of the ‘Mantrayāna and Mantranaya’.” Pacific World: Journal of the Institute of Buddhist Studies, Third Series, No. 13, Special Section: Recent Research on Esoteric Buddhism, Fall: 95–111. ———. 2013. “Practical Applications of the Perfection of Wisdom Sūtra and Madhyamaka in the Kālacakra Tantric Tradition.” In Companion to Buddhist Philosophy, edited by Steven Emmanuel, 164–79. Oxford: Blackwell Publishing. ———. 2017. “Linguistic Arguments and Exegesis in Indian Tantric Buddhism: Intention and Interpretation.” In Buddhism and Linguistics: Theory and Philosophy, edited by Manel Herat, 101–22. Basingstoke, UK: Palgrave MacMillan.

63

4 MAITREYA The Future Buddha as an Author Klaus-Dieter Mathes

In most Buddhist traditions, Maitreya is a bodhisattva who is prophesized to be the fifth, and coming, Buddha of our present auspicious eon (bhadrakalpa).1 He is held to currently reside as the future Buddha in Tuṣita heaven, from whence, like his predecessors, he will descend into the world after the teachings of the previous Buddha, in this case the Buddha Śākyamuni, are no longer practiced or known. In the Kashmir of the first centuries of our common era, Maitreya’s cult was at the center of the shift from mainstream to Mahāyāna Buddhism, when Yogācāra (Yoga Conduct) masters such as Saṅgharakṣa (second century CE) and Asaṅga (c. 315–390 CE) were believed to have sought and received inspiration from Maitreya (Demiéville 1954, 341). Followers of the Maitreya cult aspired to be reborn in Maitreya’s time and place. Maitreya, whose name means the “Embodiment of Love,” also came to be considered, together with Asaṅga and Vasubandhu, as the founder of Yogācāra, one of the two main branches of the Great Vehicle (Mahāyāna). His status as an author, however, is a bit controversial given his celestial nature. His works are generally said to have been first taught to Asaṅga, who then promulgated them in the world. The accounts of how Asaṅga received the teachings from Maitreya vary. According to Paramārtha’s biography of Vasubandhu (c. fourth or fifth century CE),2 Maitreya came down to earth from Tuṣita to teach Asaṅga the Stages of Yogic Practice (Yogācārabhūmi). Most sources specify, however, that Asaṅga was instructed by Maitreya in Tuṣita heaven (Demiéville 1954, 381–82). Since Tuṣita is considered to be part of the desire realm, it is within reach of human beings, even those who are still below the bodhisattva levels. During the transition from mainstream to Mahāyāna Buddhism, it was already considered possible for practitioners to become bodhisattvas and eventually full-fledged buddhas, but the Sarvāstivāda idea that there can be only one buddha at a time still remained widespread. Although the Mahāsaṅghikas and Sautrāntikas admitted the simultaneous existence of multiple buddhas, they nevertheless held that such buddhas must exist in different universes (Demiéville 1954, 371). Thus, bodhisattvas on the verge of buddhahood had to wait in Tuṣita (the realm of desire gods), engaging in celestial pleasure, study, and meditation, for their turn to descend on earth as the buddha of their respective era. In the meantime, however, travels to and from Tuṣita were imagined as occurring frequently, as, for example, when Śākyamuni, the buddha of our era, traveled there to give Abhidharma (metaphysics) teachings to his deceased mother, who had been reborn as a deity there. In as much as it was part of this world, human beings could also reach it without taking rebirth there, and in China it had been argued that 64

DOI: 10.4324/9781351030908-7

Maitreya

Tuṣita is easier to reach then the better-known “Blissful Paradise” (Sukhāvatī), for the latter is a pure realm accessible only for those on the bodhisattva levels.3 The question of Maitreya’s role as an author is thus vexed even for traditional scholars, since he is seen as the source of the teachings but not as their promulgator in the human realm. In his commentary on An Analysis of the Middle and Extremes (Madhyāntavibhāga),4 one of the texts Maitreya is said to have revealed to Asaṅga, Vasubandhu praises the author of the root text as “born from the Buddha himself ” (sugatātmaja),5 indicating an exalted status for the author, while Sthiramati (sixth century CE) in his subcommentary specifies Maitreya as the author of the root text.6 Sthiramati describes Maitreya as follows: Being only one birth away from buddhahood, he has reached the highest perfection in terms of all bodhisattvas’ supernatural knowledge, memory, analytical knowledge, meditative absorption, power, patience, and liberation, his hindrances on all bodhisattva levels having been completely left behind.7 As Demiéville (1954, 386) notes, Asaṅga is said to have been on the path of preparation, an initial stage of the Buddhist path, when he first received teachings from the bodhisattva Maitreya. Having received the transmission of An Analysis of the Middle and Extremes, he then proceeded to the first bodhisattva level.

The Five Maitreya Works Tibetan tradition maintains that Asaṅga traveled to Tuṣita, where he received from the future Buddha a set of texts that came to be known as the Five Maitreya Works (Mathes 1996, 11–17). In partial support of this legend, namely that Asaṅga is not the origin of these works, Frauwallner (1994, 296) observes that the doctrines expounded in the Maitreya Works show a clearly pronounced originality and distinguish themselves noticeably from Asaṅga’s own works. The Five Maitreya Works include a summary and analysis of the Discourses on the Perfection of Insight (Prajñāpāramitāsūtras), an analysis of “buddha nature” (tathāgatagarbha), and teachings on the Great Vehicle in general. Back in the human world, Asaṅga is said to have written down these teachings. On this account, Maitreya and Asaṅga, or rather Maitreya through Asaṅga, can be credited with the following five works, given in the order in which they are traditionally said to have been taught to Asaṅga. As we shall see, these five texts may indeed all be by a single hand: The Ornament of Clear Realization (Abhisamayālaṃkāra) The Ornament of Great Vehicle Discourses (Mahāyānasūtrālaṃkāra) An Analysis of the Precious Potential (Ratnagotravibhāga) An Analysis of the Middle and Extreme (Madhyāntavibhāga) An Analysis of Phenomena and their True Nature (Dharmadharmatāvibhāga) The first text, The Ornament of Clear Realization,8 summarizes the Discourses on the Perfection of Insight in a way compatible with Yogācāra and Tathāgatagarbha9 thought. This is followed by The Ornament of Great Vehicle Discourses,10 which groups mainly Yogācāra topics into twenty-one chapters. The remaining three texts are so-called analyses (vibhāgas); namely, An Analysis of the Precious Potential,11 An Analysis of the Middle and Extreme, and An Analysis of Phenomena and their True Nature.12 The first analysis is a standard Indian treatise on buddha nature and buddhahood, the second defines a “Yogācāra middle way” based on 65

Klaus-Dieter Mathes

the Discourses on the Perfection of Insight, and the last is a Yogācāra work that focuses on the fundamental transformation of the ordinary states of mind, and the phenomena (dharmas) they display, into the true nature of these phenomena (dharmatā), which amounts to buddhahood. Some Tibetan traditions place An Analysis of the Precious Potential at the end of the list.13 To present the two ornaments and three analyses as Maitreya’s revelation invests the Yogācāra summary and interpretation of Great Vehicle Discourses an extremely high level of authority. This must be seen in the context of newly forming Yogācāra school’s competition with the rival Madhyamaka (Centrist Way) interpretation of the Great Vehicle.14 The Chinese tradition knows neither The Ornament of Clear Realization nor An Analysis of Phenomena and their True Nature (Seyfort Ruegg 1969, 43), and attributes An Analysis of the Precious Potential to Sāramati (Frauwallner 1994, 255). Instead of these three texts, the Chinese corpus of the Five Maitreya Works includes the Stages of Yogic Practice (the same one attributed to Asaṅga via Maitreya and mentioned earlier); An Analysis of Yoga (Yogavibhāga), which is now lost; and A Commentary on the Diamond Cutter (Vajracchedikāvyākhyā) (Nakamura 1989, 256). The Stages of Yogic Practice differs in general tenor from the Five Maitreya Works of the Tibetan tradition, as it contains numerous mainstream Buddhism strands and is dominated by detailed Abhidharma-based descriptions of psycho-physical aggregates and mental factors of existence. The Tibetan Maitreya Works are more centered on a positively described ultimate “sphere of qualities” (dharmadhātu), natural luminosity, or buddha nature (Frauwallner 1994, 33; Mathes 1996, 14–16). They thus seem more likely to be by a single author than does the list of five works as given in the Chinese tradition. The remainder of this chapter provisionally considers “Maitreya” as the author of the Five Maitreya Works according to the Tibetan tradition. Our work will be to see whether these five texts might reasonably be attributed to a single hand.

The Ornament of Clear Realization In his commentary on the first among the Five Maitreya Works, The Ornament of Clear Realization, the eighth-century Indian author Haribhadra writes that while Asaṅga had understood the meaning of the Buddha’s words and had gained realization, he still was not able to determine the meaning of the Discourses on the Perfection of Insight due to their profundity and numerous repetitions. When he saw that this depressed Asaṅga, Maitreya expounded on these Discourses on the Perfection of Insight, composing the verses of The Ornament of Clear Realization so that Asaṅga might more easily understand the essence of the scriptures (Brunnhölzl 2010, 47). In his initial description of the foundation of the path and practice, Maitreya makes it clear that an all-pervading, indivisible sphere of qualities does not allow for a distinction into definite, unchangeable potentials or dispositions (gotra), such as to restrict a particular practitioner to developing on one particular vehicle ( yāna) only. In The Ornament of Clear Realization, we thus find: Given the indivisibility of the sphere of qualities, A distinction between potentials is not acceptable.15 Needless to say, this excludes the possibility of an absolute cut-off potential (agotra, lit. “those without a family”). Such a harsh category, which is occasionally found in Yogācāra, contradicts the central Great Vehicle aspiration to lead all sentient beings to the ultimate level of buddhahood. The softening of any rigid typology of potentials in evidence here reflects a thread running throughout the Five Maitreya Works in which we find a heavy emphasis on the inherent capacity of all beings to attain buddhahood. Maitreya’s teaching thus is best described 66

Maitreya

as a synthesis of the Yogācāra and Tathāgatagarbha models of reality, a synthesis that seeks to avoid the potential flaws of both systems. Even though several passages in The Ornament of Clear Realization can be interpreted along the lines of Madhyamaka16 – and numerous Tibetan masters such as Rin chen bzang po (958–1055) have taken the text as a commentary on the second Dharma Wheel in keeping with the Madhyamaka School of Autonomous Reasoning (Svātantrika-Madhyamaka, Seyfort Ruegg 1969, 65) – the text’s proximity to Yogācāra concepts cannot be denied (Makransky 1997). Emptiness is thus thought of in terms of the absence of a perceived object and perceiving subject rather than in terms of the absence of an own nature (svabhāva). This becomes clear in the following two verses in The Ornament of Clear Realization, which distinguish how ordinary beings and the noble ones apprehend sentient beings as substantially and nominally existent, respectively: The two concepts about substantially and nominally existent sentient beings Are the two perceiving subjects. Divided into ordinary beings and noble ones, Each of them is ninefold in terms of their perceived objects. If the two types of perceived objects do not exist the way they appear, Of what are these two types of subjects then perceivers? The emptiness of being a perceiving subject Is thus the defining characteristic of both of them.17 In the same vein, Brunnhölzl (2010, 84–85) observes that the relinquishment of the perceived and perceiver, which results in a wisdom beyond these two, is as classical a Yogācāra paradigm as can be imagined. Moreover, the list of twenty-two examples of enlightened mind (bodhicitta) in The Ornament of Clear Realization (I.19–20) is not found in the Discourses on the Perfection of Insight but is almost identical with a list found in The Ornament of Great Vehicle Discourses (IV.15–20). Likewise, the description of potential in The Ornament of Clear Realization (I.38– 39) is not sheer emptiness, but rather the sphere of qualities serving as a basis for various accomplishments on the path. The Yogācāra and Tathāgatagarbha leanings of this text are thus evident. When it comes to the bodies of the buddha, we find the presentation of the “body of qualities” (dharmakāya) in terms of the “embodiment of true nature” (svābhāvikakāya), “the body of enjoyment” (sāmbhogikakāya), and the “emanation body” (nairmāṇikakāya). This presentation matches similar passages in The Ornament of Great Vehicle Discourses and An Analysis of the Precious Potential. According to The Ornament of Clear Realization, the embodiment of true nature, which is itself the true nature of the body of qualities, thus contains all buddha qualities: The Sage’s embodiment of true nature Has the natural defining characteristic of all uncontaminated qualities, Which have attained to purity In every respect.18 In other words, the embodiment of true nature is not conceptualized as the mere emptiness of the body of qualities, as we find in the four-body system of Haribhadra. In that four-body system, the embodiment of true nature is an independent body and not simply another way to talk about the body of qualities (Mathes 2008, 168; Makransky 1997, 289ff.). But here, the presentation of the buddha bodies recalls the Tathāgatagarbha theory of An Analysis of the Precious Potential, in which the buddha qualities are inseparably connected with a buddha element that is equated with the body of qualities and described in terms of the embodiment of true nature 67

Klaus-Dieter Mathes

(Mathes 2008, 8–9). It should also be noted that in The Ornament of Clear Realization, buddhahood is described as all-pervading and permanent, as well: Given its vast activity, Buddhahood is termed “all-pervading”; And being exempt from decay, It is called permanent.19 This brings to mind the initial verse of praise of An Analysis of the Precious Potential, where buddhahood is described as being discovered, in the same sense as the historical Buddha with the given name Siddhārtha opened up to, or found, buddhahood: Homage to you who opened up to buddhahood, which is without beginning, middle, or end, and is peaceful, And who, after awakening, taught the fearless, eternal path for the sake of the awakening of those who are not yet awake; To you, who holds the supreme sword and thunderbolt of wisdom and love, who cuts the tumor of suffering into pieces, And destroys the rampart of doubts enclosed by a thicket of various views.20 According to a subsequent commentary verse, being “without beginning, middle, or end” means that buddhahood is unconditioned.21

The Ornament of Great Vehicle Discourses The next work attributed to Maitreya is The Ornament of Great Vehicle Discourses. Even though accepted as a Maitreya work by both the Tibetan and Chinese traditions, the colophon of the only Sanskrit manuscript mentions the unidentified great bodhisattva Vyavadātasamaya (Lévi 1911, 7), adding further mystery to the question of authorship. The Ornament of Great Vehicle Discourses treats various traditional Yogācāra topics and notably includes a theory that accepts the notion of a cut-off potential. Within his description of different types of cut-off potentials, Maitreya distinguishes a group that has no “cause” or chance of liberation from four groups of sentient beings, which are only temporarily cut off. Some are solely bent on bad conduct. Then there are those whose positive qualities have been destroyed, Or those who lack the virtue conducive to liberation. And some have few positive qualities. But there are also those for whom there is no cause at all.22 Vasubandhu explains: What is meant here regarding those who are without the capacity to attain perfect nirvāṇa is the cut-off potential. In short, there are two types, those who are cut off temporarily and those who are cut off completely. Of those who are cut off temporarily, there are four types: those who are solely bent on bad conduct, those whose roots of virtue are cut off, those who are without the roots conducive to liberation, 68

Maitreya

and those who have few roots of virtue. They are those whose accumulations are incomplete. Those who are entirely (atyanta-) without the capacity to attain perfect nirvāṇa – without any cause, so to speak – lack any potential to attain perfect nirvāṇa at all (eva).23 An absolute (atyanta-) cut-off potential, however, would contradict the statement in The Ornament of Great Vehicle Discourses that all sentient beings have buddha nature: Even though suchness is undifferentiated in all sentient beings, In its purified form it is the state of the Tathāgata. Therefore all living beings Have the seed/nature (garbha) of him (i.e., the Tathāgata).24 In other words, we have here a strict potential-based system with the explicitly permanent exclusion of one group of sentient beings from liberation over against an element from An Analysis of the Precious Potential – buddha nature being defined there as well as in The Ornament of Great Vehicle Discourses as “suchness accompanied by stains” (samalā tathatā). However, if that which constitutes a specific potential – i.e., the dependent nature ( paratantrasvabhāva) within the Yogācāra three-natures model of reality – does not exist ultimately, the ultimate goodness of an all-pervading sphere of qualities could still inhere as the ultimate nature of all sentient beings, even those with a cut-off potential (D’Amato 2003, 126ff.). The gotra issue thus is related to the problem the traditional three-nature model has when the dependent nature is taken to have substantial, ultimate existence. The problem with this model is that it posits that something dependently arising really exists, a position which has been rightfully refuted in Madhyamaka philosophy. The three-nature theory lies at the center of the Yogācāra model of reality (even though Maitreya himself does not make much use of it25) and reflects the old Abhidharma distinction of relative and ultimate truths/realities in terms of nominal ( prajñaptisat) and substantial (dravyasat) existence. In a modified form – there being no longer material, but only “mental” substance in fully developed Yogācāra – a dependently arising stream of false imagining (abhūtaparikalpa), i.e., the substantially existing dependent nature, projects the imagined nature ( parikalpitasvabhāva) of a perceived object and perceiving subject, which exist only nominally. Sthiramati follows along these lines in relating the imagined and dependent natures to the relative and ultimate truths/realities (Salvini 2015, 44–50). The third – the perfect – nature ( pariniṣpannasvabhāva) is the dependent empty of the imagined. In the Maitreya works, this model is made compatible with Tathāgatagarbha thought by reducing the substantial existence of the dependent nature to the level of relative truth. In his commentary on the first verse of the True Reality (tattva) Chapter in The Ornament of Great Vehicle Discourses, Vasubandhu excludes both existence and nonexistence from true reality, and in this context takes the ultimate to exist as neither the imagined nor the dependent. The first line of the first verse in the True Reality Chapter is as follows: Neither existent nor nonexistent . . .26 Vasubandhu comments: The ultimate has the meaning of nonduality. It is taught in five points. Neither existent – in terms of what is called the imagined and dependent – nor nonexistent – in terms of what is called the perfect characteristic.27 69

Klaus-Dieter Mathes

In other words, the exclusion of what exists nominally and substantially – i.e., the imagined and dependent natures – from the ultimate and the equation of the latter with buddha nature work together to smooth out the differences between The Ornament of Great Vehicle Discourses and An Analysis of the Precious Potential. The problem with the traditional potential (gotra) model and cut-off potential could be interpreted, as already indicated, along similar lines. One possible solution would thus be to regard it as a teaching of provisional meaning (neyārtha). But the problem with that approach is that in The Ornament of Great Vehicle Discourses, it is stated that the single-vehicle (ekayāna) doctrine – which is a crucial element of the Tathāgatagarbha teachings – was taught with a purpose, which implies that it has provisional meaning: The perfect Buddhas have taught The unity of the vehicle (ekayānatā) For the sake of those who are unsettled, To attract some and to hold others.28 In his commentary, Vasubandhu introduces this verse with the following remark: Buddhahood is the single-vehicle. Thus the unity of the vehicle must be understood with this or that intent in this or that discourse. But it is not that the three vehicles do not exist. Why, again, did the buddhas teach the unity of the vehicle with this or that intent?29 In his subcommentary, Sthiramati concludes: As for the explanation of “single-vehicle” here, it must be taken to have provisional meaning,30 because he (i.e., the Buddha) taught it with an intent, namely for the sake of sentient beings. The teaching of three vehicles has definitive meaning.31 It is not necessary to accept Sthiramati’s conclusion, though, for if one goes by the Yogācāra hermeneutics of Vasubandhu’s Reasoning of Interpretation (Vyākhyāyukti), every discourse has an aim or intent, even a discourse of definitive meaning. In other words, the status of provisional meaning cannot be concluded from the fact that an intent is mentioned (see Mathes 2008, 14–15).

An Analysis of the Precious Potential In An Analysis of the Precious Potential, the next of the Five Maitreya Works,32 Takasaki (1966) and Schmithausen (1971) identified older textual layers, which reach back to the beginning of the fourth century CE, the time when the doctrinally close ninth chapter of The Ornament of Great Vehicle Discourses was composed. This is nothing unusual, though, as a teacher typically draws on earlier memorized passages. The final versions of both texts display a remarkable synthesis of Yogācāra and Tathāgatagarbha thought, which is indicative of a single author. In An Analysis of the Precious Potential, the cut-off potential (and thus the strict potential distinctions) is said to be taught with a hidden intent (and therefore to be of provisional meaning), since in fact everybody has the naturally pure potential. In his commentary on An Analysis of the Precious Potential, Asaṅga says: Again, the saying: “Those with great desire (icchantikas) do not have at all the capacity to attain perfect nirvāṇa” is taught with the hidden intent of another time to 70

Maitreya

remove hatred towards the Great Vehicle doctrine, this hatred being the reason why they themselves are those with great desire. Indeed, since everybody has the naturally pure potential, it cannot be that there should be anybody whose nature is the exact opposite of purity.33 This, in combination with the restriction of the existence of false imagining (i.e., the dependent nature) to the level of relative truth, allows for the notion that all sentient beings possess buddha nature, the dependent nature being included within the adventitious stains that cover that nature. The latter is then, as we have seen already in this chapter, identified with mind’s luminous nature or suchness (i.e., the perfect nature). The main thesis of An Analysis of the Precious Potential is that all sentient beings possess buddha nature, which in turn is inseparably endowed with innumerable buddha qualities. This comes out most clearly in the following verses: There is nothing to be removed from it And nothing to be added. The real should be seen as real, And seeing the real, one becomes liberated. The buddha-element is empty of adventitious stains, Which have the defining characteristic of being separable; But it is not empty of unsurpassable qualities, Which have the defining characteristic of not being separable.34 This position is mainly based on the nine similes of buddha nature in The Discourse on Buddha Nature (Tathāgatagarbhasūtra,35 Zimmermann 2002, 64) and other doctrinally related discourses such as The Discourse for Queen Śrīmālā (Śrīmālādevīsūtra)36 and The Chapter on the Absence of Increase and Decrease (Anūnatvāpūrṇatvanirdeśaparivarta).37 It is not clear, however, whether all buddha qualities are inseparably contained in one’s nature or only a subset of them, such as the largely space-like qualities of the embodiment of true nature (svābhāvikakāya) presented and discussed in An Analysis of the Precious Potential (II.29–47). In other words, the buddha nature of sentient beings would then not be identical with the body of qualities of a fully enlightened mind. That this was the intent of An Analysis of the Precious Potential is clear from its explanation of the fifth simile (a treasure buried under the floor of a poor man’s house) of The Discourse on Buddha Nature, where buddha nature is fully equated with the thirty-two qualities of the body of qualities (Zimmermann 2002, 121). An Analysis of the Precious Potential, which otherwise faithfully renders the nine similes of The Discourse on Buddha Nature, speaks only of the “treasure of qualities” (dharmanidhi). That this is not simply an unintentional inaccuracy is clear from An Analysis of the Precious Potential I.149–52, where the treasure illustrates the naturally present potential, from which the thirty-two qualities of the body of qualities emerge. In other words, the treasure of buddha nature no longer stands for these thirty-two qualities, but rather for their cause. It thus is very likely that Maitreya’s An Analysis of the Precious Potential is a Yogācāra interpretation of older strands of the text that had more closely followed the intent of the Discourses on Buddha Nature (Tathāgatagarbhasūtras).38 It should be noted, however, that already two similes from The Discourse on Buddha Nature – namely those of a tree grown from a seed and the future monarch (cakravartin) in a womb – indicate a growth of the buddha qualities in terms of strength (Mathes 2008, 8–11). Whether buddha nature is fully equated with the body of qualities or not, its analysis in terms of three natures39 results in a luminous perfect nature that is empty of adventitious stains. 71

Klaus-Dieter Mathes

These stains include the imagined nature and dependent nature of false imagining (Mathes 2012, 194). This would not be, then, the original three-nature model of An Analysis of the Middle and Extreme, the next in line of our Five Maitreya Works, but that of the Great Commentary (Bṛhaṭṭīkā), a large Perfection of Insight (Prajñāpāramitā) commentary usually ascribed to Vasubandhu and occasionally to Daṃṣṭrasena (Brunnhölzl 2011b, 9–14). If Vasubandhu truly was the author, the Great Commentary would lend further support to limit the validity of the three-nature model of An Analysis of the Middle and Extreme and its ontological distinction between the imagined and dependent natures to the level of relative truth.

An Analysis of the Middle and Extreme The presentation of false imagining (i.e., the dependent nature) in the beginning of the first chapter of the An Analysis of the Middle and Extreme is structured around the original Yogācāra model with its dominant Abhidharma distinction between nominal and substantial existence. As already mentioned, this translates into a merely nominal existence of the imagined and a substantially existing dependent. Salvini (2015, 44–50) shows that substantial existence is equivalent, at least for Stir­ amati, to the Abhidharma ultimate. It has been argued that the first part of this chapter also negates the real existence of nondual mind because in the formulas defining the fourfold Yogācāra practice, which leads to the realization of a state free from perceived and perceiver, “mind-only” (cittamātra) and “cognition-only” (vijñaptimātra) are also left behind. False imagining (i.e., “mind-only” as the dependent) is said to exist, however, and in the formulas describing the fourfold practice, vijñaptimātra is not the technical term for the Yogācāra tenet of everything existing as cognitiononly but rather simply expresses the logical impossibility of cognition without any object. This is clear from Vasubandhu’s commentary on An Analysis of the Middle and Extreme I.6cd: Based on the non-recognition of a perceived object, the non-recognition of mere cognition (vijñaptimātra) arises.40 Still, the dependent is said to require abandoning in An Analysis of the Middle and Extreme III.9c,41 and so does not qualify as the ultimate. Line III.10d of the same text goes on to say that only the perfect nature is accepted as ultimate truth,42 while false imagining (i.e., the dependent) needs to be abandoned. The obvious solution to the seeming contradiction between an ultimately and not ultimately existing dependent is that in the description of false imagining in the first part of the first chapter, the qualifier “ultimately” only refers to an ultimate existence in Abhidharma terms (where momentary factors of existence constitute the ultimate), while the ultimate in the third chapter of An Analysis of the Middle and Extreme, as in the sixth chapter of the Ornament of Great Vehicle Discourses, must be taken as the ultimate in the Tathāgatagarbha system, i.e., buddha nature itself. What I propose here, in other words, is that Maitreya attempted a synthesis of Yogācāra and Tathāgatagarbha thought, a synthesis that can be also found in An Analysis of the Precious Potential and An Analysis of Phenomena and their True Nature. This requires a shift from the original Yogācāra model of an ultimate dependent to a Tathāgatagarbha ultimate beyond the imagined and dependent natures. Understood in this way, the initial verse of An Analysis of the Middle and Extreme would then present the thesis of such a Yogācāra-Tathāgatagarbha synthesis: False imagining exists. Duality is not found in it. Emptiness is found there (i.e., in false imagining) And the latter is found in the former.43 72

Maitreya

The first two lines would thus only distinguish nominal and substantial existence on the level of a Tathāgatagarbha-based relative truth, while the third and fourth lines would be an explanation of a Tathāgatagarbha-based understanding of the relative and ultimate truths. In other words, emptiness is not only the absence of duality, but is also understood as luminosity (see An Analysis of the Middle and Extremes I.16 and I.22). Luminous emptiness pervades the dependent nature’s bearers of properties (dharmin)44 as their true nature (dharmatā), while false imagining exists in emptiness as bearers of properties. This is as explained in Sthiramati’s commentary: The existence of emptiness in false imagining must be understood in terms of the latter’s true nature. False imagining, in turn, exists in emptiness in the form of the bearers of properties.45

An Analysis of Phenomena and their True Nature The same relation between phenomena and their true nature is also at the center of the An Analysis of Phenomena and their True Nature and its commentary by Vasubandhu.46 Their definitions are as follows: As to the defining characteristics of phenomena, they are dualistic appearances and that which appears in accord with expressions; all of them are false imagining.47 As to the defining characteristic of the true nature of phenomena, it is suchness, in which there is no differentiation between a perceived object and the perceiving subject, an expressed object and the expression.48 The relation between the two is explained as follows: The two (i.e., phenomena and their true nature) are neither identical nor separate, because there is, and also is not, a difference between the existent true nature and nonexistent phenomena.49 Even though the technical terms neither for relative and ultimate truths/realities nor for the three natures are found here (and are absent in the commentary, as well), one is left with having to include false imagining and the duality it manifests within relative truth, false imagining being the defining characteristic of non-existing phenomena. Still, false imagining is not completely negated, since it exists as mere delusion, i.e., that which generates dualistic appearances (see Mathes 1996, 255). Just as in An Analysis of the Middle and Extreme, it must be eventually abandoned: Even though false imagining always appears to consciousness, its phenomenal content is simply nonexistent. Moreover, “when it appears, the true nature does not appear” means that suchness does not appear. When [false imagining] does not appear, the true nature appears.50 This happens upon the completion of one’s fundamental transformation (āśrayaparivṛtti): One comprehends the nature of fundamental transformation known as the stainlessness of suchness so that adventitious stains do not appear anymore, and only suchness appears.51 73

Klaus-Dieter Mathes

To be sure, the term buddha nature (tathāgatagarbha) is not found in the whole of An Analysis of Phenomena and their True Nature or its commentary. Still, the proximity of this Yogācāra text to An Analysis of the Precious Potential is evident, as can be gathered from Vasubandhu’s final summary, where a permanent true nature (dharmatā) and natural luminosity are disclosed in just the same way as is the purity of gold, water, and space: Likewise, in the case of fundamental transformation, it is not that natural luminosity did not exist before. It only did not appear due to the manifestation of adventitious hindrances. . . .52 Since that change does not exist, the true nature and the fundamental transformation (āśrayaparivṛtti), which is characterized by it, are permanent. Here, with the examples of gold and water, only the quality of these objects of comparison, not their substance, was taught as being analogous to the transformation. With the example of space, the transformation was taught completely.53 An Analysis of Phenomena and their True Nature and its commentary display the same synthesis of Yogācāra and Tathāgatagarbha thought as in the first Maitreya works. This synthesis can be identified as a continuous thread that remedies the flaws Yogācāra has in the eyes of the followers of the Madhyamaka, namely that a considerable group of sentient beings is completely cut off from liberation or that a dependently arising mind exists on the level of ultimate truth. The influence of Yogācāra on the original Tathāgatagarbha doctrine, on the other hand, prevents a too-substantialist notion of buddha nature. Whether revealed by a celestial bodhisattva (as tradition would have it) or taught by a Maitreyainspired Asaṅga (as David Seyfort Ruegg maintains) or by Maitreyanātha and Sāramati, humans of flesh and bone (as argued by Erich Frauwallner), there is enough ground for the hypothesis that the two ornaments and the three analyses, the Five Maitreya Works of the Tibetan tradition, are from one hand. It is difficult to see how Asaṅga could have played an active role in their composition, as his own works are largely influenced by the concepts of Śrāvakayāna scholasticism. Instead of Maitreya’s main topic of an inconceivable ultimate empty of duality, Asaṅga prefers to chart a complex psychological map of mental factors of existence; different types of consciousness, especially the ground consciousness (ālayavijñāna); and mind’s associated factors. Following the Chinese tradition, though, Frauwallner argues that the Analysis of the Precious Potential is by Sāramati while the remaining two analyses and the two ornaments are by Maitreyanātha. David Seyfort Ruegg replies that the Ornament of Clear Realization and the group of three Yogācāra works (i.e., The Ornament of Great Vehicle Discourses, The Analysis of the Middle and the Extreme, and An Analysis of Phenomena and Their True Nature) differ to a greater extent than the latter three from the An Analysis of the Precious Potential and that if Frauwallner tolerates the difference between the three Yogācāra works and The Ornament of Clear Realization, he should also accept the differences with The Analysis of the Precious Potential. In line with Seyfort Ruegg, I argue that the Yogācāra influence on the latter and the Precious Potential’s Tathāgatagarbha influence on the three Yogācāra works can be best explained by assuming the same author with the agenda of a systematic interpretation of Great Vehicle Buddhism.

Notes 1 That is, following the influential Discourse on the Auspicious Eon (Bhadrakalpikasūtra). See Tournier (2019, 101). 2 For the dates of Vasubandhu, see Mathes (2008, 45).

74

Maitreya 3 See Vibhāṣā T. 1545, xxxix, 152c (quoted after Demiéville 1954, 376, fn. 4). 4 For a translation of this text, see D’Amato (2012). 5 MAVBh 173: “Having venerated the author of this treatise, who was born from the Sugata himself,” et passim: The translation of all quoted material is my own. 6 MAVṬ 24: “ Maitreya is the author of this treatise in verse form.” 7 MAVṬ 24–7. 8 For an English translation, see Brunnhölzl (2011a, 475–505). 9 In this contribution, the term Tathāgatagarbha indicates the stream of thought, whereas the term “buddha nature” indicates the idea that is at the heart of that stream of thought. 10 For an English translation, see The Dharmacakra Translation Committee (2014). 11 For an English translation, see Brunnhölzl (2014, 331–460). 12 For an English translation, see Brunnhölzl (2012, 157–71). 13 For the sequence of the Five Maitreya Works, see Mathes (1996, 16–17). 14 It should be noted that the related Yogācāra hermeneutics of Vasubandhu’s “Reasoning of Interpretation” (Vyākhyāyukti) is at times also taken to stem from Maitreya (see Mathes 2008, 427 fn. 73). 15 AA I.40ab (AA 88). 16 See, for example, AA I.31ab (AA 79): “The fact that matter and the rest lack an own nature refers to the state of having nonexistence as their nature.” 17 AA V.6–7 (AA 311–4). 18 AA VIII.1 (AA 383–4). 19 AA VIII.11 (AA 397–8). 20 RGVV 79–12. 21 RGV I.6ab (RGVV 81). 22 MSA III.11 (MSABh 1221–24). 23 MSABh 1225–132. 24 MSA IX.37 (MSABh 4013–14). 25 Maitreya invokes the three-nature model of reality only once, in An Analysis of the Middle and Extremes I.5, where he equates the imagined ( parikalpita) with the cognitive object (artha), the dependent ( paratantra) with false imagining (abhūtaparikalpa), and the perfect ( pariniṣpanna) with the nonexistence of duality (dvayābhāva). Generally, though, Maitreya prefers to work with the categories duality, false imagining, and emptiness. See Mathes (2000, 200ff). 26 MSA VI.1 (MSABh 2212). 27 MSABh 2214–1. 28 MSA XI.54 (MSABh 693–4). 29 MSABh 6827– 692. 30 Tib. bkri ba’i don being an alternative translation to drang ba’i don, Skt. neyārtha. 31 MSAVBh 196a6–7. 32 For an English translation see Brunnhölzl (2014, 337–460). 33 RGVV 371–4. 34 RGV I.154–55 (RGVV 761–4). 35 For an English translation of the Tathāgatagarbhasūtra, see Zimmermann (2002, 93–161). 36 For an English translation, see Wayman and Wayman (1974). 37 For an English translation, see Silk (2015). 38 Tathāgatagarbhasūtras in the plural is a generic term for a group of discourses that teach buddha nature. The Tathāgatagarbhasūtra is one of them. 39 It should be noted that RGV does not make direct use of three-nature terminology. 40 MAVBh 203–4. 41 MAVBh 4021: “With regard to its comprehension and abandonment,” ( parijñāyāṃ prahāṇe ca). 42 MAV III.10d (MAVBh 4114): “But the ultimate is to be viewed in terms of only one.” Vasubandhu (MAVBh 4115–16) comments: “Ultimate truth should be understood in terms of the perfect nature alone.” Vasubandhu comments (MAVBh 412–3): “With regard to the comprehension and abandonment of the dependent” ( paratantrasya parijñāne prahāṇe ca |). 43 MAVBh 1716–17. 44 Lit., the bearers of properties. 45 MAVṬ 1517–18. 46 For an English translation, see Brunnhölzl (2012), 173–98.

75

Klaus-Dieter Mathes 4 7 DhDhVK 19–22. 48 DhDhVK 26–29. 49 DhDhVK 38–41. The additions in brackets are in accordance with Vasubandhu’s commentary (see Mathes 1996, 122). 50 DhDhVV 571–76. 51 DhDhVVS 12–13. 52 DhDhVV 701–702. 53 DhDhVV 706–708.

Bibliography Primary Sources AA DhDhVK DhDhVV DhDhVVS MAV MAVBh MAVṬ MSA MSABh MSAVBh RGV RGVV

Tripathi, Ramshankar, ed. 1993. Abhisamayālaṃkāra. Bibliotheca Indo-Tibetica Series 2. Sarnath: Central Institute of Higher Tibetan Studies. Dharmadharmatāvibhāgakārikā (Tibetan translation) Ed. by Klaus-Dieter Mathes. See Mathes 1996, 104–14. Dharmadharmatāvibhāgavṛtti (Tibetan translation) Ed. by Klaus-Dieter Mathes. See Mathes 1996, 69–98. Dharmadharmatāvibhāgavṛtti (Sanskrit fragment) Ed. by Klaus-Dieter Mathes. See Mathes 1996, 99–103. Madhyāntavibhāga See MAVBh Nagao, Gadjin M., ed. 1964. Madhyāntavibhāgabhāṣya. Tokyo: Suzuki Research Foundation. Yamaguchi, Susumu, ed. 1934. Madhyāntavibhāgaṭīkā. Nagoya: Librairie Hajinkaku. Lévi, Sylvain, ed. 1907. Mahāyānasūtrālaṃkāra. Bibliothèque de l’École des Hautes Études. Sciences historiques et philologiques 159. Paris: Librairie Honoré Champion. Mahāyānasūtrālaṃkārabhāṣya See MSA Mahāyānasūtrālaṃkāravṛttibhāṣya (Tibetan translation) Derge bsTan ’gyur 4034, sems tsam, vol. mi, 1b1–283a7 Ratnagotravibhāga See RGVV Johnston, Edward H. 1950. Ratnagotravibhāgavyākhyā. Patna: The Bihar Research Society, 1950. [The manuscripts A and B on which Johnston’s edition is based are described in Johnston 1950, vi–vii. See also Bandurski et al. 1994, 12–13.]

References Bandurski, Frank, Bhikkhu Pāsādika, Michael Schmidt, and Bangwei Wang. 1994. Untersuchungen zur buddhistischen Literatur. Sanskrit-Wörterbuch der buddhistischen Texte aus den Turfan-Funden. Beiheft 5. Göttingen: Vandenhoeck & Ruprecht in Göttingen. Brunnhölzl, Karl. 2010. Gone Beyond, vol. 1. Ithaca, NY: Snow Lion Publications. ———. 2011a. Gone Beyond, vol. 2. Ithaca, NY: Snow Lion Publications. ———. 2011b. Prajñāpāramitā, Indian “gzhan stong pas”, and the Beginning of Tibetan gzhan stong. Wiener Studien zur Tibetologie und Buddhismuskunde 74. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien. ———. 2012. Mining for Wisdom Within Delusion. Boston and London: Snow Lion Publications. ———. 2014. When the Clouds Part. Boston and London: Snow Lion Publications.

76

Maitreya D’Amato, Mario. 2003. “Can All Beings Potentially Attain Awakening? Gotra-Theory in the Mahāyānasūtrālaṃkāra.” Journal of the International Association of Buddhist Studies 26 (1): 115–38. ———. 2012. Maitreya’s Distinguishing the Middle from the Extremes (Madhyāntavibhāga). Along with Vasubandhu’s Commentary (Madhyāntavibhāgabhāṣya). A  Study and Translation. New York: The American Institute of Buddhist Studies. Demiéville, Paul. 1954. “La Yogācārabhūmi de Saṅgharakṣa.” Bulletin de l’École Française d’Extrême Orient 44 (2): 339–436. The Dharmacakra Translation Committee. 2014. Ornament of the Great Vehicle Sutras: Maitreya’s Mahāyānasūtralaṃkāra with Commentaries by Khenpo Shenga and Ju Mipham. Maitreya Texts 3. Boston and London: Snow Lion Publications. Frauwallner, Erich. 1994. Die Philosophie des Buddhismus. Berlin: Akademie Verlag. Johnston, Edward H. 1950. The Ratnagotravibhāga Mahāyānottaratantraśāstra. Patna: The Bihar Research Society. Lévi, Sylvain. 1911. Mahāyānasūtrālaṃkāra, vol. 2: Traduction-Introduction-Index. Bibliothèque de l’École des Hautes Études. Sciences historiques et philologiques 190. Paris: Librairie Honoré Champion. Makransky, John J. 1997. Buddhahood Embodied: Sources of Controversy in India and Tibet. SUNY Series in Buddhist Studies. Albany, NY: State University of New York Press. Mathes, Klaus-Dieter. 1996. Unterscheidung der Gegebenheiten von ihrem wahren Wesen (Dharmadharmatāvibhāga). Indica et Tibetica 26. Swisttal-Odendorf: Indica et Tibetica Verlag. ———. 2000. “Tāranātha’s Presentation of trisvabhāva in the gŹan stoṅ sñiṅ po.” Journal of the International Association of Buddhist Studies 23 (2): 195–223. ———. 2008. A Direct Path to the Buddha Within: Go Lotsāwa’s Mahāmudrā Interpretation of the Ratnagotravibhāga. Studies in Indian and Tibetan Buddhism. Boston: Wisdom Publications. ———. 2012. “The Gzhan Stong Model of Reality – Some More Material on Its Origin, Transmission, and Interpretation.” Journal of the International Association of Buddhist Studies 34 (1–2): 187–226. Nakamura, Hajime. 1989. Indian Buddhism: A Survey with Bibliographical Notes. New Delhi: Motilal Banarsidass. Salvini, Mattia. 2015. “Language and Existence in Madhyamaka and Yogācāra: Preliminary Reflections.” In Madhyamaka and Yogācāra: Allies or Rivals? edited by Jay L. Garfield and Jan Westerhof, 29–71. Oxford: Oxford University Press. Schmithausen, Lambert. 1971. “Philologische Bemerkungen zum Ratnagotravibhāga.” Wiener Zeitschrift für die Kunde Südasiens 15: 123–77. Seyfort Ruegg, David. 1969. La Théorie du Tathāgatagarbha et du Gotra: Études sur la Sotériologie et la Gnoséologie du Bouddhism. Paris: École française d’Extrême Orient. Silk, Jonathan A. 2015. Buddhist Cosmic Unity: An Edition, Translation and Study of the Anūnatvāpūrṇatvanirdeśaparivarta. Hamburg Buddhist Studies 4. Hamburg: Hamburg University Press. Takasaki, Jikido. 1966. A Study on the Ratnagotravibhāga (Uttaratantra) Being a Treatise on the Tathāgatagarbha Theory of Mahāyāna Buddhism. Rome Oriental Series 33. Rome: Istituto Italiano per il Medio ed Estremo Oriente. Tournier, Vincent. 2019. “Buddhas of the Past: South Asia.” In Brill’s Encyclopedia of Buddhism. Vol. 2: Lives, edited by Jonathan A. Silk, 95–108. Leiden and Boston: Brill. Wayman, Alex, and Hideko Wayman. 1974. The Lion’s Roar of Queen Śrīmālā. New York: Columbia Press. Zimmermann, Michael. 2002. A Buddha Within: The Tathāgatagarbhasūtra. The Earliest Exposition of the Buddha-Nature Teaching in India. Bibliotheca Philologica et Philosophica Buddhica 6. Tokyo: The International Institute for Advanced Buddhology.

77

PART 2

Poet Philosophers

POET PHILOSOPHERS Introduction to Part 2

Texts taught in a typical, contemporary academic philosophy curriculum include a wide variety of genres. While philosophers today are primarily taught to write academic articles and monographs, philosophers have historically experimented with a much broader range of literary forms, including dialogues, aphorisms, personal essays, confessions, epistles, meditations, autobiographies, and commentaries, in addition to treatises. Some Western philosophers, perhaps most famously Lucretius and Boethius, expressed their philosophical thinking in poetic form. For many philosophers, their chosen literary form is not merely one among several possible ways of clothing their thinking, but is essential to their thought. Consider Plato’s dialogues, Montaigne’s essays, Kant’s treatises, or Nietzsche’s aphorisms: in each case, the philosophy is inherently connected to the literary form and would be different if embodied in another genre. A similar diversity of literary form appears in Indian philosophical traditions, which have also included treatises, commentaries, dialogues, narratives, and poems. Much Indian Buddhist philosophy was written according to models of Sanskrit verse. Nāgārjuna, Vasubandhu, and Śāntideva, for example, and many other prominent Indian Buddhist philosophers, wrote some of their most important works following Sanskrit poetic forms and integrated elements of dialogue. Thus, in some sense, many Indian Buddhist authors can be considered “poet philosophers.” The poet philosophers we consider in this part, however, are poetic in a stronger sense: their philosophy is not merely articulated in verse but embodied in song and narrative, and their poetic and literary forms do important philosophical work. In his chapter on Aśvaghoṣa, Vincent Eltschinger notes that this poet from the first century of the Common Era  – known for Buddhist dramatic works and refined court poetry (mahākāvya) – was the first Buddhist philosopher to articulate systematic and sophisticated critiques of other religious-philosophical systems. In Life of the Buddha (Buddhacarita), an epico-lyrical poem, Aśvaghoṣa defends the rationality of the Buddhist path – in contrast to other formulations of the path to liberation, especially Sāṃkhya and Vaiśeṣika – and therefore its effectiveness. In this account, the Buddha is a philosopher who justifies Buddhist doctrine and practice on the basis of reason, and reason is necessary for morality and liberation. Aśvaghoṣa’s views anticipate ideas that are more fully developed by later thinkers associated with Sautrāntika and Yogācāra. Moreover, Eltschinger argues, Aśvaghoṣa may be the first Buddhist philosopher we can identify who pursues a distinctive intellectual agenda along with an argumentative and literary style to achieve it. DOI: 10.4324/9781351030908-9

81

Poet Philosophers

The following chapter, by Sonam Kachru, is a literary and philosophical reading of the Questions of Milinda (Milindapañha), a text that presents a dialogue between the Buddhist monk Nāgasena and the Bactrian King Menander (“Milinda” in Pāli). Kachru is interested in reading this text as a work of literature that can help us understand what it might be like to think and feel as a Buddhist. He wants to explore how the Questions of Milinda might be encountered in ways that can change us as readers. He therefore focuses not only on the arguments, but also on aesthetic features such as the use of examples and the drama of the dialogue. Kachru reads the drama as an exploration of the role of thought in our lives, and the tragedy of living a life for which we can give no justification. Kachru also suggests that the text itself can teach us how to read well and how to reason wisely about our own lives. In the next chapter, Anne E. Monius addresses philosophy in the work of the Tamil author Cāttaṉār (ca. sixth–seventh century). Cāttaṉār is said to be the author of The Renunciation of Maṇimēkalai (Maṇimēkalai Tuṟavu), an extended poetic narrative that traces the path of Maṇimēkalai from a youthful courtesan to her ordination as a Buddhist nun. It is the first extant Tamil text of formal philosophy and is representative of the ways in which many Tamil texts integrated philosophical reasoning into poetic narratives. Three of the chapters are explicitly philosophical: one critiques non-Buddhist views; the final two present a Buddhist account of inferential reasoning and interdependent origination. Maṇimēkalai’s understanding of inferential reasoning and especially interdependent origination as presented by her teacher, Aṟavaṇaṉ, is what eventually leads her to commit to a life of renunciation. Like Maṇimēkalai, the reader can look back and see how everything in her life was interdependently originated and what kinds of conditions can lead to freedom and renunciation of worldly life. As Monius argues, the beauty and rhetorical power of the narrative help the reader cultivate both an understanding and an emotional response intended to turn us toward renunciation and freedom. In the final chapter of this part, Roger R. Jackson presents the late–first millennium Indian tantric poet, Saraha, as a philosopher. Saraha is informed by the metaphysical and ontological thought of Madhyamaka, Yogācāra, and Vajrayāna. Still, as Jackson argues, Saraha is a kind of philosophical anti-philosopher, as he is critical of philosophical study as a method to achieve awakening ( just as he is also critical of meditation, pilgrimage, and tantric ritual and practice – all of which different Buddhists have held to be necessary for liberation). It is only when we let go of our attachment to conventional forms of practice, Saraha insists, that we can open ourselves to the ultimate reality beyond language and thought. In his songs, while arguing that the ultimate is beyond the grasp of words, he still seeks to somehow gesture toward it. This gesturing is accomplished with many terms, including: the innate, stainless mind, inmost nature, the real, the great seal, nondual mind, emptiness, awakening, Buddha. According to Saraha’s nondual philosophy, everything in our lives  – even what we take to be polluting and polluted – is ultimately somehow part of the radiant, pure, primordial reality. Instead of employing sustained rational arguments, Saraha seeks a direct understanding of reality and poetically presents the human potential for awakening.



82

5 AŚVAGHOS․A

The Dawn of Indian Buddhist Philosophy Vincent Eltschinger

Aśvaghoṣa (100 CE?) is widely acknowledged as one of the earliest Indian and Buddhist representatives of ornate court poetry (kāvya).1 But the celebrated poet can with equal right be regarded as the first Buddhist philosopher – the first, at least, with a clearly identifiable literary personality, argumentative style, and intellectual agenda. To be sure, the Buddhist canon on which Aśvaghoṣa so heavily relied contained many interesting arguments against non-Buddhist – mostly Brahmanical – ideas and practices such as the caste-classes and ritual violence. But Aśvaghoṣa is the first known Indian Buddhist intellectual to have developed systematic and often quite sophisticated arguments against concurrent salvation systems, predominantly Sāṃkhya, likely Buddhism’s most dangerous rival in his time and milieu. Whereas the poet provides an eloquent outline of the Buddhist path to salvation in his lyrical poem Handsome Nanda (Saundarananda, SNa) – which, like his Drama of Śāriputra (Śāriputraprakaraṇa), a play dedicated to the conversion of the Buddha’s great disciples Śāriputra and Maudgalyāyana, contains only very little philosophical discussion – he shows, in his epico-lyrical Life of the Buddha (Buddhacarita, BC or Life), that the path identified and experienced by the Bodhisattva is the only rational – hence, legitimate – method for obtaining salvation. In doing so, Aśvaghoṣa portrays the future Buddha as a philosopher assessing concurrent systems and establishing his own method by means of reasoned arguments alone.

Background According to the fairly unanimous colophons of his works, Aśvaghoṣa was a native of Sāketa (or Ayodhyā), the son of lady Suvarṇākṣī (hence probably a brahmin), an eloquent Buddhist monk, teacher, and poet – perhaps even a great dialectician, depending on the proper interpretation of mahāvādin.2 Chinese legends, some of which likely go back to Indian prototypes, portray him as a (likely Śaivite) brahmin who converted to Buddhism after being defeated in debate by the Sarvāstivādin hierarch Pārśva.3 Be that as it may, Aśvaghoṣa likely was ordained as a monk in the Sarvāstivāda monastic order. In spite of legends associating him with King Kaniṣka (crowned in 127 CE) and the so-called Kashmir council, where he would have been in charge of writing down the Mahāvibhāṣā (a commentary on one of the six Sarvāstivāda canonical Abhi‑ dharma works, the Method for Knowledge (Jñānaprasthāna)),4 there are good reasons to believe that Aśvaghoṣa was close to a milieu of Sarvāstivādins whose doctrines at times DOI: 10.4324/9781351030908-10

83

Vincent Eltschinger

conflicted with “orthodox” Sarvāstivādin-Vaibhāṣika tenets and stood much closer to what later came to be labeled Sautrāntika and/or Yogācāra.5 This is, in fact, hardly surprising, for Aśvaghoṣa’s Handsome Nanda has been rightly regarded as an early witness to a milieu of Sarvāstivādins that laid strong emphasis on meditation and visualization practices, the maturation of which eventually resulted in the Yogācārabhūmi (stages of yoga practice/practitioners) treatises.6 Whatever the case may be, Aśvaghoṣa’s exclusive reliance on sūtras while discussing doctrinal issues makes him literally – though anachronistically – a Sautrāntika.7 Aśvaghoṣa perhaps would have recognized himself in the description of the Sarvāstivādins made in a later scripture, the Mahāvaitulya-Mahāsannipāta,8 where the members of this order appear as experts in Buddhist dogmatics and textual exegesis, but also, and this is perhaps more surprising, as specialists in the controversy against the non-Buddhists, an area almost completely absent in the Sarvāstivāda Abhidharma. That the controversy against the non-Buddhists formed an essential part of the self-representation of the (Mūla)sarvāstivāda order is made perfectly clear in a remarkable passage from the Vinaya of the Mūlasarvāstivādins accounting for the last existence of the Bodhisattva.9 About to leave the Tuṣita heaven where he had spent his penultimate existence, the future Buddha is held back by the gods, who inform him about the poisonous conditions obtaining on earth due to the presence of six allodox “reasoners” (tārkika), six brahmin “traditionalists” (ānuśrāvika), and six “meditators” (samāpattṛ). Far from being frightened, the Bodhisattva resolutely descends to earth to defeat these perturbators, and only then to reach nirvāṇa. The Mūlasarvāstivāda Vinaya thus features an agonistic Bodhisattva coming to earth in order to vanquish non-Buddhist teachers, and there is every reason to suspect that the (Mūla)sarvāstivādin intellectual elite was thus encouraged to view philosophical controversy against the non-Buddhists as a welcome imitation of the founder. If there is no compelling evidence that Aśvaghoṣa was aware of this passage, there are good reasons to believe that he was fully sympathetic with its message. This can be seen from his summary of the Buddha’s career in BC 25.9 (P103a6–7), according to which the Buddha, “after refuting the allodox erroneous paths, proceeded on such a path that he could teach the right path,” or in the claim that 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. (BC 27.30–32, P115b7–116a2) Aśvaghoṣa’s martial repertoire is thus perfectly in tune with the preceding passage from the Mūlasarvāstivāda Vinaya and shows that, for him, too, defeating the non-Buddhists was an essential feature of the Bodhisattva’s spiritual career and a prerequisite for salvation.

Philosophy and Apologetics Aśvaghoṣa did not build a system of philosophy. Nor did he develop any particular method. However, if there is nothing like “Aśvaghoṣian” philosophy, his works exhibit a great many rational arguments concerning issues that eventually became the object of heated philosophical controversies, as well as a philosophical agenda mirroring distinctly apologetic concerns. As shown by Patrick Olivelle and others, the highly polysemic notion of Dharma is one of the Life’s main sites of contention.10 In Chapters 9–10 of the Life, Aśvaghoṣa dismisses orthodox Brahmanical attempts to prove the Bodhisattva’s abandonment of princely and domestic life to be morally wrong. For all this, Aśvaghoṣa has nothing to object to the normativity of Dharma 84

Aśvaghos∙a

in the socio-political realm11 as long as – contrary, for instance, to caste regulations – its prescriptions are rationally justified and therefore morally satisfactory. But a significant part of the Life of the Buddha is intended to demonstrate that no law outside the one identified, experienced, and preached by the (future) Buddha – i.e., Buddhism – is relevant as far as religion as a system of salvation is concerned. In other words, Aśvaghoṣa’s Life of the Buddha is destined to demonstrate Buddhism’s monopoly over salvation, and it is in this apologetic perspective that Aśvaghoṣa’s activity as a philosopher is best understood. Let us briefly go back to the canonical accounts of the main events that took place between the Bodhisattva’s Great Departure and the attainment of Awakening. Gautama had first been initiated by the teachers Arāḍa Kālāma and Udraka Rāmaputra to higher meditative attainments – the stage of nothingness and the stage of neither-consciousness-nor-unconsciousness. Judging these practices unsatisfactory, i.e., not conducive to liberation, the Bodhisattva left these teachers and turned to asceticism, which he subsequently abandoned on the grounds that it made salvation impossible. However, the canonical accounts provide no argument for the Bodhisattva’s rejection of these practices. Aśvaghoṣa’s original contribution was to supply reasons (hetu, yukti) in order to account for the future Buddha’s dissatisfaction, thus ascribing him a systematic assessment of concurrent systems for salvation – in Sanskrit, a parīkṣā, “critical examination,” or vicāra, “critical analysis.” In Handsome Nanda, the Bodhisattva is described as a path expert in search of certainty concerning the (un)reliability of the available methods for salvation and critically analyzing “which of the various scriptural traditions (āgama) in the world is the highest” (SNa 3.1–5). Critical examination of concurrent truth claims proved to be of decisive importance for the development of philosophy in the Buddhist environment (at least) – this is what the Levels of Yoga Practice (Yogācārabhūmi) subjects the “allodoxies” to;12 what logic and dialectics are designed for according to our earliest extant Buddhist witnesses; the title of numerous Madhyamaka and epistemological works and chapters; the very way in which Dignāga (480–540) and Dharmakīrti (around 600?) understand their own philosophical endeavors. Buddhist philosophy essentially consists in a “critical examination,” i.e., in the use of reasoning ( yukti) to evaluate concurrent religio-philosophical systems and to establish Buddhism, or a particular interpretation thereof, as the only rationally founded soteriological method. If critical examination ( parīkṣā) is the closest one gets to a designation of philosophy in the early Buddhist environment, can the parīkṣaka – the person who carries out a critical examination – be regarded as the Indian Buddhist counterpart of the philosopher? Interesting in this connection is what the formerly Sāṃkhya-oriented ascetic Subhadra says to the Buddha in the Life (BC 26.7ab, P107a8–b19): “It is said that you have gained a path of salvation other than that of parīkṣakas like myself.” Significantly, Johnston translates parīkṣaka with “philosophers,” quite rightly so, considering Subhadra’s acknowledgment that previously he had held with respect to that which is manifested 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. (BC 26.17, P108a3–4)13 Subhadra is thus a parīkṣaka in that he holds speculative (and, in his case, ultimately wrong) views about the world and human nature. Aśvaghoṣa’s parīkṣakas are also particularly interested in psychology. Accordingly, parīkṣakas who know the soul are said to be doctors for the mind, and the Buddha is regarded as a parīkṣaka for analyzing the “dispositions, tendencies, and feelings of all beings” (SNa 8.5 and 9.52cd). It is due to not being a parīkṣaka, i.e., due to not analyzing things, that the Buddha’s half-brother Nanda is said to have formerly believed 85

Vincent Eltschinger

that heaven was eternal (SNa 12.4ab). Analyzing reality (tattva) is what characterizes parīkṣā as an intellectual activity and the parīkṣaka as one who devotes themself to it. Whereas Handsome Nanda 14.48 shows that being a parīkṣaka does not necessarily presuppose a direct intellectual intuition of the true nature of things, other passages in the same work clearly link the parīkṣaka with a soteriologically relevant analysis of reality.14 This was typically the case of the newly converted Nanda as he finally “examined things according to their prerequisites, their causation, their nature, the sensation with which they are experienced and their individual defects” (SNa 17.15). Extrapolating a little, one could perhaps say that parīkṣā as a discursive operation refers both to philosophy as the critical examination of concurrent speculative views and to philosophy as a soteriologically oriented analysis of reality. The parīkṣaka’s analysis of reality causes them both to criticize alternative accounts of the world and to set out on the path to mental tranquility, insight, and salvation.

A Critique of Rituals and Asceticism Shortly after departing from his father’s palace, the Bodhisattva visits a forest hermitage (āśrama) with various sorts of brahmin ascetics and their wives, all “desirous of Paradise and working to accumulate merit” (BC 7.10). According to Aśvaghoṣa’s lively description, this penance grove was in full activity, a workshop as it were of dharma, with the transference elsewhere of the blazing sacrificial fires, with its throngs of seers who had completed their ablutions and with the shrines of the gods humming with the din of prayers. (BC 7.33) To achieve their religious ends, the hermits dedicate themselves to austerities (tapas) and practice various observances and rituals – what the Bodhisattva calls “methods of dharma” (dharmavidhi). “With such austerities accumulated for the due time,” one of the forest-dwellers says, “they win by the higher to Paradise, by the lower to the world of men. For bliss is obtained by the path of suffering; for bliss, they say, is the ultimate end of dharma” (BC 7.18). Diverse as they may be, however, all these ascetic and ritual practices have in common to be aimed at better forms of rebirth – ideally heaven – but not at getting rid of existence itself, which is the Bodhisattva’s purpose, he who has “made a vow for the annihilation of existence” (BC 7.44). Since, at this early stage of his career, the future Buddha has no direct intuition of the truth, he subjects the hermits’ implicit assumptions to a rational examination. In Life 7.20–31, the Bodhisattva’s main line of argument against these Brahmanical practices is that they are absurd as long as they are not aimed at complete liberation from existence (the heavens they are conducive to are – admittedly ethereal and pleasurable – forms of existence). According to him, it is absurd to expect heaven, which is itself painful in as much as it is transient (see what follows), from painful ascetic practices; absurd, to practice restraint in order to achieve heaven, which, as a “karmically” conditioned state, is nothing but a form of bondage; absurd, to long for the continuance of existence, with all that is known of the miseries of saṃsāra; absurd, finally, to expect to rid oneself of the fear of death by ascetic practices that bring about new births and deaths. As a consequence, “living beings, making themselves miserable in their hopes of bliss, miss their goal and fall into calamity” (BC 7.24). Moreover, if the mortification of the body is dharma, then the corporeal bliss that one longs for in heaven is adharma, and therefore “dharma in this world bears as its fruit what is contrary to dharma” (BC 7.26; and “in as much as it is under the direction of the mind that the body acts and ceases to act, therefore it is the taming of the mind only that is required,” BC 7.27). Does religious merit accrue 86

Aśvaghos∙a

from the purity of food? Then deer and poor people will have much merit. Moreover, water is just water and will not make a sinner pure. As an ascetic says at the end of the chapter, the Bodhisattva, “having properly analyzed heaven and release, has opted for release,” and this on the basis of various arguments (bahuyuktiyukta; BC 7.52 and 32).

The Case of Arāda Before leaving the forest hermitage, one of the anchorites – likely a Śaiva professing Sāṃkhya views15 – invites the Bodhisattva to pay a visit to Arāḍa Kālāma, a sage teaching the “path of the tattvas” (BC 7.55). According to canonical literature, Arāḍa and Udraka Rāmaputra – Gautama’s first instructors in meditation – were teaching stages of ecstatic trance called the “incorporeal realms.” These practices likely belonged to what Johannes Bronkhorst has called “mainstream meditation,” i.e., Jaina and “Hindu” ascetic traditions that  – contrary to Buddhism – understood meditation as a painful and “forceful effort to restrain the mind and bring it to a standstill” and as a medium “by which the practitioner gradually puts an end to all ideations” (Bronkhorst 1993, 22, 81). Aśvaghoṣa thus unsurprisingly presents Arāḍa as an exponent of liberation (BC 11.69 and SNa 3.3) who has “gained insight into final beatitude” (BC 7.54). But contrary to the canonical narratives, which remain entirely silent on Arāḍa’s and Udraka’s theoretical commitments, Aśvaghoṣa introduces Arāḍa’s doctrine as a philosophical system (darśana, śāstra) entailing both a theoretical and a practical aspect. Thus, far from being just the meditation specialist depicted in the canonical sources, Aśvaghoṣa’s Arāḍa is, first and foremost, a philosopher – more precisely, a representative of the type of “early” Sāṃkhya associated with the teachers Vārṣagaṇya and Pañcaśikha.16 To the best of my knowledge, however, no prior Buddhist account of Arāḍa associates him with Sāṃkhya. Moreover, Arāḍa is recorded in none of the traditional lists of Sāṃkhya and/or Yoga teachers. Arāḍa’s system, as presented by Aśvaghoṣa, differs markedly from “classical” Sāṃkhya, yet bears close affinities with Sāṃkhya as it is expounded in the Carakasaṃhitā and in different passages of the Mahābhārata.17 What is striking about Aśvaghoṣa’s reworking of the story is that, far from rejecting Arāḍa’s and Udraka’s soteriologies after practicing them, as the traditional accounts have it, the Bodhisattva dismisses them on purely theoretical grounds after submitting them to rational examination. His argument starts as follows (BC 12.69): “I have listened to this doctrine of yours, which grows more subtle and auspicious in its successive stages, but I consider it not to lead to final beatitude, since the field-knower [i.e., the self] is not abandoned.” Arāḍa’s Sāṃkhyayoga is a soteriological system aiming at the self’s complete isolation from whatever is not the self, such as passion, corporeality, and mental processes. Once the culmination of the path – the stage of nothingness – has been reached, the self escapes from the body as a bird from its cage. In Life 12.69–82 and especially 69–76, the Bodhisattva attempts to demonstrate that this very conception makes salvation altogether impossible.18 Although most of the subsequent argument operates, I believe, on the ontological level, it can also be interpreted as suggesting that, since the knower of the field is not abandoned, the very notion of a self – which, according to most Buddhists, is what enslaves beings to existence and suffering – cannot be eliminated, either. In his play Drama of Śāriputra, the poet, after “quoting” Life 12.75 (see what follows), declares that “cessation is not definitive if there still is a belief in a self, and is due to the perception of selflessness alone.”19 And in Life 17.12–14, Śāriputra – who previously had been an adherent of Sāṃkhya – criticizes his former creed in the following way: When one embraces the idea of the soul (ātman) as the origin, there is no abandonment of the ego-principle, and therefore no abandonment of the ego. When a lamp 87

Vincent Eltschinger

and the sun are both present, what is to be known as a cause of the destruction of light? Just as if one cuts off the roots of a lotus, the fine fibers remain entangled with one another, so he deemed the [Sāṃkhya] method of salvation not to lead to finality, while the Buddha’s method was like cutting a stone. (BC 17.12–14) Positing a self amounts to reinforcing an already deeply ingrained sense of ego. The perception of selflessness is the only way to counteract the belief in a self; it is only when the latter has been eradicated that salvation can be considered truly final. The non-abandonment of the self makes this impossible. In Life 12.70, the Bodhisattva gives the reason why in his opinion the self (hence the belief in a self) cannot be entirely eliminated (BC 12.70): “For I am of opinion that the field-knower, although liberated from the primary and secondary constituents, still possesses the quality of giving birth and also of being a seed.” As pointed out by Johnston (1984, 178, n. 70), giving birth and being a seed are well-known expressions in Sāṃkhya sources, some of which regard the unmanifest (avyakta) – which, according to them, is none other than the self – as productive. They then explain how the intellect, the ego-principle, the five great elements, etc., are successively evolved from the unmanifest, the self. Life 12.70 seems to presuppose similar ideas on the self: far from being the unproductive witness of classical dualistic Sāṃkhya, the self is the dynamic source of everything else. But if the qualities of giving birth and being a seed belong to the very essence of the self/field-knower, they will so remain even beyond liberation. In other words, the self will continue to produce that which it is supposed to rid itself of, at least provided the causal conditions are present. This is apparently what Aśvaghoṣa suggests in the next stanzas: For although the soul by reason of its purity is conceived as being liberated, it will again become bound from the continued existence of the causal conditions. Just as a seed does not grow for want of the proper season, soil or water, but does grow when these causal conditions are present, such I deem to be the case of the soul. (BC 12.71–72) To better understand this, let us turn to Life of the Buddha 26.11–14. In BC 26.11, which presents a preclassical Sāṃkhya path to liberation, the release of the self is claimed to come about by the increase of sattva (“goodness”) and the destruction of the result of past action (karman) through the elimination of rajas (“passion”) and tamas (“darkness”).20 Against these tenets of his former Sāṃkhya persuasion, according to which prakṛti (“Nature”) is permanent, Subhadra argues as follows: For in the world they attribute darkness and passion, which delude the mind, to Nature. Since Nature is acknowledged to be permanent, those two equally do not cease to exist, being necessarily also permanent. Even if by uniting oneself with goodness those two cease to exist, they will come into being again under the compulsion of time, just as water, which gradually becomes ice at night, returns to its natural state in the course of time. Since goodness is permanent by nature, therefore learning, wisdom and effort have no power to increase it; and since it does not increase, the other two are not destroyed, and since they are not destroyed, there is no final peace. (BC 26.12–14, P107b3–108a6)

88

Aśvaghos∙a

The parallelism between the arguments of BC 12.71–72 and 26.12–14 is striking. In addition, as for the statement that liberation is deemed to come by severance from the power of the act, from ignorance and from desire, there is no complete severance from them so long as the soul persists. It is true that advance is obtained by the progressive abandonment of these three, but where the soul still remains, there these three remain in a subtle state. (BC 12.73–74) To be sure, progress on the path and liberation presuppose the gradual/total elimination of ignorance, craving, and (bad) action, but the belief in a self makes such an elimination impossible, for this belief is the very content of ignorance and craving, its consequence. In other words, ignorance, craving, and action can perhaps be reduced, but certainly not uprooted. Besides, the liberation claimed by Arāḍa to take place upon reaching the stage of nothingness is purely illusory, as the Bodhisattva makes clear in BC 12.75: “But such liberation is a creation of the imagination based on the subtlety of the faults, the inactivity of the mind and the length of life in that state.” The inactivity of the mind and the length of life alluded to here are to be understood in the context of the fourth dhyāna, which plays a major role both in the Buddhist and in Arāḍa’s analysis. Whereas the inactivity of the mind likely refers to the suspension of intellectual and affective events arrived at by the yogin, the length of life is that of the deities inhabiting those ethereal realms. And indeed, life expectancy in the incorporeal realms is well codified in Abhidharma: 20,000 cosmic periods in the stage of the infinity of space; 40,000 in the stage of the infinity of cognition; 60,000 in the stage of nothingness; and 80,000 in the highest state of existence.21 According to the Bodhisattva, then, those rarefied cognitive and affective events occurring in next to endless existences can easily be mistaken for liberation. But from a Buddhist perspective, there can be no question of liberation at that stage. However subtle, defilements are still polluting those living in these realms. From a Buddhist perspective, the elimination of defilements by non-Buddhist (“mundane”) paths is neither sufficient nor capable to lead to salvation; the elimination of false views – first and foremost, the false view of a self – is required, and this can only occur by perceiving the Buddhist truths. BC 12.77 opens a new section in the Bodhisattva’s argument, with Vaiśeṣika now under attack, as suggested by Bronkhorst: “And as the [soul] is not released from number, etc., it is not devoid of qualities; therefore, as it is not devoid of qualities, it is not admitted to be liberated.”22 In Vaiśeṣika, numbers are qualities (guṇa), and “even a liberated soul will, from the Vaiśeṣika perspective, possess the quality ‘number’ by virtue of the fact that it has a number: each liberated soul by itself is one in number” (Bronkhorst 2005, 595). And among the many qualities that can reside in the soul, some (including number) are explicitly said to remain in a liberated soul. As remarked further by Bronkhorst, the next stanza, BC 12.78, is also to be interpreted against a Vaiśeṣika background: “For no distinction exists between the qualities and the possessor of the qualities; for instance, fire is not perceived, when devoid of outward appearance (rūpa) and heat (uṣṇa).” This argument would make little sense against Sāṃkhya, according to which everything merely consists of a collection of qualities with no distinction between the qualities (guṇa) and their possessor (guṇin). According to the Bodhisattva, not only does the soul possess qualities such as number, but it also must possess them since nothing can be perceived independently

89

Vincent Eltschinger

of its qualities. With BC 12.79, Aśvaghoṣa seems to move one step further, claiming that no guṇin can be said to exist without qualities, which corresponds to the Vaiśeṣika understanding of substances (dravya): Before a conglomerate mass exists, there cannot be a possessor of the mass; so, before qualities exist, there cannot be a possessor of the qualities. Therefore the soul, as possessor of the body, being first released, is subsequently bound to it again. The Bodhisattva’s last argument against Arāḍa’s position is in the form of a dilemma (BC 12.80–81): And the knower of the field, when without a body, must be either knowing or unknowing. If it is knowing, there is something for it to know, and if there is something for it to know, it is not liberated. Or if your teaching is that it is unknowing, what then is the use of inventing the existence of a soul? For even without a soul the existence of the quality of not-knowing is well established as in the case of a log or a wall. As far as I can see, the Sāṃkhya, both “early” and “classical,” is immune to the first branch of the dilemma. The position of classical Sāṃkhya on the issue of the self’s liberated state can be summarized as follows: although it is conscious by its very nature, the self does not cognize anything due to its dissociation from prakṛti and its evolutes – the intellect, the mind, the sense organs, their objects, etc. The liberated self is thus characterized by a kind of nonintentional, contentless, or non-referential consciousness. This position exposes the Sāṃkhya to the second branch of the dilemma, following which an unknowing liberated self would be similar to a log or a wall and thus both unattractive and useless. Such a liberated self is frequently compared to a stone by the opponents of the Vaiśeṣika and the Naiyāyika view of liberation.23 And indeed, that the liberated self is insentient (thus incapable of experiencing pain, but also pleasure) and unknowing is a characteristic position of the Vaiśeṣika and Nyāya schools.24 As objections found in Pakṣilasvāmin’s and Uddyotakara’s commentaries on Nyāyasūtra 1.1.22 clearly testify, this view has been criticized from very early times on the grounds that such a liberated state cannot be striven for. Aśvaghoṣa certainly belongs to the first philosophers to criticize these Sāṃkhya – and, most probably, Vaiśeṣika – ideas on the liberated state. The Bodhisattva concludes his critique of Arāḍa’s position on the following stanza (BC 12.82): “But since this successive abandonment is declared to be meritorious, therefore I deem complete success in reaching the goal to derive from the abandonment of everything.” In accepting that a self is liberated, Arāḍa’s system fails to abandon everything, and thus makes the elimination of the belief in a self, hence liberation, impossible. Liberation presupposes riddance not only of what is falsely believed to belong to the self – the Sāṃkhya prakṛti-cumvikāra, the Vaiśeṣika guṇas, and so on – but also of the self.

Against the Self In his critique of Arāḍa’s Sāṃkhya system, however, Aśvaghoṣa does not refute the self. His arguments appear somewhat later in the poem, in canto 16, as the newly awakened Buddha preaches selflessness to King Bimbasāra and his subjects while arriving at Rājagṛha.25

90

Aśvaghos∙a

According to the Buddha, if a self existed, it would be either permanent or impermanent; great problems (doṣa) follow in both these hypotheses. To begin with, if it were impermanent, there would be no fruit of the act; and, since there would thus be no rebirth, salvation would come without effort. (BC 16.80–81) An ephemeral agent bears no moral responsibility for his own actions, for there is no continuum to link up actions and their results by underlying the maturation process. But insofar as it is past actions that bind to saṃsāra, there would be no bondage, which simply amounts to liberation. In other words, the final aim of religious life would be achieved without effort.26 Aśvaghoṣa continues (BC 16.82): “But if the self were permanent and all-pervading, there would be neither absence of birth nor birth; for space, which is all-pervading and permanent, neither passes away nor is born.” By definition, whatever is permanent and thus undergoes no change knows neither rise nor fall. However, living beings are seen to be born and to die.27 If the self is all-pervading but not permanent, the following absurdity ensues (BC 16.83): “If this self were all-pervading in nature, there would be no place where it is not; and when it passed away, there would simultaneously be salvation for everyone together.” Suppose the self is all-pervading: it will, then, be present in all beings engaged in all retributive destinies, and its passing will entail the same consequences as those outlined in Life 16.81, except for their collective, universal application: for all living beings engaged in transmigration, at the time of this universal self’s destruction, bondage will stop; hence, liberation be achieved, for want of karmic results and rebirth. In addition, “as being all-pervading by nature, the self would be inactive and there would be no doing of the act; and without the doing of acts, how could there be the union with the fruit of them?” (BC 16.84). Action involves an agent, an instrument, something that is acted upon, and a specific location where it takes place. An all-pervading entity would be either all this at the same time, or rule out the possibility of anything other than itself. But the absence of both action and agent makes karmic retribution and an experiencer impossible. Suppose, however, that the self – which both Buddhist and Brahmanical sources take to be autonomous (svatantra) – is an agent. Then “if this self did perform deeds, it would cause no suffering to itself; for who, that is absolutely autonomous, would cause suffering to himself?” (BC 16.85).28 Let it be recalled here that the prototype of Aśvaghoṣa’s argument is the so-called second sermon of the Buddha in Benares, in which the Buddha speaks as follows (SBhV I.138,10–20): Corporeality, O monks, is not the self. For if corporeality were the self, O monks, corporeality would tend neither to harm nor to suffering, and with regard to corporeality wishes such as: “I wish my body were so and so,” or “may it not be so and so” would be fulfilled. But since corporeality is not the self, therefore corporeality tends to harm and suffering, and with regard to corporeality wishes such as: “I wish my body were so and so,” or “may it not be so and so” are not fulfilled. Similarly, affective sensation, ideation, the conditioning factors and awareness are not the self.29 Aśvaghoṣa continues (BC 16.86): “Due to its being permanent, the self undergoes no change; but since it experiences pleasure and suffering, we see that it does in fact undergo change.” The idea expressed in this stanza underlies the entire Buddhist critique of permanent

91

Vincent Eltschinger

entities such as God and the self: either such an entity is permanent – and hence, it cannot undergo any change – or it undergoes change – and hence, it cannot be regarded as permanent.30 The permanence of the self is not only challenged by the many changes this self is bound to undergo as an agent, an experiencer, and the very substratum of psychophysical life. Another threat consists in the fact that salvation presupposes change in the form of a process of gradual improvement and purification. Aśvaghoṣa spells this out as follows (BC 16.87): “Salvation comes from the winning of knowledge and the abandonment of defilements; but since the self is inactive and all-pervading, there is no salvation for it.” As we have seen previously, a permanent and all-pervading self neither changes nor acts. However, transformation and action are necessary conditions for achieving liberation. Indeed, soteriology – and especially Buddhist soteriology – entails both a via illuminativa and a via purgativa: the gradual acquisition of gnosis through scriptural instruction, philosophical investigation, and meditative practice parallels (and more often than not conditions) the progressive neutralization and elimination of gross and subtle defilements. Now, according to Aśvaghoṣa, the self’s all-pervasiveness makes liberation impossible in that it precludes agency and action.31

On Some Metaphysical Assumptions In SNa 16.17, Aśvaghoṣa claims that the cause of this suffering from active being in the world is to be found in the category of the vices such as desire and the rest, not in a Creator or Primordial Matter or Time or the Nature of Things or Fate or Chance. The same topos underlies BC 18.18–56, a fascinating but difficult passage32 in which Aśvaghoṣa (alias Anāthapiṇḍada/Sudatta, the rich merchant who gave the Buddhist community its first monastery, the Jetavana in Śrāvastī) criticizes all available metaphysical hypotheses concerning the origin and/or the nature of the universe: God, Nature, Man, Time, Chance/Causelessness. In these two passages, the poet resorts to a fairly widespread motif whose locus classicus (and origin?) is often believed to be Śvetāśvatara Upaniṣad 1.1–2ab: What is the cause of brahman? Why were we born? By what do we live? On what are we established? Governed by whom, O you who know brahman, do we live in pleasure and pain, each in our respective situation? Should we regard it as time, as inherent nature, as necessity, as chance, as the elements, as the source of birth, or as the Person? Or is it a combination of these? This ubiquitous motif likely was intended to map the various theoretical possibilities in metaphysical matters and, perhaps as a sort of scholarly exercise, to allow one to find one’s place in the picture and to criticize the other hypotheses. Be that as it may, the topos can be seen to structure much of philosophical controversy until at least Dharmakīrti (PV 2.179ff.) and Śāntarakṣita’s Tattvasaṅgraha (8th century CE). Thanks to a sermon delivered by the Buddha (BC 18.3–14), Sudatta has reached the first fruit of religious life. “With the correct view born in him, he shed the various wrong views, like an autumnal cloud shedding a shower of stones, and he did not hold that the world proceeded from a wrong cause, such as a Creator and the like, or that it was uncaused” (BC 18.18, P79a6–7). The wrong causes (viṣamahetu) criticized in canto 18 consist in a creator God (īśvara, verses 19–28), Nature (either svabhāva or prakṛti, verses 29–41),33 Time (kāla, verse 42), and Man 92

Aśvaghos∙a

( puruṣa, verses 47–51); as for the absence of cause, it is refuted in verses 52–54. According to Aśvaghoṣa, if God were the creator, the world would know neither process nor change, and the living beings no transmigration; men would have no doubt about him and love him like a father; they would have no free will and bear no responsibility. If Nature is one (ekātman), all-pervading (vyāpin, hence the universal cause of everything), without attribute (nirguṇa), permanent (nitya, hence not liable to change), productive in essence (  janaka), as its advocates believe, then its effects should be one, unlimited, without attribute, undifferentiated, and indestructible. However, one observes that, contrary to this, they are manifold, limited, endowed with specific qualities, differentiated, and ephemeral. “If Time is postulated as the creator of the world, then there is no liberation for seekers. For the cause of the world would be perpetually productive, so that men would have no end” (BC 18.42, P80b3–4). Is man, then, in control of his own destiny? Certainly not, for, as already argued in the case of the self, he would certainly do only what is desirable to himself, obtain whatever he desires, and not let himself be reborn as an animal. Finally, that all things and events obey causes and conditions can be ascertained by mere empirical observation (BC 18.52, P81a6–7): does not “corn grow from the seed with the support of soil and water and by union with the right season”? Verse 56 (P81b2–3) spells out Sudatta’s conclusion: “The various beings too, moving and stationary, come into existence in dependence on various causes; there is nothing in the world without a cause, yet the world does not know the universal cause.”

Conclusion As we can see, Aśvaghoṣa was very well acquainted with contemporary forms of Sāṃkhya and a witness to the rise of Vaiśeṣika, two philosophical systems that remained the main targets of the Buddhist intellectual elites until the sixth century. In addition, Aśvaghoṣa likely took an active part in the formation of Sautrāntika and Yogācāra ideas. An uncompromising defender of Buddhism, he fought against all non-Buddhist religio-philosophical systems to impose Buddhism as a legitimate form of life and as the only source of salvific truth. Like his near contemporary Mātṛceṭa, another celebrated poet, Aśvaghoṣa deeply and permanently impregnated the Buddhists’ imaginaire, aesthetic sensibility, and philosophical repertoire. Even if pramāṇas play virtually no role in his philosophy and the Madhyamaka was apparently not known to him, Aśvaghoṣa surely was one of those who most decisively paved the way for later, “mature” expressions of Buddhist philosophy. Throughout his works, Aśvaghoṣa insisted that philosophical analysis is a precondition of religious commitment, and reason, a sine qua non of liberation. This is a lesson many among our contemporaries would be well advised to meditate upon.

Abbreviations BC

(Buddhacarita, by Aśvaghoṣa). See Johnston 1984. Unless otherwise mentioned, all translations of the BC are Johnston’s. P (Peking Edition of the Tibetan Buddhist Canon). See Suzuki 1957. PV (Pramāṇavārttika, by Dharmakīrti). See Miyasaka 1971–1972. SBhV (Saṅghabhedavastu). See Gnoli 1977. SNa (Saundarananda, by Aśvaghoṣa). See Johnston 1928 for the Sanskrit and Johnston 1932 for the English translation. Unless otherwise mentioned, all translations of the SNa are Johnston’s. ŚvUp (Śvetāśvatara Upaniṣad). See Olivelle 1998, 413–33. T. (Taisho Edition of the Chinese Buddhist Canon). See Takakusu and Watanabe 1924–1932. 93

Vincent Eltschinger

Notes 1 On Aśvaghoṣa, see Salomon (2015). 2 On these colophons, see Eltschinger (2012, 172, n. 3). 3 On these legends, see Young (2015). 4 See Takakusu (1904, 276–79; Li 2002, 9–13); concerning Kaniṣka, see Lévi (1896, 448–49, 472–75). 5 See especially Yamabe (2003); Eltschinger (2020). 6 See Demiéville’s classic exposition in Demiéville (1951); Yamabe and Sueki (2009, xiii–xviii). 7 See Eltschinger (2020). 8 T. 397 (大方等大集經), XIII.159a16–21. For a translation, see Karashima (2015, 138–39). 9 For a more detailed discussion of this passage, see Eltschinger (2019). 10 See Olivelle (2009, xvii–xlix, especially xliii–xlix); Brocquet (2015). 11 See Eltschinger (2018). 12 On the paravāda section of the Yogācārabhūmi, see Eltschinger (2013–2014). 13 For similar statements, see BC 17.4 (P75b7; Johnston 1984, III.24) and 17.10 (P76a5–6; Johnston 1984, III.25). 14 SNa 16.3, tattvaparīkṣaṇa; 17.15, dharmeṣu cakre  . . . parīkṣām; 18.42, vijñātatattvasya parīkṣakasya; 16.47, dhātūn. . . parīkṣamāṇaḥ. On the soteriological aspects of parīkṣā/parīkṣaka, see SNa 9.52, 14.7, 16.3, and 14.7, as well as 18.42 and 18.60. 15 See BC 7.51 and 53, and Johnston (1937, 35–36). 16 Johnston (1937, 9–10). For a description of these teachings, see Johnston (1937, 82–83); on Aśvaghoṣa and Vārṣagaṇya, see Johnston (1937, 66, 71, n. 1, 78, 88). 17 On the Sāṃkhya known to Aśvaghoṣa, see Ramakrishna Rao (1964); Kent (1982). In my opinion, these two scholars forcefully interpret Aśvaghoṣa’s description of Arāḍa’s Sāṃkhya according to the Sāṃkhyakārikā’s dualistic scheme. For another interpretation, see Eltschinger (forthcoming). 18 Dharmakīrti’s critique of the self as the basis of Brahmanical soteriologies, though much more systematic and sophisticated, is based on the same principle. See Eltschinger and Ratié (2013, 187–283). 19 See Lüders (1911, 194). 20 See previously in chapter, and Johnston (1937, 35–36). 21 See La Vallée Poussin (1923–1931, II.173–174). 22 Translation by Bronkhorst (2005, 594). 23 See Eltschinger and Ratié (2013, 235, n. 188). 24 See Eltschinger and Ratié (2013, 231–36). 25 For a contextualized discussion of these arguments, see Eltschinger (2013). 26 For a similar argument, see Lamotte (1949, 742–43). 27 For a similar argument, see Lamotte (1949, 742). To the self’s permanence, Aśvaghoṣa adds allpervasiveness. In other words, there is no point in either time or space that the self would not occupy, so that this self neither changes nor moves. 28 For a similar argument, see Lamotte (1949, 743); Eltschinger (2010, 318). 29 On this sermon, see Wynne (2009). 30 For a similar argument, see Lamotte (1949, 742). 31 Note, however, that most of the later Buddhist arguments do not argue for the impossibility of liberation due to the self’s omnipresence, but due to its permanence. According to Dharmakīrti (PV 2.202), the self’s permanence makes both bondage and liberation impossible: [There can be neither bondage nor liberation for something permanent.] Because the cause of the arising of suffering is [what we call] ‘bondage’: [since what is permanent is devoid of the capacity to bring about anything,] how [could this belong] to [something] permanent? [And] being the cause of the nonarising of suffering is [what we call] ‘liberation’: how [could this belong] to [something] permanent? (Translation by Eltschinger and Ratié 2010, 204–5) See also Eltschinger (2010, 321); Lamotte (1949, 742). 32 In spite of Johnston’s heroic and truly remarkable rendering, Aśvaghoṣa’s critique of these five hypotheses remains very terse due to the Tibetan translation in which it has to be read. Only a short overview will be provided here. Note that, perhaps due to an interpolation, a form of early Sāṃkhya is criticized as a sixth view in verses 43–46 (44–46 are missing in Chinese). 33 Johnston suggested that Tibetan rang bzhin renders Sanskrit svabhāva; I believe that it rather renders Skt. prakṛti. A Sanskrit fragment or quotation would settle the matter.

94

Aśvaghos∙a

References Brocquet, Sylvain. 2015. “Le mot dharma dans le Buddhacarita. Polysémie et rhétorique de la conversion.” In Poïkiloï Karpoï. Exégèses païennes, juives et chrétiennes. Études réunies en hommage à Gilles Dorival, edited by Mireille Loubet and Didier Pralon, 283–311. Aix-en-Provence: Presses Universitaires de Provence. Bronkhorst, Johannes. 1993. The Two Traditions of Meditation in Ancient India. New Delhi: Motilal Banarsidass. ———. 2005. “Aśvaghoṣa and Vaiśeṣika.” In Buddhism and Jainism, Essays in Honour of Dr. Hojun Nagasaki on His Seventieth Birthday, edited by “Editorial committee,” 596–90. Kyoto: Heirakuji Shoten. Demiéville, Paul. 1951. “La Yogācārabhūmi de Saṅgharakṣa.” Bulletin de l’École française d’ExtrêmeOrient 44 (2): 339–436. Eltschinger, Vincent. 2010. “On a Hitherto Neglected Text against Buddhist Personalism: Mahāyānasūtrālaṅkāra 18.92–103 and Its Bhāṣya.” Études Asiatiques/Asiatische Studien 64 (2): 291–340. ———. 2012. “Aśvaghoṣa and His Canonical Sources: II. Yaśas, the Kāśyapa Brothers and the Buddha’s Arrival in Rājagṛha (Buddhacarita 16.3–71).” Journal of the International Association of Buddhist Studies 35 (1–2): 171–224. ———. 2013. “Aśvaghoṣa and His Canonical Sources: I. Preaching Selflessness to King Bimbisāra and the Magadhans (Buddhacarita 16.73–93).” Journal of Indian Philosophy 41 (2): 167–94. ———. 2013–2014. “The Yogācārabhūmi Against Allodoxies ( paravāda): 1. Introduction and Doxography.” Wiener Zeitschrift für die Kunde Südasiens 55: 191–234. ———. 2018. “Aśvaghoṣa on Kings and Kingship.” Indo-Iranian Journal 61: 311–52. ———. 2019. “Why Did the Buddha Come to Earth? Monastic Lineage, Interreligious Polemics, and Philosophy.” In Reasons and Lives in Buddhist Traditions: Studies in Honor of Matthew Kapstein, edited by Dan Arnold, Cécile Ducher, and Pierre-Julien Harter, 293–306. Somerville, MA: Wisdom Publications. ———. 2020. “The ‘dhyāna-Master’ Aśvaghoṣa on the Path, Mindfulness, and Concentration.” In Mārga. Paths to Liberation in South Asian Buddhist Traditions, edited by Cristina Pecchia and Vincent Eltschinger, 99–176. Vienna: Austrian Academy of Sciences Press. ———. Forthcoming. “Aśvaghoṣa on Sāṅkhya.” Eltschinger, Vincent, and Isabelle Ratié. 2010. “Dharmakīrti Against the pudgala.” Journal of Indian and Tibetan Studies (Indogaku Chibettogaku Kenkyu) 14: 185–215. ———. 2013. Self, No-Self, and Salvation. Dharmakīrti’s Critique of the Notions of Self and Person. Vienna: Austrian Academy of Sciences Press. Gnoli, Raniero. 1977. The Gilgit Manuscript of the Saṅghabhedavastu. Being the 17th and Last Section of the Vinaya of the Mūlasarvāstivādin. Part I. Rome: Istituto Italiano per il Medio ed Estremo Oriente. Johnston, E. H. 1928. The Saundarananda of Aśvaghoṣa. Critically Edited with Notes. London: Humphrey Milford, Oxford University Press. ———. 1932. The Saundarananda or Nanda the Fair. Translated from the Original Sanskrit of Aśvaghoṣa. London: Humphrey Milford, Oxford University Press. ———. 1937. Early Sāṃkhya. An Essay on Its Historical Development According to the Texts. London: The Royal Asiatic Society. ———. 1984 (1936). Aśvaghoṣa’s Buddhacarita or Acts of the Buddha in Three Parts: Sanskrit Text of Cantos I – XIV with English Translation of Cantos I – XXVIII, Cantos I to XIV Translated from the Original Sanskrit Supplemented by the Tibetan Version and Cantos XV to XXVIII from the Tibetan and Chinese Versions. New Delhi: Motilal Banarsidass. Karashima, Seishi. 2015. “Who Composed the Mahāyāna Scriptures? – The Mahāsāṃghikas and Vai­ tulya Scriptures.” Annual Report of the International Institute for Advanced Buddhology at Soka University (for the Academic Year 2014) 18: 113–62. Kent, Stephen. 1982. “Early Sāṃkhya in the Buddhacarita.” Philosophy East and West 32 (3): 259–78. La Vallée Poussin, Louis de. 1923–1931. L’Abhidharmakośa de Vasubandhu, 6 vols. Paris and Louvain: Paul Geuthner, J.-B. Istas. Lamotte, Étienne. 1949. Le Traité de la Grande Vertu de Sagesse de Nāgārjuna (Mahāprajñāpāramitāśāstra), vol. II. Louvain-la-Neuve: Université de Louvain, Institut Orientaliste.

95

Vincent Eltschinger Lévi, Sylvain. 1896. “Notes sur les Indo-Scythes.” Journal Asiatique [9e Série] 8: 444–84. Li, Rongxi. 2002. “The Life of Aśvaghoṣa Bodhisattva. Translated from the Chinese of Kumārajīva (Taishō Volume 50, Number 2046).” In Lives of Great Monks and Nuns, 3–13. Berkeley: Numata Center for Buddhist Translation and Research. Lüders, Heinrich. 1911. “Das Śāriputraprakaraṇa, ein Drama des Aśvaghoṣa.” Sitzungsberichte der königlich preussischen Akademie der Wissenschaften zu Berlin: 388–411. Miyasaka, Yusho. 1971–1972. “Pramāṇavārttika-kārikā (Sanskrit and Tibetan).” Acta Indologica 2: 1–206. Olivelle, Patrick. 1998. The Early Upaniṣads. Annotated Text and Translation. New Delhi: Munshiram Manoharlal. ———. 2009. Life of the Buddha by Aśvaghoṣa. New York: New York University Press, JJC Foundation. Ramakrishna Rao, K. B. 1964. “The Buddhacarita and the Sāṃkhya of Arāḍa Kālāma.” Adyar Library Bulletin 28 (3–4): 231–41. Salomon, Richard. 2015. “Narratives and Long Poetry: Aśvaghoṣa.” In Brill’s Encyclopedia of Buddhism, edited by Jonathan Silk, Vincent Eltschinger, and Oskar von Hinüber, vol. I, 507–14. Leiden: Brill. Suzuki, Daisetz T. 1957. The Tibetan Tripitaka, Peking Edition, Kept in the Library of the Otani University, Kyoto. Tokyo and Kyoto: Tibetan Tripitaka Research Institute. Takakusu, Junjiro. 1904. “The Life of Vasubandhu by Paramārtha.” T’oung Pao [Second Series] 5: 269–96. Takakusu, Junjirō 高楠順次郎, and Watanabe Kaikyoku 渡邊海旭. 1924–1932. Taishō shinshū daizōkyō 大正新脩大蔵経. Tokyo: Taishō Issaikyō Kankōkai 大正一切經刊行. Wynne, Alexander. 2009. “Miraculous Transformation and Personal Identity: A Note on the First anātman Teaching of the Second Sermon.” Thai International Journal for Buddhist Studies 1: 85–113. Yamabe, Nobuyoshi. 2003. “On the School Affiliation of Aśvaghoṣa: ‘Sautrāntika’ or ‘Yogācāra’?” Journal of the International Association of Buddhist Studies 26 (2): 225–54. Yamabe, Nobuyoshi, and Fumihiko Sueki. 2009. The Sutra on the Concentration of Sitting Meditation (Taishō Volume 15, Number 614). Translated from the Chinese of Kumārajīva. Berkeley: Numata Center for Buddhist Translation and Research. Young, Stuart H. 2015. Conceiving the Indian Buddhist Patriarchs in China. Honolulu: University of Hawai’i Press, Kuroda Institute.

96

6 THE MILINDAPAÑHA How to Use a Philosophical Resource and Find a Literary Gem Sonam Kachru

– “Reverend Nāgasena, will you converse with me?” – “I will . . . if you converse in the speech of the wise and learned, but not if you converse in the speech of kings.” The Questions of Milinda (after Horner 1963, 38–39)

The Questions of Milinda – or Milindapañha in Pāli1 – is the most widely known version today of The Questions of Menander, the name I will initially use for what scholars believe to have been a genre of works from ancient South Asia. The Questions of Menander ostensibly records for us conversations held at Sāgala (today, Sialkot) over a few days, featuring the Bactrian King Menander and a Buddhist monk named Nāgasena. (The name “Milinda” is how “Menander” survives in Pāli.) In addition to the Pāli work, we also possess in Chinese The Sūtra of the Monk Nāgasena (a fourth-century CE translation prepared from an earlier thirdcentury CE translation), possibly reflecting materials from Gandhāra (Demiéville 1924; Fussman 1993, 63–66; Baums 2017, 33). There is evidence that there were other representatives of this genre, now – except for fragmentary traces – lost (Skilling 1998, 91–92). It is important to recall that Menander enjoyed a life off the page, as it were. He reigned over the northwest of the subcontinent and the Punjab in 155–130 BCE, and Sāgala may indeed have been his capital (Bopearachchi 1990, 1992). If Menander is remembered in Greek sources primarily for the military victories which extended his power over northern India (Tarn 1951, 225, 266), for Buddhists he was “one of theirs” (Lamotte 1988, 425). I am unsure as to whether Richard Stoneman is right to say that “there is little doubt that the King himself was a Buddhist” (Stoneman 2019, 365); the numismatic evidence is silent on the issue of his personal rather than official convictions (Bopearachchi 1990, 48; Fussman 1993, 85–90). Indic sources, however, do emphasize his fostering of Buddhist institutions (Neelis 2011, 105), and in the Pāli tradition, as the ending of the Questions of Milinda would have it, Menander converts to Buddhism (Neelis 2011, 105). Though we are not compelled to believe that, it is hard to shake off an intriguing fact noticed very early by James Prinsep (see Tarn 1951, 266): Menander’s funeral is described by Plutarch in a way that cannot but now be reminiscent of the distribution of the relics of the Buddha.2

DOI: 10.4324/9781351030908-11

97

Sonam Kachru

Unlike Menander, we have no independent record of the Nāgasena of this dialogue; nor need we believe that he ever existed as an individual, despite the philosopher Vasubandhu in the late fourth and early fifth century CE treating him as a sthavira, an authoritative Buddhist elder and forebear to several traditions (Kapstein 2001, 363; see also Rhys Davids 1891). For our purposes, we can take it that Nāgasena lives, where he lives, as a dramatic character in The Questions of Menander. What of The Questions of Menander? Some might wish to treat the dialogues as invested in documenting rather than imagining a world. Others may hope to read it like detectives, sleuthing for inadvertent (if not deliberate) traces of a multicultural and multi-ethnic past worthy of remembrance.3 Here, however, I ask the reader to practice a variety of forgetting. I want the reader to treat both Menander and Nāgasena as dramatic characters, no more and no less. Forgetting can be productive. I hope that it will allow us to attend to the work as a literary achievement exploring, among other things, what it can be like to learn to think and feel as a Buddhist. I will confine my attention to the Pāli work The Questions of Milinda, to which I will make reference (for the most part) using or adapting the translations in the two volumes of Rhys Davids (1890, 1894, hereafter RD1 or RD2, respectively) or the two volumes by I. B. Horner (1963, 1964, hereafter H1 and H2, respectively).4

The Questions of Milinda and This Guide Not being an expert in this work, I begin like Menander: as an enthusiastic outsider. And like him, I find myself increasingly pulled into the textual universe this work holds inside of itself. My goal is practical: how shall a student make use of this remarkable though forbidding work? The Pāli work can seem heterogenous. Though it is presented as a single work consisting in seven books, one may also treat it as a library of five works redacted possibly over centuries (from the second–fifth centuries CE). On one view (following Von Hinüber 2000, 84), the first three books of the work we now have were once a complete unit. In Book I, we get backstory, featuring the past lives of the protagonists and social and cultural contexts for the conversations. The dialogue proper takes place in Books II–III, which are arguably older still (cf. Norman 1983, 110–12), ending on the evening of the second day. Book III then concludes with a meeting on the morning of the third day as the protagonists offer a retrospective appraisal of the conversations, thus concluding this larger unit. With Book IV, On Difficult Questions, the conversation and text resume on the morning of the fourth day. From this point on, each book introduces a new theme and constitutes for philologists a new historical addition. Book V introduces another day and a new topic, the uniqueness of the Buddha. Book VI discusses laypeople, asceticism, and the way to nibbāna, while Book VII takes a new route, which I call Thinking With Similes. There is then another formal conclusion. The following guide is divided into two parts, each part emphasizing distinct ways of approaching and reading (sometimes overlapping parts of) the work. Part One deals with Books II–VI, and Part Two with the literary unity of Books II–III, which I will refer to as *The Questions of Menander to highlight a difference.5 Reading the work through from Books I–VII can be a very different experience than reading Books II–III alone. The former is triumphal; the latter tragic. Furthermore, reading The Questions of Milinda as a whole, from Books I–VII, does not obviously present itself as a dramatic unity. Nevertheless, one may discern the following trajectory: Milinda travels a pathway from ignorance and skepticism into greater familiarity with the world as Buddhists conceive it to be. Such a trajectory involves the “thickening”

98

The Milindapañha

of theory (discussed in what follows), one feature of which is a growing familiarity with the imagery and values of a Buddhist textual universe.

Part One: The Triumph of Wisdom (Books II-VI) As one goes farther and farther into the work, Milinda seems to gain knowledge not only about Buddhist concepts but also of Buddhist texts in a wide variety of genres, exhibiting thereby considerable skill as a reader. To begin with, however, the focus is on the acquisition of initially unfamiliar conceptual vocabularies and the exercise of skill in reasoning. In a meta-poetic moment in the opening verses of the prologue (Trenckner 1880, 1), the discourse (kathā) of Nāgasena is described as being aesthetically captivating (citra) through the use of illustrative examples (opamma) and arguments – or, more generally, epistemic stances and strategies (naya). This is a characterization with which the Buddha, later in the prologue, will agree (RD1, 6). The Questions of Milinda is not merely a dictionary, conceptual lexicon, or encyclopedic FAQ.6 To be sure, historians may use the work to elicit patterns of intellectual concern. For example, why was it important to ask whether we may ascribe a variety of attention that is appraisable according to ethical and epistemological norms to animals like sheep and goats? Or other complex epistemic states, like wisdom? (II.1.6 in RD1, 51). Such use will not, however, help us address the distinctive analytic method enshrined in conversations employing illustrative cases. Though much more work needs to be done to get anything approaching an inventory of the diverse ways in which illustrations are used, I offer two ways of using illustrations that strike me as being important.

Intelligibility and Acknowledgment Milinda does not only ask questions of the form “What is [the definition of] X?”; he also asks questions like “How could anyone not already committed to Buddhism believe that P?” Here is an example (III.7.2; cf. RD1, 124): “Your people, Nāgasena, say that though a man should live an evil life for a hundred years, at the moment of death, were a thought of the Buddha to enter his mind, he will be reborn among the gods. This I don’t believe.’ ” Sometimes his questions take the form, “I cannot believe P.” Milinda’s questions can challenge the intelligibility and not alone the factuality of certain positions. To not find a proposition intelligible might mean not knowing what someone is saying, not knowing how they could consistently say it, or not knowing why anyone would wish for it to be true. In such cases, conversation often serves to reveal that there is a substantive principle, belief in which blocks the intelligibility of the proposition in question. For example, the King once elicits from Nāgasena that a monk ought to act so that “this dissatisfaction should cease, and no dissatisfaction arise” (III.7.3; cf. RD1, 124–25). The King takes this to entail that something nonexistent, like what is to come, can figure as a reason, and so as a cause, for action. But the King takes it on principle that nothing nonexistent can play that role. Nāgasena uses an illustration to bring the King to see how he himself already acts as if he does not always believe the principle. For example, the King’s prudential reliance on precautionary measures – exemplified in fortification, defense, or the digging of wells – commit him to endorsing sentences of the same logical shape as Nāgasena’s. A more famous example of this occurs in the first conversation. Milinda finds incredible Nāgasena’s saying that his name does not refer to any single entity – no single entity, as it turns out, from among the parts into which he might be resolved, nor any combination of these parts. Milinda finds this incredible

99

Sonam Kachru

because he appears to believe that words must correspond to referents in the world if they are to be meaningful and that metaphysical analysis can always supply us with the referents. In reply, Nāgasena elicits from this, in combination with the King’s use of the word “chariot” and his beliefs about the parts of a chariot, that the King will say about chariots what Nāgasena has said about selves – it is not selflessness that is the source of unintelligibility, but the King’s untested, inconsistently held, and therefore unwise reliance on a dubious principle of meaning (RD1, 43–45). Such illustrations provide conceptual and psychological gains. By proposing counterexamples to a general principle, they suggest that principle’s possible falsity; and by getting someone to acknowledge that their reliance on a principal has been superficial – and, in fact, inconsistent with commitments tacitly held – an illustration may redirect the use of reason. Instead of being used to challenge an interlocutor, reasons may be used to realize that what seems unfamiliar in an interlocutor may have to do with an imperfect understanding that one has of oneself. Reading the arguments in sequence, it is helpful to imagine them finding traction in the mind of a dramatic character. By having his own (hitherto tacit) commitments rendered perspicuous with the help of the illustrations Nāgasena has drawn from ways of life familiar to the King, Nāgasena’s reasons help Milinda become less hostile and more open: more open to the intelligibility of vocabularies and beliefs hitherto alien to him, more open to formerly tacit propositions to which he is really committed and of which he was hitherto ignorant, and thus, more open to the claims of another’s beliefs on him as being possibilities for his own life.

Conceptual Innovation Illustrations may also be used more narrowly: to modulate one’s concepts, for example. At one point, for instance, the King effectively asks if it is worse to do something bad knowingly or unknowingly (III.7.8; RD1, 129). Forensic practices of ascription of blame and punishment – to which the King makes reference in the discussion – as well as received Buddhist concepts of action underscoring the intent with which something is done7 might lead one to reason as follows: if one punishes what is done with some requisite kind of awareness – and typically, we do not punish what is done without some kind of awareness or intent behind it – it must be the case that this is because we account it worse when someone does bad knowingly. Look, however, at Nāgasena’s example: “But what do you think, O King: Take a heated iron ball glowing with flame and heat – one person takes hold of it unknowingly; another takes hold of it knowingly. Who is more severely burnt?” (cf. RD1, 129) What “worse” means is here fixed by the example of “being burnt” – a relatively direct, deleterious and causal consequence of the action for the agent. This is not quite the moral notion defined by reliance on social practices of blame or punishment. Nāgasena has used a clever example to shift the concept of action from a normative context to a (somewhat) more naturalized one. This latter context also modifies what “knowingly” might mean: it need not mean prospective intent, but rather something like being attentive with awareness of what one is doing, here defined by knowledge of the nature of what one is acting with and what it can do.8 Nāgasena’s point, I take it, is not to answer an old question (and so yielding an answer conflicting with orthodoxy). The point, instead, through the proliferation of cases, might be the creation of more context-specific variegated concepts of action.9 Earlier, Nāgasena identifies the following as the hallmark of reasoning as a paṇḍit (II.1.3; RD1, 46): When paṇḍits talk a matter over with one another, there is a winding up, an unwinding . . . distinctions are drawn, and contra-distinctions. . . . 100

The Milindapañha

Winding up and unwinding appear to be phrases in a process of getting an interlocutor to clarify their position by virtue of its presuppositions and entailments (cf. Ganeri 2012, 20–25). I take this to correspond to what is involved in clarifying the intelligibility of a position. The careful drawing of distinctions and corresponding conceptual refinement (and innovation, even) is, on my reading of The Questions of Milinda, essential to the wise exercise of reason. The word “paṇḍita” – a word translated in my epigraph as someone “wise and learned” – combines the virtues of long study, analytic intelligence, and action guiding wisdom (see Collins 2020, 5, 75–76): a paṇḍita/ā knows his or her way around difficult terrain, so to speak, by exercising the virtues which the conversations of Nāgasena and Milinda exemplify, and which a reader, perhaps, is intended to enact.

Logical Topography of the Doctrine In Books IV–VI, the questions acquire more logical structure. The title of Book IV, “meṇḍakapañha,” literally means “A question about a ram.” In the Ummagga Jātaka (Cowell and Rouse 1907, 177–78), the future Buddha solves a puzzle involving a strange fact, the friendship between a dog and a ram observed by a King and encoded in an enigmatic question. By extension, the phrase “ram-question” can mean a riddle, or anything difficult to interpret without wide experience or knowledge. Milinda does not ask riddles, exactly  – he poses dilemmas (ubhato-koṭika). For example: Did the Buddha not answer a certain kind of question because he did not know the answer or because he wished to keep back certain truths? Either way, inconsistency looms for a Buddhist (RD1, 204). There is some precedent in the Pāli canon for the use of dilemmas, though interestingly, they are not principally associated with either the Buddha or his students. The leader of the Jains, Nigaṇṭha Nātaputta, is shown to be keen on having his followers use them in debate and is portrayed as holding that being put a dilemma is like having something stuck in the throat that one can neither swallow nor spit out (see Abhayarājakumāra Sutta, Majjhima Nikāya, i.392; Ñāṇamoli and Bodhi 1995, 498). Unlike what this story suggests of the role of dilemmas, Milinda does not use them as a coercive tool in debate – or if he does, Nāgasena soon repurposes them as tools of clarification. A dilemma exposes a person to prima facie incompatible commitments. In previous books, Nāgasena helps Milinda to put his own commitments into rational order; here, the dilemmas force Nāgasena to contour doctrinal statements into a space of reasons and exhibit their conceptual structure. We begin to see statements of the form: “The Buddha said P because of x, y, z”; “His saying Q* is an instance of believing (the more general principle) Q”; “He said P in the sense of P*, which can be made consistent with ~P”; and so on. Take the following case (IV.5.16; RD2, 16–18). Milinda asks a question that reveals him to be a meticulous reader of Buddhist texts in that he picks up on a contradiction in the Buddha having said the following (which I offer in paraphrase) at different times and in different texts: 1. In former births when human, I acquired the habit of nonharming. 2. When I was Lomasa Kassapa, I had hundreds of creatures killed while performing the Vājapeyya ritual. The Vājapeyya (in Sanskrit), an important and complex Vedic ritual, is sometimes (as here) said to have involved the sacrifice of animals. Milinda is right to note that it is inconsistent to say, “Considering all actions in all lives, I have never Ø-ed,” while also maintaining that 101

Sonam Kachru

“I have Ø-ed.” But there is no inconsistency, says Nāgasena, if the second sentence is read as saying: 2.* When I was Lomasa Kassapa, I Ø-ed, being out of my mind because of lust. Lomassa performed the ritual for a King as a way of securing the latter’s daughter in return. The interesting thing here is not only that Nāgasena produces a sentence which avoids the appearance of flat-out inconsistency; Nāgasena is also using the dilemma to reconstruct a principle with far-reaching implications: What does it mean to have acted, but not as oneself? The King does not understand this. And this may not be his fault. He goes on to provide a reason for his incomprehension (RD2, 17). He argues that it is only possible to Ø out of some strong motivating pro-attitude or condition – involving states like lust, anger, pride, avarice, poverty, and so on – when these express the kind of person one is. On his view, it is a cruel person, for example, who is said to act out of some such state as anger, as that is part of what it means, on this account, to be a cruel person. And if that is true, the clause “by way of strong emotion” or the like can never be exculpatory. Nāgasena disagrees. One can Ø through a motivating force when it neither befits one nor expresses who one is. Even if the motivating state is internal to a person, for example, it can be a state from which the person is alienated. One needs a more refined criterion for internal source-hood of an action, one which recognizes that desires which lie within the boundaries of a person’s psychology may nonetheless be external to who that person is (for a contemporary account, see Sripada 2016). Nāgasena does not only see the need for such an account; he shows us the kind of questions to which such an account will lead and the implications of such a view. How can one know that alienation has taken place? One must learn to listen to another as a literary critic might listen to a poem. Nāgasena cites a verse ascribed to Lomasa and says of it that the sentiment it expresses is inconsistent with any desire on the future Buddha’s part to engage in violence. It shows, instead, that as Lomasa, the future Buddha wished to act one way, but that at the sight of his beloved “lust deranged his mind” (cf. RD2, 18). An unwilling madman and addict through lust, as it were, he was constrained to act harmfully. Nāgasena then goes on to explain the implicit analogy through an extended discussion on the loss of prudential self-interest and moral compass in madness. If to act when alienated from the motivational sources for the actions one commits is analogous to derangement, there is reason to withhold ascribing responsibility to the agent. But then, readers might well ask: is talk of responsibility ultimately tenable? According to Nāgasena, and other Buddhist texts, as long as we do not acquire meta-cognitive control of our action-motivating states – such control being a variety of achievement of regimens of training – we all, in some sense, act as if mad (Collins 2015, 197–202). Recently, there has been growing interest among contemporary philosophers in such issues. The consequences of this “as if ” for ethics, after all, are far-reaching.

Thickening Theory The exploration of the logical topography can be an involved affair. But there are ways in which things can become more involved that do not have to do with logical structure. Consider the following questions. Do wondrous events take place at the shrines memorializing the relics of the worthy dead of the Buddhists (IV.8.51; RD2, 174)? Can the dead derive any benefit from offerings made in their name (IV.8.29; RD2, 151)? Did the Buddha think that trees speak, or not (IV.3.20; RD2, 241)? 102

The Milindapañha

Such questions exhibit the “thickening” of theory. By that I mean to say that the concepts they use and the facts they seek to bring into view have become more “world” involving, with the content increasingly coming to reflect the meanings and values of participants in a particular social world. Such normatively inflected content evinces a particular emphasis on such features as are capable of rendering intelligible or motivating action in the social world they are suited to capture in thought (cf. Scanlon 2003). The thickening of theory in The Questions of Milinda, beginning with Book IV, is reflected in the topics (increasingly having to do with ritual as well as cosmology), as well as in the greater variety of sources to which Milinda and Nāgasena appeal, spanning verses, customs, and prose narratives. For a good example of why one needs to pay attention to thinking “thickly,” we can turn to Milinda’s questions about when the Buddha was Vessantara, which we might term a “master class” in thinking with narrative (IV.8.1, RD2, 114). Vessantara’s birth story is a popular story of the Buddha’s penultimate life as a human in which, as Vessantara, he practiced the ethical excellence of generosity to the extreme of giving away his wife and children. Steven Collins describes it as a story not only extolling certain virtues but thinking critically about them (Collins 2016, 1–2). As Milinda sees it, the ethical difficulty presented by the narrative is magnified by the way the future Buddha gives away his children in that Vessantara (a) gave them without their consent while (b) seeing the suffering this entailed for them and (c) remaining impassive in the face of the suffering his actions manifestly brought about. How can goodness consist in this? I will quote Milinda’s way of making this point at some length with only a few ellipses. The appeal to narrative detail is not incidental to the argument. In accord with what I meant previously by “thickening,” it serves as a paradigmatic way of bringing a moral fact into view (cf. RD2, 115; H2, 96): Nāgasena, the future Buddha did something hard to do when he gave away his own beloved children to the Brahmin as slaves. And the second thing he did was harder still: He bound his own children . . . young and tender though they were, with creeping vines, and seeing them being dragged along by the Brahmin, the Brahmin flogging them with vines, he could bear to look on. And third . . . When his boy came back to him having struggled to free himself from his bonds, the future Buddha again bound him with vine and again gave him away. Fourth . . . When the children, weeping, cried: “Daddy, this ghoul is leading us away to eat us!” he did not comfort them by saying “Don’t be afraid.” Fifth . . . The prince Jāli, weeping, fell at his feet and begged him: “Enough, Daddy, be content with this. Get back Kaṇhājinā. I’ll go with the ghoul. Let him eat me!” The future Buddha did not agree even to this. Sixth . . . He showed no compassion as the boy Jāli lamented: “Have you a heart of stone, Daddy, that you can look upon us, suffering, being led away by the ghoul into the vast, wild forest and not hold us back?” And the seventh thing he did was even harder still – when his children were led away out of sight to something frightening and terrifying his heart did not split into a hundred pieces, a thousand pieces! Milinda’s appeal to the narrative is masterly. It places us in the narrative by inviting us to inhabit alternative points of view. Reading this, we begin with a narrator’s third-personal point of view on an event thinly described. We are then invited to contrast the perspective of Vessantara’s boy with that of the future Buddha. There is also the invoked, but not explored, mute and terrified perspective of Jāli’s little sister, Kaṇhājinā. We look on the future Buddha as his children see 103

Sonam Kachru

him, and with them, perhaps, interrogate his perspective on this horror. Unlike the boy, he is silent. Unlike the girl, he appears initially to be silent and indifferent, without compassion or moral concern. As we watch, at last, the children disappear from his point of view, what might we have felt were we in his place? What do we feel? What ought he (or we) to feel? I. B. Horner has Milinda grant, at the end, that Vessantara was “in great distress and terror” seeing his children being led out of sight (H2, 97). I am not convinced. The text is unclear here, perhaps deliberately so. The frightfulness and terribleness evoked by the sentence describing, from his point of view, the disappearance of his children – “When his children were led away out of sight” – might more naturally be read as qualifying, instead, the murky destiny to which the children are being led, or, though this is strained, the ghoulish Brahmin leading them away; perhaps the ambiguity reflects the limit of the future Buddha’s sight and his growing horror as they vanish (RD2, 115).10 We do know that Vessantara said and did nothing overtly and, perhaps, that he ought to be feeling something; else, Milinda’s narrative argument implicitly asks what distinguishes the future Buddha’s silence from that of the ghoulish Brahmin who leads the children away. Doesn’t Vessantara also stand to gain from this travesty? “What,” Milinda asks, “has a man desiring merit to do with bringing anguish to others? Should he not have given himself away?” (cf. H2, 97). After long discussion, including an argument we shall briefly consider in what follows, Nāgasena, too, appeals to narrative. He does this by invoking Vessantara’s experience after the terrible deed has happened: Vessantara, having given away his wife and children, entered the leaf hut and lay down. Powerful grief rose in him, anguished as he was because of his exceeding love for them. His heart grew hot, and his breath, hot, could not pass through his nose. He breathed in and out of his mouth. The tears which fell were drops of blood that poured from his eyes. (cf. H2, 105; RD2, 126; cf. Cone and Gombrich 2010, 55) For Nāgasena, the grief is caused by his love for his children, but it may not be directed at the suffering of his children and his wife alone. It may signify a revolt at what the supererogatory ideal and excellence of giving requires of him; it may also signify embodiment of the cost of excellence (RD2, 126; see also Heim 2003, 540, 545). Narrative, as appealed to by Milinda and Nāgasena, provokes but does not settle interpretation. Before Nāgasena turns to narrative, he offers some striking ways of conceptualizing the moral character of the life of a being directed at awakening. He suggests thinking of such a being’s life as a public shareable good – a public feast, for example. He also goes on to offer some thoughts on awakening as a goal, readjusting the scope and weight of traditional moral goods. Each of these arguments will require careful unpacking, as Nāgasena himself anticipates. He imagines the moral contestability of Vessantara’s actions and their ambiguous and transgenerational emotional impact as being bound up together (Heim 2003, 546). In fact, when Nāgasena speaks to Milinda of how successive tradition gradually makes it possible for such an event to reach “our meeting here today so that, defaming and disparaging that gift, we question whether it were well given or ill given” (H2, 97), I like to think that he also has in mind future readers of the dialogue, identifying moral deliberation rather than definitive opinion as the vital part of tradition (Heim 2003, 546). While there is much to be said about this, I wish to note just this. Nāgasena’s decision to contest Milinda’s arguments by appealing to the point of view and the reality of psychological cost is telling, for Nāgasena thereby directs us to the following questions. Why would one wish to do something which wrenches 104

The Milindapañha

one’s entire being this way? What is the world (and its moral structure) like that realizing our ideals must exact such a price? The King assumes a kind of indifference on the part of the future Buddha and conceives of such indifference as part of the pursuit of this excellence. Perhaps this is because he cannot imagine a world in which doing good and feeling good do not align. Is he thinking of our world? Here, Nāgasena is effectively telling us that the story of Vessantara presents us with a world wherein, to adapt a phrase from John Rawls (and Isaiah Berlin), there can be no moral life or world without loss (cf. Pogge 2007, 187). Nāgasena is instructing Milinda in tragic choice. We cannot pursue or maintain all that matters to us, and doing what one has overriding reason to do is consistent with ruinous psychological and physical cost. Were one to push back on either of these points, one would inhabit a different moral universe from that in which Vessantara and Nāgasena live, a point that such appeals to narrative, with their richly felt perspectives, help make apparent.

Part Two: The Tragedy of Wisdom (Books II – III) The Questions of Milinda ends in Book VII with the King requesting to be ordained as a monk, though not before building a monastery for Nāgasena. This secures the success of Buddhism in the marketplace of ideas, an image invoked at the very beginning of the prologue, where we are told how Sāgala, rich in markets, features innumerable varieties of well-displayed goods, along with the competing cries of welcome from teachers of every philosophical tradition and many ascetic practices (H1, 2). Discourse is for “the stability of true Dhamma” (cf. H2, 305) – so says the first of the last verses of the text. On this view, it makes sense that the conversations recorded in this text be seen as a means to shore up the faith against such times as when “the tradition of the Buddha is in danger of crumbling away” (to use, once more, the language of the prologue) (RD1, 14). But that is not the only story we might tell. At the end of *The Questions of Menander – or at the end of what is now Book III of the Questions of Milinda11 – the camera does not, so to speak, pan back far enough to take in the historical vicissitudes of the Buddhist tradition. Instead, we find two men who have grown close during their exchange of questions and answers, their virtues and excellences beginning to mirror one another’s. They are said to spend the night examining themselves, rehearsing their conversation. The next morning, the King is keen to convey that the felicity of mind (somanassa) with which he passed the night did not derive from the fact of having questioned the monk; rather, there are virtues beyond victory and defeat: “I was debating with myself as to whether I had asked felicitously, and whether I had been rightly answered” (cf. RD1, 135). Nāgasena concurs. To speak well together, and to examine oneself after, is what matters. On this ending, Milinda does not convert to Buddhism. If he changes at all, he is changed by coming into a cooperative relationship with Nāgasena based on the felicity of mind philosophy offers when exercised as a way of life.12

A More Intimate Drama Milinda’s conversion to the art of wise, cooperative reasoning is the more intimate drama and the text’s minor masterpiece. More specifically, Books II–III offer an intriguing response to the following narrative problem: how shall thinking be represented as a dramatic activity? *The Questions of Menander creates dramatic movement by invoking transitions in what I shall call the context, the mode, and the character of thought. The work leads its readers from an initially public and intrinsically confrontational mode of thought concerning the beliefs of 105

Sonam Kachru

individuals to a mode of thought that is private, dialogical and cooperative, and increasingly impersonal. Here I  mean two things by “impersonal”: thinking, as the text has it, involves being able to master a vocabulary and way of framing our experience on which (a) there is no reference to selves or what belongs to a self; and (b) what is offered can be used to bring any individual under a description and not only biographically individuated persons. At the end of Book III, Milinda wishes to lavish gifts on Nāgasena, but Nāgasena refuses. The King reminds Nāgasena that they are constrained by public opinion. Milinda and Nāgasena have been speaking oriented by their desire to arrive at truth. They have done so in the King’s chambers, beyond the reach of the public. But the public still exists as a horizon of possible relevance, and as an available way of reframing the conversation – as an exchange of goods, for example, or as earlier, at the very beginning of the text, when the conversation began in public and was framed as a public contest and trial. It is only after Nāgasena successfully defeats Milinda in a public debate that the conversation changes (see Vasil’kov 1993). The next day, they choose to speak in private and not public; Nāgasena insists that they continue to speak only as pandits and not as Kings.13 Among other things, this means the following: by holding different opinions on a topic one may cooperate with and not only challenge another; one may reason not only to win an argument but to get at truth; an argument need not entail contestation; and a new model of arguments exchanged in conversation may involve a process of trying to expose error in one another’s positions, with such exposure being productive rather than threatening and to be met with acknowledgment and not anger (RD1, 46). Outside the constraints of unreasonable power, conversation, analysis, and the mutual exploration of possible positions may lead to truth. Such cooperation need not eschew hierarchy. As Milinda puts it at the end, they have reasoned together as master and student to get at the truth (RD1, 88). Where may such intimate conversations of truth unconstrained by power and fear be held? At the beginning, Menander suggests speaking as one might with a brother, a novice, a lay disciple, or a servant (RD1, 88). It is, perhaps, into some such intimate and edifying theater that the text invites the reader, as well. The mode of conversation shifts, as does the topic. We move from discursive frames, like our conventional ways of speaking, which presuppose selves, to frames which do not. More generally, the conversation involves Milinda making himself more at home with the existential context for our lives provided by Buddhism in its understanding of rebirth, saṃsāra, and time and making himself at home with the virtues emphasized by Buddhists as a means to freedom within this context. If the going gets tough here, and the questions too numerous and too disparate, one may hold to the following dramatic thread: Milinda’s continuing trouble with both aspects of impersonalism. Consider his question about the self as a principle of cognizance or awareness, for which he uses the Pāli word vedagu (RD1, 86).14 Apparently unconvinced by the impersonal vocabulary for experience Nāgasena has been laying out and in an attempt to motivate an idea of a self as perceiver, Milinda helps himself to a striking illustration: Imagine a person in a house with many windows. The self might be like such a person: as one might see using the windows, the self sees, or hears, and the like, with the relevant sense capacity. One might think that Milinda has learned something from Nāgasena about the use of illustrations. But perhaps he has not learned the lesson perfectly. Nāgasena uses Milinda’s illustration to highlight an important disanalogy with the case of sensory experience. Experience might be analogous to a person in a house with many windows were something sensible an object like the King’s manservant, for example, something capable of coming in and out of view: something, that is, that might be (a) reidentified over time (b) through any window equally and (c) available to us to be aware of wherever they (and we) are located. But this, 106

The Milindapañha

Nāgasena believes, is not so. Sensible things are too tightly individuated to their senses to motivate the sense of Milinda’s autonomous perceiver and the content-indifferent windows. This argument needs more careful study. It is strikingly reminiscent of Socrates’ thematically continuous discussion with Theaetetus on the difference it makes to think of the soul as seeing with in contrast to seeing through the senses (Burnyeat 1976).15 Whether in Greek or Pāli, impersonalism can impose changes to the way we speak and think about ourselves. These changes are difficult for Milinda to fully accept. As he effectively admits in his next question (RD1, 89–92), what he is having difficulty with may be put this way. He is accustomed to speaking and thinking about experience in terms of sentences structured by appeal to actions and their agents. Even when he withholds speaking of a self, he still thinks of events as actions brought about by agents, albeit insentient ones. Nāgasena, in attempting to teach him Buddhist thought (Abhidhamma), is trying to teach him an entirely new conceptual grammar, featuring causal and structural explanations rather than agential ones. There is another source for the difficulty of impersonalism based on a more familiar sense of “personal.” As he is learning to think in this new impersonal idiom for exploring experience, Milinda is not content to learn what may hold true of anyone’s experience; he wants to know about Nāgasena in particular. Why did Nāgasena, rather than anyone else, leave home? (II.1.5, RD1, 49). It is a pattern of concern that does not leave him. Later, Milinda happens to raise again a question which he had asked sometime earlier in the conversation. Tellingly, it is a highly personal question: “Will you, Nāgasena, be born again?” Milinda waits to ask his question again just after a lengthy exposition of an impersonal (in the sense of “without self ”) account of rebirth and continuity. Nāgasena has to rebuke him into dropping the matter for good (II.2.7; see I, 6; RD1, 76–77; 50). The incident is, to me, suggestive. Milinda asks some fifteen questions before raising again the one that so obviously matters to him. As with music, the silences between questions, and what is not said in them, can be as significant as what is said.

The Lion in a Gilded Cage Shifts in the context, mode, and character of conversation, I have suggested, are dramatically significant. They allow us to track what it is like to begin thinking with – and as – a Buddhist about the world as Buddhism describes it. A later literary critic on the subcontinent might ask: what aesthetic savor does such drama sustain? A variety of tragic bittersweetness, I would venture to say. Unlike The Questions of Milinda, *The Questions of Menander is not triumphal. On the night before the formal conclusion of their private dialogue, when Nāgasena and Milinda agree in their assessment of the virtues in play in philosophical conversation and acknowledge the need to take into account the wider social world to which they must eventually return, the narrative has Milinda say this (3.7.18; cf. RD1, 135; see also Sick 2007, 276): Just as . . . a lion, the king of beasts, trapped in a golden cage is ever staring outward, I . . . too, though I dwell on the inside, ever stare outward. Philosophy need not always result in freedom. For some, it can result in a sense of imprisonment. Let me elucidate the poetic argument that is suggested by the image of the lion (also found in the Chinese work). An inscription in Kharoṣṭhi found in Tillya Tepe, Afghanistan, refers to “the lion who has driven away fear,” perhaps meaning the Buddha and his first sermon (see DeCaroli 2015, 20). One might also expect it to be a symbol of royalty, as it was for the Mauryan Aśoka, as well as for Menander’s descendants: the seated lion is found on the Indian-standard square coins of the apparently 107

Sonam Kachru

Buddhist Menander II (90–85 BCE).16 In Pāli, however, the King’s first glimpse of Nāgasena has the latter, quixotically enough for a bald monk, appear “like a shaggy lion who knows no fear” (RD1, 39). I recommend reading Milinda’s image as an allusion to a popular Buddhist text which exhorts the celibate ascetic to “not tremble, as a lion does not tremble at sounds, not caught up with others, as the wind is not caught in a net” (verse 71, Suttanipāta, in Norman 2001, 9; see also Salomon 2000, 108; 38). The text continues: “Living victorious, having overcome like a strong-toothed lion, the king of beasts, one should resort to secluded dwellings” (verse 72, Suttanipāta, in Norman 2001, 9). Given this background, Milinda’s use of the image is surely ironic: the lion, unlike the wind, now finds himself caged “inside” his own quotidian life. Redescribing oneself with such commitments one has acquired through reasoning in conversation, but without a corresponding change in one’s way of life, is not straightforward. If it is philosophical reasoning that lets Milinda see through the bars of the cage to a broader vista of possibilities, the same reasons now led him to experience his life as a cage in the first place. When they begin their private dialogue, Nāgasena effectively explains to Milinda the difference between reasons and causes. A monk is said to leave home with nibbāna as the goal or reason, though individual monks have any number of causes that have led them to leave home, including, as Nāgasena wryly mentions, the tyranny of kings (RD1, 49–50). At the end of their conversation, Milinda cites political necessities as holding him back. Is this to cite a mere cause? Could one ever have normatively weighted reasons not to take up the way of life Nāgasena has worked so hard to render intelligible? If Milinda finds himself locked into a life for which he knows he has no justification, that might suggest a quiet tragedy.

Conclusion “[T]here are . . . passages in the early part of the Questions of King Milinda,” said H. H. Price, once Wykeham Professor of Logic in Oxford, “which have a very modern ring, and might almost have been written in Cambridge in the 1920s” (Price 1955, 229). For the most part, the attention of philosophers continues to be directed only at early moments in The Questions of Milinda, employing the same criterion of relevance: what topics might appear in the leading journals of the discipline today? This is too restrictive. There is much cutting-edge material to be found elsewhere in the text. But it is perhaps better to do away with the parochial criterion of relevance altogether and read more of the text and with greater sensitivity to the way in which it directs us to read it. If we incline to Milinda’s point of view, the work might seem to offer us a path from doubt through reason farther and farther into the imaginative world of Buddhism; from Nāgasena’s (rather less variable) vantage point, the work may be read as the enactment of faith through the victory of reason in service of Buddhism. But there is a danger in identifying too much with either character at the expense of the other. While the answers to difficult questions are important, so too is the art of questioning, which is why the Buddha in the prologue, predicting the conversations to come in the future, says that both Milinda and Nāgasena shall explain his teaching, unraveling and disentangling its difficulties through questions and illustrations (cf. RD1, 6). Another way to put this is to say that the true protagonist might be wisdom itself, and that the true stage for this dialogue might lie off the page. In a work dedicated to the skillful use of similes and images, it is fitting to consider the image with which the work concludes (cf. H2, 305; see also Drewes 2007, 105–6): Therefore, let one who is wise, beholding his own good, 108

The Milindapañha

reverence and honor those who have wisdom As a memorial ought to be honored. Notice the invitational quality: a reader is expected to exemplify wisdom in the sense of prudential reason, enough to keep in view wisdom – now meaning something like analytic skill – as something worthy of regard as a memorial housing the Buddha’s relics is worthy of regard, and for the same reason: honoring such a site brings merit. A memorial (cetiya) housing relics of the Buddha or other venerated teachers, whether in the literary universe of Pāli or Sanskrit, can trigger strong emotions and make available transformative experiences. From about the first century CE, by analogy, books containing scripture could be such memorials (Drewes 2007); any body (khandho), in fact, as the conclusion of The Questions of Milinda has it, might serve in this capacity, as long as it embodies wisdom. Thus, we are to treat the fictional characters who are wise – pandits embodying wisdom ( paññā) – and, presumably, the text with its wise depiction of their wisdom-exemplifying conversation, as being effectively memorial shrines. To honor and reverence a work this way implies a certain way of reading. I leave it to the reader to imagine ways in which one might read this text worshipfully, another possible accent of the verbal root pūj-, along with “honoring” and “reverencing.” To me, being asked to read a text as if venerating a memorial shrine suggests, at the very least, that the reader is being encouraged to enter into an intimate, affectively charged (and possibly long-term) transformative relationship with it.17 One might here recall the conclusion of *The Questions of Menander and the suggestion that analytic conversation must be accompanied by self-examination. To echo Nāgasena’s characterization of the dialogue of pandits, perhaps we are to read forward and backward, occupying now the role of Milinda/Menander, now that of Nāgasena, alternatively touching freedom and seeing our own gilded cages through which possible forms of life beckon.

Notes 1 The work is included within the Pāli canon in Burma as the eighteenth book of the Khuddaka Nikāya, the latter being a part of the Group of Scriptures (Sutta-piṭaka) and rendered either as The Short Group of [Canonical] Texts, or The Group of Short [Canonical] Texts; neither name, as Steven Collins wryly remarks, is accurate (Collins 2020, 2). In the introduction to the standard edition of the work, V. Trenckner records that in Sri Lanka, it is sometimes known as Milindapañho, which uses the masculine ending as collective singular; manuscripts typically refer to it as Milindapañham (using the neuter collective singular) or Milindapañhā, which is either plural or a collective singular feminine ending (Trenckner 1880, vi). 2 See Hall (2015, 200), citing Plutarch, Moralia, 21d-2c, who has Menander incorrectly be a King of Bactria (Stoneman 2019, 365); though see also Fussman (1993, 65) and Lamotte (1988, 421), quoted in Neelis (2011, 105, n117). On the connection with Hellenistic hero cults and not alone Buddhist relic cults, see Neelis (2011, 105), citing Burkert (1985, 203). 3 For students interested in such questions, I recommend now beginning with Chapters 12–13 in Stoneman (2019), supplemented with Baums (2017). This is not, however, to endorse every claim in Stoneman (2019). For example, I do not think it is right to say that The Questions of Milinda “contains nothing that would remind one of Greek philosophy” (Stoneman 2019, 366). 4 On translations into other European languages, see de Jong (1996, 382–83). Excitingly, a muchneeded new translation into English is being prepared by Maria Heim for the Murty Classical Library of India and Harvard University Press. 5 I also wish with this title to signal the likely provenance for the latter: this stretch of text – where the Pāli and the aforementioned Chinese work, The Sūtra of the Monk Nāgasena, overlap most – may well reflect materials held in common the second century C.E. (see Baums 2017, 33; possibly earlier, according to Nolot 1995, 9). 6 It is not merely, that is, a Buddhist iteration of the South Asian narrative penchant – one also found in late-antique Mediterranean genres (erotapokriseis or problemata; see Volgers and Zamagni 2004) – for organizing knowledge through questions and answers.

109

Sonam Kachru 7 In standard historiography, Jains and Buddhists present their views of action as offering contrasting perspectives of the nature of action and the part of action which ought to receive emphasis and weight. It is claimed by Buddhists that Jains emphasize what is done (by overt action), sometimes even at the expense of awareness or intent, while it is claimed by Jains that Buddhists downplay the overt action and what is done thereby. For rival perspectives on the same examples, see Jacobi (1895, 414–15); Pruden (1991, 650–51); the issue is raised as early as the Upāli Sutta (Ñāṇamoli and Bodhi 1995, 477–93). 8 Compare the example of the iron ball in the prudential advice offered in verse 371 of the Dhammapada (Norman 2000, 53). 9 This kind of conceptual variegation is only strengthened when one reads over the several cases in the work. See Rhys Davids (1890, Vol. I., 224; 1894, Vol. II., 78). See also McDermott (1977). My account is informed by the methodological commitments to context-sensitivity and resistance to generalization which Maria Heim has reconstructed for Buddhaghosa in Heim (2018, especially pp. 50, 149, 167). 10 The entire description of the seventh difficult thing to do requires careful analysis, well beyond the scope of our discussion. I should say that I am inclined to follow Rhys Davids and not Horner, and I am grateful to Maria Heim for helping me think through the possible readings, including those of the commentary. 11 A terminological reminder: Earlier, I called Books II–III *The Questions of Menander to highlight a philological and literary point. This stretch of text antedates the Pāli work, as it is found in different languages, and it is literarily distinctive. As we consider this part of the work, I shall use the name “Milinda” to say things that are true of the character of the King we find in the Pāli work as well as in the The Sūtra of the Monk Nāgasena, the work surviving in Chinese which overlaps with Books II–III. 12 A model of such friendship (and this style of conversation) may be found in the conversation of Sāriputta and Puṇṇa Mantāṇiputta, who speak employing similes and as “two great beings rejoining in each other’s good words” (Ñāṇamoli and Bodhi 1995, §14, 245). 13 There is canonical precedent for the implied contrast between truth and power as orienting values in conversation. In the Upāli Sutta (Majjhima Nikāya i.376), the Buddha asks the householder Upāli, someone keen on adversarial debate, whether or not he “will debate on the basis of truth,” given which “[they] might have some conversation on this” (Ñāṇamoli and Bodhi 1995, 481). The sense that some topics are best treated in private conversation and not public debate is found in the way that the narrative changes in the Bṛhadāraṇyaka-Upaniṣad 3.2.13, when Yājñavalkya takes Ārtabhāga aside to discuss karma. On thinking in conversation more generally, see the papers in Black and Patton (2016). 14 The word has an interesting history. In Brahmanical usage, it meant “one who knows the Vedas”; in Buddhist usage, it was reinterpreted to mean “One who has achieved knowledge of release from samsara.” See Norman (2012, 198). 15 Though Menander in Pāli uses the instrumental case to make his point – the self is “the living principle which sees hue with the visual sense (  jīvo cakkhunā rūpaṃ passata . . .)” – his appeal to the example of windows, and his subsequent explanation, suggest instead that he agrees with Socrates’ case for the soul seeing “through” the senses (see Burnyeat 1976, 32–33, 37). Nāgasena, meanwhile, adopts a position reminiscent of the Heraclitean flux of unowned sensations of transient proper sensibles against which Socrates sets himself. 16 Whitehead (1914, 62). See also the maneless lions of Pantaleon (Whitehead 1914, 16). 17 My thanks to Ishan Chakrabarti for conversation on this. Caley Smith has drawn my attention in conversation to a similarly metatextual moment in the Bhagavad Gītā (verse 18.70) when study or reading of certain texts can take the form of an epistemologically transformative variety of (inner) ritual, a jñāna-yajña, and result in new forms of relations to what is real. See Flood and Martin (2012, 190).

Bibliography Baums, Stefan. 2017. “Greek or Indian? The Questions of Menander and Onomastic Patterns in Early Gandhāra.” In Buddhism and Gandhara: An Archaeology of Museum Collections, edited by Himanshu Prabha Ray, 33–47. London: Routledge.

110

The Milindapañha Black, Brian, and Laurie Patton. 2016. Dialogue in Early South Asian Religions. London and New York: Routledge. Bopearachchi, Osmund. 1990. “Ménandre Sôter, un roi Indo-grec. Observations chronologiques et géographiques.” Studia Iranica 19: 39–85. ———. 1992. “Was Sagala Menander’s Capital?” In South Asian Archaeology 1989: Papers from the Tenth International Conference of South Asian Archaeologists in Western Europe, Musée national des Arts asiatiques – Guimet, Paris, 3–7 July 1989, edited by Catherine Jarrige, 327–37. Madison, WI: Prehistory Press. Burkert, Walter. 1985. Greek Religion. Trans. John Raffan. Cambridge, MA: Harvard University Press. Burnyeat, Myles. 1976. “Plato on the Grammar of Perceiving.” The Classical Quarterly 26 (1): 29–51. Collins, Steven. 1982. Selfless Persons: Imagery and Thought in Theravāda Buddhism. Cambridge: Cambridge University Press. ———. 2015. “Madness and Possession in Pāli Texts.” Buddhist Studies Review 31 (2): 195–214. ———, ed. 2016. Readings of the Vessantara Jātaka. New York: Columbia University Press. ———. 2020. Wisdom as a Way of Life: Theravāda Buddhism Reimagined. Edited by Justin McDaniel. New York: Columbia University Press. Cone, Margaret, and Richard F. Gombrich. 2010. The Perfect Generosity of Prince Vessantara: A Buddhist Epic. Bristol: Pali Text Society. Cowell, E. B., and W. H. D. Rouse. 1907. The Jataka, or Stories of the Buddhas Former Births, vol. 4. Cambridge: Cambridge University Press. DeCaroli, Robert. 2015. Image Problems: The Origin and Development of the Buddha’s Image in Early South Asia. Seattle and London: University of Washington Press. de Jong, J. W. 1996. “Review of Entrietens de Milinda et Nāgasena. Traduit du p‫‏‬âli, présenté et annoté.” Bulletin of the School of Oriental and African Studies 59 (2): 382–83. Demiéville, Paul. 1924. “Les versions chinoises du Milindapañha.” Bulletin de l’École Française d’Extrême Orient 24: 1–26. Drewes, David. 2007. “Revisiting the phrase ‘sa pṛthvīpradeśaś caityabhūto bhavet’ and the Mahāyāna cult of the book.” Indo-Iranian Journal 50 (2): 101–43. Eliot, T. S. 1919. “The Preacher as Artist.” Athenæum 4674: 1252–53. Flood, Gavin, and Charles Martin. 2012. The Bhagavad Gita. New York and London: W. W. Norton and Company. Fussman, Gérard. 1993. “L’Indo-grec Ménandre ou Paul Demiéville revisité.” Journal Asiatique 231: 61–138. Ganeri, Jonardon. 2012. Identity As Reasoned Choice: A South Asian Perspective on the Reach and Resources of Public and Practical Reason in Shaping Individual Identities. New York: Continuum International. Hall, Edith. 2015. Introducing The Ancient Greeks: From Bronze Age Seafarers to Navigators of the Western Mind. New York: W. W. Norton and Company. Heim, Maria. 2003. “The Aesthetics of Excess.” Journal of the American Academy of Religion 71 (3): 531–54. ———. 2018. Voice of the Buddha: Buddhaghosa On The Immeasurable Words. New York: Oxford University Press. Horner, I. B. 1963. Milinda’s Questions: Volume I. London: Luzac and Company. ———. 1964. Milinda’s Questions: Volume II. London: Luzac and Company. Jacobi, Hermann. 1895. Jaina Sūtras: Part II. Oxford: Clarendon Press. Kapstein, Matthew T. 2001. Reason’s Traces: Identity and Interpretation in Indian and Tibetan Buddhist Thought. Boston: Wisdom Publications. Lamotte, Étienne. 1988. History of Indian Buddhism: From the origins to the Śaka era. Translated by Sara Webb-Boin. Publications de l’Institut orientaliste de Louvain, 36. Louvain-la-Neuve: Université catholique de Louvain, Institut orientaliste. McDermott, James P. 1977. “Kamma in the Milindapañha.” Journal of the American Oriental Society 97 (4): 460–68. Ñāṇamoli, Bhikkhu, and Bhikkhu Bodhi, trans. 1995. Majjhima Nikāya. Boston: Wisdom Publications. Neelis, Jason. 2011. Early Buddhist Transmission and Trade Networks: Mobility and Exchange Within and Beyond the Northwestern Borderlands of South Asia. Leiden: Brill. Nolot, Édith, trans. 1995. Entrietens de Milinda et Nāgasena. Traduit du pâli, présenté et annoté. Paris: Galimard.

111

Sonam Kachru Norman, K. R. 1983. Pāli Literature: Including the Canonical Literature in Prakrit and Sanskrit of All the Hīnayāna Schools of Buddhism. Wiesbaden: Otto Harrassowitz. ———. 2000. The Word of the Doctrine (Dhammapada). Oxford: Pali Text Society. ———. 2001. The Group of Discourses (Sutta-nipāta). Oxford: Pali Text Society. ———. 2012. “Theravāda Buddhism and Brahmanical Hinduism: Brahmanical Terms in a Buddhist Guise.” The Buddhist Forum 11: 193–200. Pogge, Thomas. 2007. John Rawls: His Life and Theory of Justice. Translated by Michelle Kosch. New York: Oxford University Press. Price, H. H. 1955. “The Present Relations Between Eastern and Western Philosophy.” The Hibbert Journal 53: 222–29. Pruden, Leo M. 1991. Abhidharmakosabhasyam of Vasubandhu. Volume II. Berkeley: Asian Humanities Press. Rhys Davids, T. W. 1890. The Questions of King Milinda. Translated from the Pāli. Part I. Oxford: Clarendon Press. ———. 1891. “Nāgasena.” The Journal of the Royal Asiatic Society of Great Britain and Ireland (July): 476–78. ———. 1894. The Questions of King Milinda. Translated from the Pāli. Part II. Oxford: Clarendon Press. Salomon, Richard. 2000. A Gāndhārī Version of the Rhinoceros Sūtra: British Library Kharoṣṭhī Fragment 5B. Seattle and London: University of Washington Press. ———. 2018. The Buddhist Literature of Ancient Gandhāra: An Introduction with Selected Translations. Somerville, MA: Wisdom Publications. Scanlon, T. M. 2003. “Thickness and Theory.” Journal of Philosophy 100 (6): 275–87. Sick, David H. 2007. “When Socrates Met The Buddha: Greek and Indian Dialectic in Hellenistic Bactria and India.” Journal of the Royal Asiatic Society 17 (3): 253–78. Skilling, Peter. 1998. “A Note on King Milinda in the Abhidharmakośabhāṣya.” Journal of the Pāli Text Society 24: 81–101. Sripada, Chandra. 2016. “Self-Expression: A Deep Self Theory of Moral Responsibility.” Philosophical Studies 173 (5): 1203–32. Stoneman, Richard. 2019. The Greek Experience of India. From Alexander to the Indo-Greeks. Princeton and Oxford: Princeton University Press. Tarn, W. W. 1951. The Greeks in Bactria and India. Cambridge: Cambridge University Press. Trenckner, V. 1880. Milindapañha. London: Pali Text Society. Vasil’kov, Y. 1993. “Did East and West Really Meet in Milinda’s Questions?” Petersburg Journal of Cultural Studies 1: 62–77. Volgers, Annelie, and Claudio Zamagni, eds. 2004. Erotapokriseis: Early Christian Question and Answer Literature in Context. Leuven: Peeters. Von Hinüber, Oskar. 2000. A Handbook of Pāli Literature. Berlin and New York: Walter de Gruyter. Whitehead, R. B. 1914. Catalogue of Coins in the Panjab Museum, Lahore. Vol. I: Indo-Greek Coins. Oxford: Clarendon Press.

112

7 CĀTTAN ‒ ĀR

Poet-Philosopher in Tamiḻ Anne E. Monius1

Who Is Cāttan‒ār?

Since the rediscovery and first publication of his only known work, The Renunciation of Maṇimēkalai (Maṇimēkalai Tuṟavu),2 in 1898, Cāttaṉār3 has continued to generate intense debate among scholars of South Indian Tamiḻ literature, history, and religion. Who was he? When and where exactly did he live? What manner of Buddhist was he, and what texts did he know and draw upon in composing this long poetic narrative about a beautiful young courtesan, Maṇimēkalai, who becomes a Buddhist nun? Embedded in the story are three densely worded philosophical chapters in verse: one chapter is devoted to ten non-Buddhist teachings, another is on Buddhist inferential reasoning, and the final chapter provides a detailed account of interdependent origination. What purpose do these chapters serve? Do they reveal anything of the author’s own particularly Buddhist inclinations in a Tamiḻ-speaking region of southern India where the history of Buddhist thought and practice remains obscure? In the century or more since The Renunciation of Maṇimēkalai became widely known as a classic work of Tamiḻ literature and one of the earliest long narratives composed in Tamiḻ, knowledge of its author, Cāttaṉār, has advanced little. All that can be said of him with any certainty derives from the sole text that he has left to us, alongside several brief mentions in another long Tamiḻ narrative, The Story of the Anklet (Cilappatikāram), with which The Renunciation of Maṇimēkalai has long been aligned. As scholar Kamil V. Zvelebil aptly summarizes, “the only certain thing” that can be said of Cāttaṉār “is that he was a convinced and well-educated Buddhist and a great poet” (1995, 128). Cāttaṉār’s name first appears in the preface to The Story of the Anklet, the earliest longform narrative in Tamiḻ, that tells the story of a young married couple, Kōvalaṉ and Kaṇṇaki. Kaṇṇaki weathers adultery, hardship, and exhausting travel, and her husband is unjustly put to death by a local king as he attempts to sell his wife’s anklet, their sole remaining valuable. In response, the ever-faithful Kaṇṇaki tears off her breast, hurls it at the royal city of Maturai (which promptly bursts into flames), and ascends into heaven as a goddess. Dating of the text continues to be debated, but a scholarly consensus has emerged that places the narrative in the fourth or fifth century. While the preface names one Iḷaṅkō Aṭikaḷ as the author of the text, Iḷaṅkō is said to compose the story at the specific request of “Cāttaṉ, a grain merchant from Maturai” (preface, lines 88–9),4 who has already been described as “Cāttaṉ of graceful Tamiḻ” DOI: 10.4324/9781351030908-12

113

Anne E. Monius

(line 10). Cāttaṉ tells Iḷaṅkō of the events he has witnessed, and then listens as Iḷaṅkō narrates the story back to him “in poetry and song” (line 60). Cāttaṉ appears once more in the text, as he is called again to describe the events he has witnessed; here he is described as “the master of graceful Tamiḻ” (25.66). Cāttaṉ, as a character in The Story of the Anklet, emerges as an observant poet who witnesses extraordinary events first-hand, yet humbly yields the telling of the narrative to his companion, Iḷaṅkō. The preface to The Renunciation of Maṇimēkalai names Cāttaṉ as the author only once, again pairing him with Iḷaṅkō but with their respective roles reversed. The dating of The Renunciation of Maṇimēkalai is no more certain than that of The Story of the Anklet, although continuing the story of minor characters in the first text perhaps suggests, at the very least, a somewhat later date for Cāttaṉār’s work; a rough scholarly consensus has tentatively settled on the sixth or seventh century. This time it is Iḷaṅkō – here called a “king” – who requests a story from Cāttaṉ, described as “a prosperous grain merchant,” who tells the story “of the renunciation of Maṇimēkalai in thirty songs of eloquent Tamiḻ” (preface, lines 95–8).5 Beyond these sparse literary references,6 all that can be said of the historical Cāttaṉār must be gleaned from the beautifully written poetic narrative that he authored, The Renunciation of Maṇimēkalai. Continuing the story of minor characters in the earlier narrative, The Story of the Anklet (where Maṇimēkalai is the daughter of Kōvalaṉ and his mistress), the complex narrative is told in 4,758 lines arranged in thirty chapters, with sixteen “branch stories” or subplots (Richman 1988, 2–3) woven into the primary story. With its technical terminology; references to a wide range of Sanskrit, Pāli, and Tamiḻ literary sources; and dense philosophical material, Cāttaṉār clearly anticipates a sophisticated Buddhist audience for his text, well-versed in multiple poetic and philosophical traditions (see Monius 2001, 16–18).

The Basic Story of The Renunciation of Man imēkalai7 Cāttaṉār’s narrative opens in the coastal city of Pukār. Mātavi, Kōvalaṉ’s former lover and mother of Maṇimēkalai, has just heard the news of her ex-lover’s tragic end and has renounced her former life as a courtesan to follow the Buddhist teacher, Aṟavaṇaṉ; she has vowed that her young daughter will do the same. Sent out with a friend to gather flowers for a garland, Maṇimēkalai encounters a young prince of the city, who immediately decides that he must possess her. The heroine’s namesake goddess, Maṇimēkalā, whisks the girl off to the island of Maṇipallavam to save her from the prince’s advances; there Maṇimēkalai encounters a magnificently bejeweled seat of the Buddha, and she begins to realize her previous karmic trajectory that has brought her to the present-tense narrative. The goddess reappears and grants the girl three mantras that enable her to fly, to change physical form at will, and to appease her own hunger whenever necessary. Maṇimēkalai then receives a miraculous begging bowl from a mysterious lake on the island, one that never empties of food if used to feed the poor. Maṇimēkalai then returns to the city of Pukār to be reunited with her mother and friends. They meet the Buddhist teacher, Aṟavaṇaṉ, and he proceeds to tell his young disciple the history of her miraculous alms bowl. Maṇimēkalai then feeds the hungry masses of Pukār with the bowl. The prince, of course, has not forgotten the object of his desire, and he resumes his pursuit. In order to avoid the prince’s amorous advances, Maṇimēkalai deploys the power of her mantra to change form; the prince is then killed in a case of mistaken identity. Maṇimēkalai is then imprisoned for her part in the crime. The queen attempts to have her murdered, but the power of Maṇimēkalai’s mantras keeps her safe from harm. Aṟavaṇaṉ briefly relays the basics of interdependent origination before Maṇimēkalai meets the previous holder of the miraculous alms bowl. She then proceeds to the western capital of Vañci, where she visits the temple 114

Cāttan ‒ ār

dedicated to the heroes of The Story of the Anklet, Kōvalaṉ (her father) and his virtuous wife, Kaṇṇaki. She assumes the form of a male ascetic and listens to ten non-Buddhist teachers discuss their basic doctrines. Maṇimēkalai then travels east to Kāñci and, at the request of her grandfather, commands the king to build a replica of Maṇipallavam there in recognition of the power of the magical alms bowl. In the penultimate chapter, Aṟavaṇaṉ teaches proper inferential reasoning to Maṇimēkalai, and, in the final chapter, interdependent origination. Finally, in the last two lines of the text, Maṇimēkalai, “having heard the dharma, took up ascetic practice, and vowed to eradicate the karmic effects of birth” (30.263–4).

Cāttanār’s Buddhist Philosophy ¯

Coming at the very end of the narrative, the two final chapters of The Renunciation of Maṇimēkalai – on proper inferential reasoning (chapter twenty-nine) and interdependent origination (chapter thirty) – are, by position, a culmination of sorts, serving as the immediate precursors to the heroine’s final vow of renunciation. Both chapters have received significant amounts of scholarly attention over the past century, largely in an effort to place Cāttaṉār historically and in terms of doctrinal affiliation more precisely.

Buddhist Logic in The Renunciation of Manimēkalai The penultimate chapter of Cāttaṉār’s The Renunciation of Maṇimēkalai contains the most extensive discussion of inferential reasoning in Tamiḻ literature. Maṇimēkalai’s teacher, Aṟavaṇaṉ, begins his discussion of Buddhist logic with the assertion that the Buddha accepts only two valid means of knowledge: direct perception ( pirattiyam) and inference (karuttaḷavu) (29.47–48). After a very brief treatment of direct perception, Aṟavaṇaṉ states that “all other valid means of knowing properly belong to inference” (55–6). He then moves on to propose a five-limbed syllogism, using the common fire-on-the-mountain example (57–63). The proposition (  pakkam) is that “the mountain possesses fire.” The reason (ētu) is “because it possesses smoke.” “Just like a kitchen” is the example (tiṭṭāntam). The application (upanayam) is “and the mountain also possesses smoke.” The conclusion (nigamaṉam) is thus “because the mountain possesses smoke, it also possesses fire.” Aṟavaṇaṉ then proceeds to critique this five-part syllogism, ultimately arguing that “application and conclusion are connected to the example and are thus included in it” (109–10). For the remainder of his discourse on inference, Aṟavaṇaṉ focuses on the proper working of a three-part inferential reasoning process, addressing himself first to the three valid forms of proposition (113–21), three valid forms of reason (121–35), and two valid forms of example (136–41), then turning to address the nine invalid forms of proposition (147–90), fourteen types of invalid reason (191–325), and twelve varieties of invalid example (325–468). From the sheer number of verses devoted to each topic, it would seem that Cāttaṉār particularly concerns himself with guarding against faulty inferential reasoning, enumerating the varieties of incorrect proposition, reason, and example, and providing brief examples of each. A significant amount of scholarship to date has considered the system of Buddhist inferential logic explicated by Aṟavaṇaṉ in its potential relationship to other Buddhist traditions of logic, initially spurred by the recovery and publication of the Sanskrit Primer on Logic (Nyāyapraveśa), a text originally attributed to Dignāga (whom many believed hailed from Kāñcipuram, in what is now the Indian state of Tamil Nadu) in the early twentieth century (1927–1930), but now largely viewed as the work of his disciple, Śaṅkarasvāmin. Kandaswamy (1978, 253–312) correctly summarizes the many decades of research into Cāttaṉār’s possible 115

Anne E. Monius

Sanskrit sources for his twenty-ninth chapter on Buddhist inferential reasoning by describing the author as “more an interpreter than a mere translator of Nyayapravesa” (287). Kandswamy locates what he rightly terms “minor” discrepancies between the Tamiḻ and Sanskrit texts (286–90): a slight variation in the order of the fallacies of proposition (286–87), two original subcategories of fallacious example authored by Cāttaṉār (287), a host of original examples provided in the Tamiḻ text (287–88, 291–305), and some variation in the adoption of Sanskrit philosophical terms into Tamiḻ (288–90). Hikosaka (1989, 121–52), like Kandaswamy, devotes a full chapter to discussion of the Tamiḻ narrative’s presentation of inferential reasoning, noticing major similarities and minor differences between The Renunciation of Maṇimēkalai and the Primer on Logic. Overall, Hikosaka concludes that Cāttaṉār provides more of an introduction and less of a conclusion than the Primer on Logic, and that the differences in terminology and number suggest that Cāttaṉār’s reference to this particular Sanskrit text is only partial, and “that he should have used some other sources also” (135). Nearly a century of scholarship on Cāttaṉār’s earliest presentation of formal Buddhist logic in Tamiḻ demonstrates an affinity to Śaṅkarasvāmin’s Primer, yet what marks the Tamiḻ text’s wholly unique contribution to Buddhist logic is its form, its embedding of dense philosophical discourse into highly stylized poetic narrative (see in what follows).

Interdependent Origination in The Renunciation of Man·imēkalai

Cāttaṉār’s final chapter, “The Story of the Young Woman’s Renunciation in Order to Eradicate the Qualities of Existence,” features Aṟavaṇaṉ’s teachings to his young student on the nature of human existence, emphasizing the twelve-part chain of interdependent origination, with the links in the chain (nidāna) here described in Tamiḻ as “appearance of dependent things” (carpiṉ tōṉṟu) (30.17). Aṟavaṇaṉ emphasizes that “the cause of undying liberation is realizing” that “sorrow is impermanent, caused by karma and the fruits of that karma” (29–31). The twelve dependent links in the chain of causation “have no creator, nor have they been created” (39), but are rather “dependent” entirely upon “karma and its results” (39). Aṟavaṇaṉ then describes each of the twelve “dependent things” in turn, from ignorance ( pētaimai) (51–81), awareness (uṇarvu) (82–3), name and form (aruvu uruvu) (84–5), six gateways (vāyil āṟu) (86–7), sense contact (uṟu) (88), sensation (nukarvu) (89–90), craving (vēṭkai) (91), and attachment ( paṟṟu) (92), to becoming ( pavam) (93–4), birth ( piṟappu) (95–7), old age (mūppu) (98–101), and death (cākkāṭu) (102–3). Each arises in a chain of dependence (104–18) – a “circle of existence without end” (118) – that can only be undone, step by step, through the cessation of each of the links in turn, beginning with ignorance (119–33). Particular attention (thirty-one verses: 51–81) is paid by Aṟavaṇaṉ to explaining the nature of ignorance, glossed in terms of lack of awareness of the chain of interdependent origination, of “forgetting what is seen through natural reason” (53), and lack of wisdom to discern good karma (nal viṉai) from bad karma (tī viṉai). Having explained the basics of interdependent origination, Aṟavaṇaṉ then relates to Maṇimēkalai the same teaching in terms of four groupings of the links in the chain of causation and three junctures among them (134–52); three categories of birth (153–58); the three temporalities of past, present, and future (159–68); and the working out of karma (169–75). Cāttaṉār, through his learned Buddhist character, then briefly explains the Four Noble Truths (176–88), the five aggregates (189–90), six types of designation (191–215), four methods (nayam) (217–34), and four forms of question and answer (235–49). He concludes by affirming that “there is no one who fixes either bondage or liberation” (250–51), only the elements ( poruḷ ) in the chain of causation that he has just outlined. His final call to his young disciple – to contemplate what he has just taught and thus to eradicate hatred (ceṟṟam), 116

Cāttan ‒ ār

delusion (mayakkam), and darkness of mind (maṉattu iruḷ ) (256–60) – immediately results in Maṇimēkalai’s final act of renunciation (263–4). While Cāttaṉār clearly displays a wide knowledge of Buddhist forms of inferential reasoning in Sanskrit in his previous chapter, he here in the final chapter follows far more closely Theravāda-aligned traditions in outlining each of the topics described earlier. As with the twenty-ninth chapter, many scholars have sought potential sources for Cāttaṉār’s final philosophical discussion. Overall, Cāttaṉār appears to follow most closely extant Pāli sources, particularly Buddhaghosa’s Path of Purification (Visuddhimagga), and particularly the seventeenth chapter entitled “Explanation of the Soil of Wisdom” (Paññābhūminiddesa). Cāttaṉār’s organization of the twelve links in the cycle of interdependent origination, for example, follows closely Buddhaghosa’s discussion of the same in the Path of Purification (17.289–90), as does the Tamiḻ discussion of the junctures among the groupings, temporality (284–87), the Four Noble Truths (300), and the four methods (309–13). As in the chapter on inferential reasoning, however, Cāttaṉār’s presentation contains unique definitions and details, and lies wholly embedded in a long and complex poetic narrative.

Philosophy Embedded in Poetic Narrative No matter what his specific sources might have been, Cāttaṉār’s particular contribution to the history and study of Buddhist philosophy lies less in the content of his philosophical chapters and far more in the form that each takes and the narrative purposes that each serves. Cāttaṉār organizes his narrative in such a way that the story itself leads naturally to these final two sets of Buddhist teachings. Long before the final chapter, to cite but one brief example, Aṟavaṇaṉ gives a short foretaste of his later teaching on interdependent origination as he counsels the young girl’s mother on the true nature of reality (24.105–40). The gradual transformation of Maṇimēkalai’s character, to cite another example, culminates naturally in the final two chapters of the narrative that mark the final removal of, as previously, “hatred, delusion, and darkness of mind.” Without the final two chapters, the young woman’s story would be incomplete; she would not be ready to renounce the world. In its complex plot that interweaves the main narrative and multiple branching stories, The Renunciation of Maṇimēkalai slowly both reveals and embodies its sophisticated concept of karma in the context of interdependent origination, revealing an intricate pattern of interlocking lives, deeds, and events. Cāttaṉār spins a narrative web filled with characters, events, and imagery that compels its audience to confront and reflect upon both the ultimate transience and profound interconnectedness that he describes theoretically only in the final chapter. In similar fashion, a long story that Aṟavaṇaṉ relays to Maṇimēkalai before he speaks of formal logic in chapter twenty-nine – of the goddess, Maṇimēkalā, and her dramatic rescue of one of Kōvalaṉ’s ancestors as the reason for the young girl’s name – gestures toward the overall logic in the text of ripening karmic conditions and renunciation, of past deeds coming to full fruition in the future: “On that day of your birth, the goddess, Maṇimēkalā, with her body of soft petals, spoke to your father in a dream, as if he were awake, about your resolve to renounce” (29.31–3). In such a narrative context, the long and dense discussions of proper inferential reasoning and interdependent origination are especially significant.

Philosophy as Narrative Critical to full understanding of the philosophical chapters in Cāttaṉār’s The Renunciation of Maṇimēkalai is a careful consideration of the role that those chapters play in the larger narrative that each serves. As the story of a beautiful young courtesan’s journey toward her final act 117

Anne E. Monius

of renouncing worldly life, the first twenty-eight chapters of Cāttaṉār’s text narrate the eponymous heroine’s gradual awakening to the truth of the Buddha’s dharma and the challenges she faces (primarily from the infatuated prince and his royal parents) as she turns away from the lifestyle of a courtesan and embraces the Buddhist path. Two themes dominate the overall narrative: the working out of karma in human lives and the unfolding of those conditions that lead the primary character, Maṇimēkalai, to renounce in the final lines of the text. Cāttaṉār’s text repeatedly invokes karma to explain events and circumstances in which various characters find themselves. Yet karma in The Renunciation of Maṇimēkalai clearly operates in a very different way than in Iḷaṅkō’s earlier narrative, The Story of the Anklet. Early in Cāttaṉār’s text, the young prince, enchanted by a glimpse of the beautiful Maṇimēkalai, is told that “the body is the result of karma and the cause of future karma” (4.113). Later, Maṇimēkalai attributes her fear and loneliness as she finds herself alone on the isolated island of Maṇipallavam to the same bad karma (ve viṉai) that led to her father’s earlier execution (8.40–3). The violent death of the infatuated prince in hot pursuit of Maṇimēkalai is explained to his father, the king of Pukār, as a result of bad karma (tī viṉai), and karma plays a dominant role in the sequence of actions that leads up to the moment of the murder (22.193–203). Karma in Cāttaṉār’s narrative becomes a powerful force at work in a complex web of conditions generated by the processes of interdependent origination (that, as previously, are spelled out explicitly only in the text’s final chapter). The logic of karma can perhaps be seen most clearly in Cāttaṉār’s unique treatment of the two main characters in the earlier narrative, The Story of the Anklet: Kōvalaṉ and Kaṇṇaki. Kaṇṇaki, for example – eventually emerging as a great goddess in the earlier narrative – in Cāttaṉār’s hands becomes simply another karma-accumulating individual who “will return to the ocean of births and be caught up in the cycles of birth and death” after enjoying the fruits of her good karma in heaven (26.40–1). Kaṇṇaki and Kōvalaṉ will be reunited on earth to live out their remaining karma until the Buddha comes again, performing acts of kindness to all for a long time (26.54–61). In The Renunciation of Maṇimēkalai, the logic of karma across past, present, and future requires a far more complex narrative structure than in The Story of the Anklet, including flashbacks and episodes that combine past, present, and future actions and a far greater number of characters and subplots. The earlier narrative’s straightforward A-in-the-past–yields–B-in-the-present causal sequences are replaced in The Renunciation of Maṇimēkalai by an intricate tapestry of characters, narratives and subnarratives, and scenes that suggest: A yields B, provided that C, D, E, and F are present to provide the proper environment in which both A and B can come to full fruition. Another dominant theme that links the many plots and subplots of The Renunciation of Maṇimēkalai builds upon the logic of karma noted previously and ties the flow of the narrative to the text’s dense concluding chapter on interdependent origination and the heroine’s final act of renunciation. Signaled by the “causes” or “conditions” (ētu, from Sanskrit or Pāli hetu), in combination with verbal forms such as “blossoming” (etirtal), “maturing” (murtital), or “manifesting” (nikaḻtal) – literally a blossoming, maturation, or manifestation of the beneficial root conditions that lead to enlightenment – The Renunciation of Maṇimēkalai structures its complex web of narratives around its central character’s karmic ripening, around Maṇimēkalai’s growing ability to hear and truly comprehend the Buddha’s teachings (Monius 2001, 24–32). Although the Tamiḻ narrative never defines what the “good” conditions (nal ētu) are exactly, wider Theravāda Buddhist tradition holds the primary beneficial “roots” (Pāli kusalamūla) to be absence of attachment (alobha), enmity (adosa), and delusion (amoha) (Buddhaghosa 1969, 17.66–70). Cāttaṉār grants the manifestation of beneficial root conditions (ētunikaḻcci) tremendous agency throughout the story he tells, beginning with the introduction of his heroine. Overhearing her mother, Mātavi, tell a friend about the violent death of her former lover, Kōvalaṉ, Maṇimēkalai sheds 118

Cāttan ‒ ār

fateful tears that propel her outside her home and spark the main events of the narrative to come: “Because there had blossomed (etirntu uḷatu) for Maṇimēkalai a manifestation of beneficial conditions (ētunikaḻcci) – like the fragrance emanating from a flower – she felt tremendous sorrow” (3.2–4). Later in the story, Manimēkalā (the goddess) connects the appearance of these beneficial conditions with the young woman’s future commitment to the Buddhist path: “Because the conditions (ētu) that lead to the path of dharma of the First Sage have matured (mutirntatu),” she goes on to explain that she has whisked Maṇimēkalai off to an isolated island to save her from the prince’s amorous advances (7.19–22). As the young heroine moves toward her final destination of Kāñcipuram to learn of logic and interdependent origination from her teacher before renouncing the world, references to the arising of beneficial conditions shift significantly to the future. Twice, characters predict what will happen in Kāñci to result in Maṇimēkalai’s final act: “All the manifestations of the beneficial conditions (ētunikaḻcci yāvum) will occur there. They are many” (12.104–7 and 21.155–58). Immediately prior to her renunciation, Maṇimēkalai’s Buddhist teacher, Aṟavaṇaṉ – having just explained to her the doctrine of interdependent origination  – commands his student to eradicate those conditions (ētu) that are unwholesome or non-beneficial: attachment ( paṟṟu), enmity (ceṟṟam), and delusion (mayakkam) (30.254–260). After this mandate from her teacher, now with all the beneficial conditions in place for her to advance along the path toward enlightenment, only then can Maṇimēkalai make her final vow of renunciation “to eradicate the karmic effects of birth” (30.264). The manifestation of beneficial conditions – and the final eradication of the non-beneficial – explains the transformation of Maṇimēkalai’s karmic state as witnessed throughout the narrative. If the themes of karma and the manifestation of the beneficial conditions leading to the heroine’s renunciation and enlightenment tie the larger narrative to the philosophical discussions of Cāttaṉār’s final two chapters, then his twenty-seventh chapter speaks directly to the topics that those discussions address. The narrative presentation of ten distinct schools of non-Buddhist philosophical and religious thought, presented to Maṇimēkalai shortly after she arrives in Vañci, might, on the one hand, simply be read as a catalog of philosophical thinking present in sixth- or seventh-century southern India, a narrative cousin of sorts to Mādhava’s fourteenth-century Collection of All Philosophies (Sarvadarśanasaṃgraha). Yet a careful reading of the presentation of each non-Buddhist school of thought reveals the full extent to which Cāttaṉār carefully shapes what each teacher says – primarily focusing on logic, the role of divine intervention in human life, and the constituent elements of the universe – to pave the way for the “correct” Buddhist teachings on similar topics in the final two chapters of the text. For example, Maṇimēkalai first encounters in Vañci a Vedic logician (aḷavai vāti) (27.3). Citing Vedavyāsa (Vētaviyātaṉ), Kirutakōṭi, and Jaimini (Caimiṉi) (5–6), the logician enumerates ten valid means of knowledge and proceeds to explain each in some detail. Of the second, inference (karuttaḷavu), he lists three types; each is defined and an example provided. He concludes his discussion of inference by noting that inferential reasoning works validly across the three times of past, present, and future (38). The remaining modes of valid knowledge are then discussed in turn, with brief examples provided, followed by a short discussion of eight fallacies ( piramāṇāpācaṅkaḷ ) (57–77). The Vedic logician then concludes his presentation by naming six distinct religious-philosophical traditions (Lokāyata, Buddhist, Sāṃkhya, Naiyāyika, Vaiśeṣikha, and Mīmāṃsā) in terms of the modes of knowing that each holds to be valid (78–85). Maṇimēkalai simply hears all of these and departs, immediately encountering a worshipper of Śiva who boldly declares, “Īcaṉ is lord!” (86). No comment is offered on the validity – or lack thereof – of the presentation of any non-Buddhist teacher until the young woman enters Kāñcipuram to sit at the feet of her Buddhist teacher, Aṟavaṇaṉ. To him, she describes all the doctrines of other communities learned in Vañci as being “the negation of what is 119

Anne E. Monius

good” (29.44), and she pleads with the respected teacher to impart to her “the true doctrine” (45). What follows is the lengthy discussion of proper and improper inferential reasoning, culminating in Aṟavanaṉ’s confident assertion that “through these proper methods” of inferential reasoning just described to her, she will “know what to accept as truth and as falsehood without any doubt” (471–72). The Śaiva’s confident assertion of his lord that “his appearance removes sorrow” (27.93) and the Vaiṣṇava’s equally confident claim that “Nāraṇaṉ is our protection” (99) similarly provide the salient point of departure for the final chapter’s discussion of interdependent origination in which “no lord assigns bondage and liberation” (30.250–1). The presentations of five teachers – the Ājīvaka, the Jain, the Sāṃkhyan, the Vaiśeṣikan, and the materialist – all focus on the nature of substance ( poruḷ ), providing the perfect foil for the final chapter’s discussion of all things ( poruḷ ) as “dependent” (cārpu). To all the Vañci teachers, Maṇimēkalai responds simply by laughing (27.280), as the proper Buddhist understanding of all the doctrines taught is yet to come.

Philosophy as Poetry Not only does Cāttaṉār embed his philosophical chapters in a complex narrative of which they form an integral part, as described previously, but he also composes both chapters – as he does the entire text – in sophisticated poetic form. The text describes itself in terms of “songs” ( pāṭṭu) (preface, line 98), and the oldest among the surviving manuscripts refers to each chapter of the narrative with the same term. Composed primarily in one of Tamiḻ literature’s oldest meters (akavaḷ ), Cāttaṉār’s text resounds with all the stylistic and rhetorical embellishment possible in early medieval Tamiḻ poetry, rendering even the most basic statement of a proper syllogism elegantly beautiful. Consider, for example, the following straightforward rendering of a syllogism: The proposition is: “This mountain possesses fire.” “Because it possesses smoke” states the proper reason. “Just like a kitchen” states the example. (29.59–61) In Cāttaṉār’s poetic hands, however, this becomes an elegant three lines filled with playful alliteration, rhyme, and tightly structured internal rhythm: pakkam immalai neruppuṭaittu eṉṟal pukaiyuṭaittu ātalāl eṉal poruntu ētu vakaiyamai aṭukkaḷai pōl tiṭṭāntam Although a full discussion of Tamiḻ poetic form lies far beyond the scope of this chapter, the preceding lines lend the syllogism a coherence and beauty that the English translation (or a similar statement in prose) simply cannot capture. The initial rhyme of pukai and vakai ties together the second and third lines, as does the alliteration of pakkam and pukai in the first and second lines. The internal coherence of each poetic line is bound together by the repetition of “pa” sounds in the first line in pakkam and neruppuṭaittu; the long vowels of ātalāl and ētu in the second line – like the repetition of the -ai in vakaiyamai and aṭukkaḷai in the third line – similarly grant each line an elegant coherence. Cāttaṉār’s discussion of inferential reasoning, in other words  – like his final discussion of interdependent origination – is meant to be not only philosophically meaningful, but also

120

Cāttan ‒ ār

poetically beautiful. The final two chapters of the text not only perform significant work within the overall narrative, as previously described, but each also contributes to the poetic impact of The Renunciation of Maṇimēkalai as a whole. The precise nature of that poetic impact on Cāttaṉār’s learned audience cannot be known for certain, but literary theory in both Tamiḻ and Sanskrit suggests that a poetic text such as this aims to cultivate certain emotional experiences to be savored among connoisseurs. The many scenes of hunger and profound suffering throughout the narrative  – no less Maṇimēkalai’s final vow to eradicate the karmic effects of birth (30.264) – taken together suggest an overall ethos of pathos and compassion, given philosophical framing in the final two chapters (see Monius 2001, 33–57).

Conclusion The seemingly odd combination of narrative poetry and philosophy found in Cāttaṉār’s sole extant work suggests that the complexities of both the text’s vision of karma and the gradual unfolding of the beneficial conditions leading to the heroine’s renunciation require the power of story-telling to convey the full impact of such forces at work in the world. The preceding narrative does more than merely illustrate the points of doctrine noted in the final two chapters; rather, it conveys more fully how proper reasoning and interdependent origination operate in human lives. Logic and the twelve links in the chain of interdependent origination cease to be merely abstract ideas; they have been shown throughout the entire narrative to be the forces that govern everyday human existence. The dense philosophical discussions of Cāttaṉār’s final two chapters, placed at the end of an equally complex narrative of primary plot suffused with multiple branch stories, offer a theorization of sorts for the narrative that precedes it; one without the other would be insufficient. The Renunciation of Maṇimēkalai provides a rich resource for considering Tamiḻ-language forms of narrative philosophy, as Cāttaṉār seamlessly blends extensive philosophical discussion into his core narrative. Cāttaṉār’s text presents the earliest extant example of formal philosophical discourse in Tamiḻ, and the embedding of philosophy in poetic narrative appears to be a unique feature of Tamiḻ literature of this general period. Composed perhaps in the tenth century by an unknown Jain author, for example, the Nīlakēci features its eponymous heroine traveling throughout India to debate the celebrated teachers of other religious schools and convert them to the Jain path. The fourteenth-century commentary on the text, attributed to Camaṉa Tivākara Vāmaṉa Muṉivar, preserves a verse from a long narrative no longer extant, the presumably Buddhist Kuṇṭalakēci,8 to which the anonymous Jain text is taken to be a direct response (Nīlakēci viḷakka urai 1973, 138). Unlike The Renunciation of Maṇimēkalai, in which more than twenty-five chapters of beautifully wrought narrative prose precede the later chapters on the philosophies of other religious schools – the treatment of inference and the final chapter on interdependent origination – the Jain text devotes far more time to debate and considerably less space to building a narrative. Other poetic philosophical texts are known only through stray references, such as a handbook of logic in metrical form called The Book of Logic (Aḷavainūl) (Cāminātaiyar 1978, 137), perhaps pointing to a larger Tamiḻ tradition of composing philosophical argument in verse. Cāttaṉār’s The Renunciation of Maṇimēkalai provides a brief but rich resource for considering the knowledge of – and significance of – Buddhist philosophy in Tamiḻ intellectual culture in the early medieval period. Composed in verse, rather than the prose of philosophical commentary, and embedded in sophisticated poetic narrative in which both play a critical role, each chapter’s presentation of philosophical argument has few parallels in Sanskrit or the regional

121

Anne E. Monius

languages of pre-colonial South Asia. Cāttaṉār’s sole work offers significant opportunities for fresh consideration of the relations among literature and philosophy, narrative, and logic.

Notes 1 [Editors’ note:] We feel extremely fortunate to have received this unique contribution from Anne E. Monius, who inspired so many scholars with her sensitivity to literary culture in the history of religions in South Asia, not long before her unexpected passing on August 3, 2019. Anne was renowned for her quiet yet profound generosity toward her students, as well as all others who came into her orbit. As she translates elsewhere, quoting the protagonist in The Renunciation of Maṇimēkalai, “Those who (never) cease loving all human beings are those who realize the final truth that eradicates suffering” (Anne E. Monius, “Literary Theory and Moral Vision in Tamil Buddhist Literature,” Journal of Indian Philosophy 28 [2000], 211). By this logic, surely Anne is one who has realized that final truth and thus, we must believe, suffers no more. 2 The best available translation of The Renunciation of Maṇimēkalai into English is by Daniélou (Cāttaṉār 1989). 3 Although the author of The Renunciation of Maṇimēkalai is named Cāttaṉ in the preface to the text, scholarly convention adds the Tamiḻ honorific -ār to his name. This practice will be followed throughout. 4 All quotations from The Story of the Anklet are drawn from Iḷaṅkō Aṭikaḷ (1978). All translations from the Tamiḻ throughout the chapter are by the author. 5 All quotations from The Renunciation of Maṇimēkalai are drawn from Cāttaṉār (1981). 6 While many scholars have sought to align the author of The Renunciation of Maṇimēkalai with earlier classical Tamiḻ poetry (known collectively as “Caṅkam” literature and dated to the early centuries of the Common Era), no clear connection can be found thematically or stylistically with such earlier work attributed to a poet known as Cīttalai Cāttaṉār. Named as the author of roughly eleven poems from these early anthologies of themes of “inner” (akam) love and “outer” ( puṟam) war, nothing aside from the name, Cāttaṉ, seemingly ties this poet to the author of The Renunciation of Maṇimēkalai. 7 See Richman (1988, 19–23) and Monius (2001, 18–24) for more detailed plot summaries. 8 Like the Nīlakēci, presumably known by the name of its heroine.

Bibliography Buddhaghosa. 1969. Visuddhimaggo with Paramatthamañjūsaṭīkā. Edited by Revatadhamma. 4 vols. Varanasi: Varanaseya Sanskrit Vishvavidyalaya. ———. 1991. The Path of Purification (Visuddhimagga). Translated by Bhikkhu Ñāṇamoli. Kandy: Buddhist Publication Society. Cāminātaiyar, U. Vē. 1978. Caṅka tamiḻum piṟkāla tamiḻum. Ceṉṉai: Ṭākṭar U. Vē. Cāminātaiyar Nūl Nilaiyam. Cāttaṉār. 1981. Maṇimēkalai. Edited by U. Vē. Cāminātaiyar. Ceṉṉai: Ṭākṭar U. Vē. Cāminātaiyar Nūl Nilaiyam. ———. 1989. Manimekhalaï: The Dancer with the Magic Bowl. Translated by Alain Daniélou. New York: New Directions. Dignāga. 1927–1930. Nyāyapraveśa. Edited by Anandshankar B. Dhruva. Gaekwad’s Oriental Series, 38–39. Baroda: Oriental Institute. Hikosaka, Shu. 1989. Buddhism in Tamiḻnadu: A New Perspective. Madras: Institute of Asian Studies. Iḷaṅkō, Aṭikaḷ. 1978. Cilappatikāram mūlamum arumpata uraiyum aṭiyārkkunallār uraiyum. Edited by U. Vē. Cāminātaiyar. Ceṉṉai: Ṭākṭar U. Vē. Cāminātaiyar Nūl Nilaiyam. ———. 1993. The Cilappatikāram of Iḷaṅkō Aṭikaḷ: An Epic of South India. Translated by R. Parthasarathy. New York: Columbia University Press. Kandaswamy, S. N. 1978. Buddhism as Expounded in Manimekalai. Annamalainagar: Annamalai University. Krishna Raja, S. 1998. “Theory of Anumāna in The Renunciation of Maṇimēkalai.” In Buddhism in Tamiḻnadu: Collected Papers, edited by R. S. Murthy and M. S. Nagarajan, 279–300. Chennai: Institute of Asian Studies.

122

Cāttan ‒ ār Mādhva. 1997. Sarvadarśanasaṃgraha of Mādhavācārya. Edited by K. L. Joshi. Translated by E. B. Cowell and A. E. Gough. New Delhi: Parimal Publications. Monius, Anne E. 2001. Imagining a Place for Buddhism: Literary Culture and Religious Community in Tamiḻ-Speaking South India. New York: Oxford University Press. Nīlakēci viḷakka urai. 1973. Edited by Po. Vē. Cōmacuntaraṉār. Tirunelvēli: Teṉṉintiya Caivacittānta Nūṟpatippu Kaḻakam. Richman, Paula. 1988. Women, Branch Stories, and Religious Rhetoric in a Tamil Buddhist Text. Syracuse, NY: Maxwell School of Citizenship and Public Affairs, Syracuse University. Zvelebil, Kamil V. 1995. Lexicon of Tamiḻ Literature. Handbuch der Orientalistik, Zweite Abteilung: Indien, 9. New York: E. J. Brill.

123

8 SARAHA The Anti-Philosopher as Philosopher Roger R. Jackson

Introduction The late-first-millennium tantric poet Saraha (“the Archer”)1 seems an unpromising candidate for inclusion in a volume governed by an author-centered approach to Indian Buddhist philosophy. He is, to begin with – a bit like Friedrich Nietzsche in the Western tradition – more an anti-philosopher than a philosopher, at least if we conceive of philosophy as a set of rational operations and arguments utilized in an allegedly dispassionate search for objective truth. He scorns his Indian contemporaries for their arrogant posturing and unwarranted confidence, and he insists, as a gnostic and mystic, that if we are to be free, we must transcend conceptuality, with its attendant dichotomies, and directly realize the primordial purity, emptiness, luminosity, bliss, and freedom of our own mind – and then express our realization not in the desiccated prose of the academician, but in the evocative and resonant language of the poet or singer of songs. Furthermore, if Saraha is an author at all, he is a most elusive one, almost impossible to locate in place and time. Indeed, some scholars have questioned his historicity and suggested that he may be a literary invention, a name and a set of colorful stories slapped onto a collection of works that may not all have been composed by the same person. Yet for South Asian and, especially, Tibetan Buddhists of the past millennium, Saraha has been a supreme exemplar of a radical way of being human that points beyond reason to the “joyful wisdom” possessed by those courageous enough to reject social, religious, and philosophical conventions; see themselves as they truly are; and live their lives, accordingly, with spontaneity, wisdom, and compassion. Following tradition, we will conjure “Saraha” here as a single individual, do our best to locate him within the Indian and Buddhist landscape from which he emerged, and explain why this arch-critic of rationality may deserve the title of “philosopher.”

Saraha in His Setting In the pantheons of late Indian Buddhism and Tibetan Buddhism, Saraha belongs to a class of saints known as great adepts – mahāsiddhas – who flourished in South Asia in the late first and early second millennia CE. Often said to number eighty-four, they are known primarily through the colorful stories of their lives and deeds composed in India, Nepal, and Tibet, and through spiritual songs – aphoristic couplets (dohā), adamantine songs (vajragīti), and performance 124

DOI: 10.4324/9781351030908-13

Saraha

songs (caryāgīti) – handed down under their names in oral and written traditions. The adepts are notable for subverting social and religious norms, glorifying the body and senses, celebrating spiritually potent female wisdom figures called yoginīs or ḍākinīs, instructing ordinary people through vernacular songs, displaying supernormal powers, and unabashedly exalting the freedom and joy entailed by direct knowledge of ultimate reality – variously referred to as the real (tattva), gnosis (  jñāna), the connate nature (sahaja), the natural mind (nijacitta), great bliss (mahāsukha), and the great seal (mahāmudrā) – or simply “that” (tat) or “mind” (citta). Most great adepts were followers of the esoteric Mahāyāna traditions known as Vajrayāna, or Secret Mantra, that dominated Indian Buddhism from the seventh to thirteenth centuries. Based on teachings retroactively attributed to the Buddha called tantras, these restricted – yet, paradoxically, widely popular – traditions insisted that their followers could become fully awakened in this very lifetime through receiving initiation from a qualified guru and undertaking practices in which they “made the goal into the path” by seeing themselves, other beings, and their environment as divine. In the late and highly esoteric systems known as the Mahāyoga and Yoginī tantras, practitioners would transform themselves into buddhas through elaborate and sometimes transgressive yogic techniques that involve manipulating mental and physical energies within the subtle body (sukṣmaśarīra) that interpenetrates our coarse physical body, sublimating such “negativities” as desire, anger, ignorance, and even death into their awakened equivalents. No great adept is more important than Saraha. He was influential in South Asia, but it is really in the Tibetan tradition that he gained prominence, especially in his capacity as the human source of the widespread Buddhist philosophical and practical teaching known as the great seal (mahāmudrā). Originally tantric, the term came to denote many things in Buddhist circles, including a ritual hand-gesture; clear visualization of oneself as a buddha–deity; a consort for sexual yoga practice; the primordial purity of the mind; a meditation technique whereby mind settles into its own empty, luminous, and blissful nature; and the final fruition of the Mahāyāna path, perfect buddhahood. In Tibetan Buddhism, especially but not solely in the Kagyü (bka’ brgyud) order, the great seal became the basis for much contemplative experimentation, creative mythmaking, scholastic analysis, poetic expression, and philosophical debate – and amidst all this, Saraha was a central figure: not only were his songs constantly read, memorized, quoted, and discussed, but long after his death, Tibetan masters from across the religious spectrum encountered him in visions and claimed to have received teachings directly from him. Some of Saraha’s dohā collections and performance songs have been preserved in two medieval Indic languages, Apabhramśa and Old Bengali,2 which were common in northeast India late in the first and early in the second millennium. The largest trove of Saraha texts – twenty-five or so – is found in various editions of the Tengyur (bstan ’gyur), the collection of works by Indian followers of the Buddha translated into Tibetan between the eighth and fourteenth centuries; the Saraha Tengyur texts were all translated in the eleventh century or later. Many are devoted to tantric deities and rituals, but most are tantric songs. The best known of these, the Treasury of Dohā Songs (Dohākoṣagīti) – often referred to by Tibetans as the People Dohās (PD) to reflect its original audience – is typically included in an “essential trilogy” (T. snying po skor gsum) with two other collections, the Queen Dohās (QD) and King Dohās (KD),3 also named for their audience. Other Saraha songs in the Tengyur include a trilogy of adamantine songs (vajragīti) – the Body Treasury (BT), Speech Treasury (ST), and Mind Treasury (MT)4 – and numerous instruction songs on such topics as the great seal, nonconceptual awareness, and the Buddhist path. Saraha’s songs typically involve a rich mix of sardonic social criticism, ecstatic affirmation of the purity and potency of the mind, and direct instructions to his listeners, all brought indelibly to life through his skillful deployment of simile, metaphor, paradox, humor, word-play, and poetic rhythm. 125

Roger R. Jackson

The surviving biographical material on Saraha was written beginning in the eleventh century and is available only in Tibetan, though some works may have been translated from Indic originals. As noted by Kurtis Schaeffer (2005, 15–25), there are two major Saraha narratives. In the “radish-girl narrative” found in Abhayadatta’s Lives of the Eighty-Four Great Adepts, Saraha is an observant east Indian brahmin by day and a practitioner of Buddhist tantra by night. Accused by his fellow brahmins of drinking alcohol, he denies the charge, then proves his veracity by surviving various physical ordeals. Asked by his repentant accusers to teach the path to liberation, he sings dohās to the country’s king, queens, and people, respectively. He then retreats to another country with a fifteen-year-old servant girl, who assists him in his practice. One evening, as she cooks him a radish stew, he enters a contemplation that lasts twelve years. When he emerges from it, he asks the girl for his radish stew. She mocks him for his attachment to radishes and his enslavement to conceptual thought. At this, he attains liberation, and departs with the girl for the celestial realm of the ḍākinīs. In the “fletcheress narrative,” first mentioned by the eleventh-century Nepali commentator Balpo Asu and filled out by later Tibetan writers, Saraha is a south Indian brahmin who one day encounters a low-caste female arrow maker in the marketplace of his town. When he asks her about her profession, she replies, “Noble son, the intent of the Buddha can be understood through symbols and actions, not through words and texts,” and proceeds to give him an explanation of the way the arrow and archery symbolize the realization of one’s own natural mind. At this, Saraha is instantly liberated. He moves to a cremation ground with the fletcheress, where they engage in tantric rituals that soon draw the condemnation of the townspeople. In response, Saraha teaches his dohā collections to the people, queens, and king, respectively, then departs to the southern mountain of Śrī Parvata, where he resides to this day.

Saraha as Critical Philosopher We saw in the biographical traditions that by living with a low-caste woman Saraha courted the opprobrium of people obsessed with caste purity. His rejection of this cornerstone of the Indian social system is quite evident in a song attributed to him by the Tibetan historian Tāranātha (1575–1634): Oh ho! I am a brahmin and I live with a girl who works with bamboo. I see neither caste nor no caste. (Tāranātha 1983, 35) He provides further evidence of his unconcern for caste in his Queen Dohās, where he describes females suitable to serve as a yogī’s great seal (mahāmudrā), his partner for sexual yoga practice, as including “low-caste women, laundresses, whores, or pickers of rags” (QD 436; Jackson 2012, 179). Later, he speaks of how a yogī who has practiced sexual yoga with his consort “fades into conceptless space” so that even when he lives with her in a cremation ground, “he makes friends with outcastes, and draws the corpse cart along” (QD 47; Jackson 2012, 179). Similarly, he asserts in his People Dohās that “enjoying yourself in an outcaste’s home, you don’t get covered with dirt” (PD 56b, Jackson 2004, 857). And, on the most basic level, “in the connate nature where things taste the same, you’ll find neither śūdra nor brahmin” (PD 46, Jackson 2004, 80). Saraha is even more pointed in his critique of religion. His most famous attack on the practices and practitioners of his day comes at the outset of his People Dohās, where he denounces members of his own brahmin caste for their vain incantation of the Vedas, pointless oblations, 126

Saraha

and spiritual charlatanism (PD 1–5; Jackson 2004, 53–55) and goes on to question the utility of such Hindu rituals as lamp lighting, food offering, mantra practice, pilgrimage, forest penance, and immersion in holy rivers (PD 14–15; Jackson 2004, 60). He also decries the extreme asceticism of Jains, whom he describes as “mocking the path” by going naked and plucking out their hairs (PD 5; Jackson 2004, 55). Nor does Saraha spare his fellow Buddhists. He complains about “so-called novices, monks, and elders” who “seek to dry up the intellect,” as well as proponents of the Mahāyāna, “where scripture turns to sophistry and word-play” (PD 10–11a, Jackson 2004, 57–58). To those Buddhists who assert that the heart of their practice is meditation, he remarks, The whole world is deluded by meditation – and no one perceives their inmost nature. (PD 35b; Jackson 2004, 70) His scorn is also directed toward proponents of the Vajrayāna. In the Queen Dohās, he sings of how the adamantine gnosis, “lovely in its natural state,” is devoid of maṇḍalas or fire offerings, quite without mantras, mudrās, and consecrations, and can’t come about through any tantra or treatise. (QD 31, Jackson 2012, 177) Elsewhere, he dismisses advanced subtle-body practices, noting that “drawing energies up and down/in the turning cakras, you will not find . . . truth” (MT 25, Braitstein 2014, 165–66), and asks enthusiasts of tantric sexual yoga, If you don’t grasp everything as it is, how in the midst of sex will you perfect great bliss? ( PD 91, Jackson 2004, 105) Saraha’s critique of religious practice extends to nearly every form of human behavior, so that his advice sometimes seems to amount to “do nothing”: Mind at its core is free from virtues and faults. Truth itself requires no deeds at all, while the mind that’s done with deeds is great bliss sublime. (QD verse 77, Jackson 2012, 183) In short, so much for religion. Saraha also harshly criticizes philosophy and philosophers. In his Body Treasury, he attacks the four major Indian Buddhist philosophical schools: Vaibhāṣikas and Sautrāntikas, Yogācārins and Mādhyamikas and so forth, criticize each other and argue; ignorant of the real . . ., they turn their backs on the connate nature. (BT 2, Braitstein 2014, 125) 127

Roger R. Jackson

He consistently mocks the pretensions of thinkers who “intellectualize based on signs/but . . . can’t see the real,” and “die of thirst in the desert of variant texts and meanings” (PD verses 78a, 56, Jackson 2004, 98, 84). In the Queen Dohās, he notes the falsehood of “every term and symbol you impose” and the importance of “clearing away the husk of logical complication” so as to comprehend what is true, real, and essential (QD 75, 79, Jackson 2012, 183, 184). Indeed, not just intellect but mind itself is problematic: “the world is bound by mind,/and no one at all can grasp the non-mind” (PD 78, Jackson 2004, 97) because “the mind . . ./can’t be considered anything by anyone, there is no mind” (QD verse 24, Jackson 2012, 176). Furthermore, that which is to be realized – the real, the connate nature, the great seal – is without syllable or color or qualities . . .; it can’t be spoken or known. ( PD 58, Jackson 2004, 87) One important consequence of the rejection of language and thought is that the dichotomies through which we understand and operate in the world are effaced. Thus, the distinctions between self and other, saṃsāra and nirvāṇa, pure and impure, suffering and pleasure, and even good and bad, simply do not apply. Saraha describes reality as a state in which “there is no self or other” and we are “free from existence and nirvāṇa” (PD 96, 110, Jackson 2004, 108, 115). Regarding ethics, he observes: Even the label ‘good’ is a sickness plunging you into saṃsāra, and when you label an act as ‘evil,’ you can’t prevent its fruition. (QD 75, Jackson 2012, 183) The consequences of this stance for a yogī’s conduct are dramatic: Behaving without regard for dos and don’ts, like an elephant plunging impulsively into a pond, his mind perpetually crazed, he performs the basest of deeds yet is free. (QD 50, Jackson 2012, 180) Saraha rejects language and thought – and by extension social, religious, and philosophical norms – because the ultimate nature of all phenomena is emptiness (śūnyatā): their lack of self-existence, their inconceivability, their unfindability apart from the mind. As he sings in the King Dohās, There’s nothing to be negated, nothing to be affirmed or grasped; for it never can be conceived. (KD 35, Guenther 1969, 70) As deluded beings, however, we believe – either instinctively or on the basis of reflection – that there must exist a true self – either “in here” or “out there” – that is the real, metaphysical foundation of ourselves and the world. We are reinforced in this belief by the immediacy of our senses, the persistence of our body and mind, and the apparent precision and coherence of our language and concepts. But all these deceive us: nothing has a real foundation, no such self will ever be found without or within, and all our fine-spun theories are vain exercises in imputing complexity where there is only simplicity, externality where there is only mind, and 128

Saraha

something where there is nothing. Indeed, to be undeceived, we must simply recognize, with Jack Kerouac, that the Buddha is “philosophy’s dreadful murderer” (Kerouac 1959, 179).

Saraha as Mystical Philosopher Our discussion thus far might make Saraha seem like a nihilist, as Nietzsche and Nāgārjuna appeared to critics in their own eras. However, just as Nietzsche and Nāgārjuna turn out on closer inspection not to be nihilists but simply radical thinkers who use their critiques to clear the field for a fresh way of seeing and being in the world, so Saraha’s rhetoric of critique and negation reflects his attempt to cut through conventional thinking and practice so as to lay bare the fundamental transrational reality that must be realized directly if we are to be free. In this respect, Saraha is like the mystics of many traditions, who insist that the divine only can be reached by silencing the deluded, dichotomous chatter of words and ideas and letting it shine on its own. And, like other mystics, he must paradoxically assert that the divine is utterly beyond conception or description, but because it is the most important thing we can possibly know, it must somehow be indicated – or perhaps sung – so that we can experience it, not through a miasma of concepts, but face-to-face.

Mind and Reality in Saraha As we have seen, the single thing Saraha says we must understand goes under many names, but most often, perhaps, it is simply “mind” (citta). Mind here refers not to a set of capricious emotions, rational operations, fanciful or deliberate imaginings, or “experiences” of any sort, but mind as it really is – pure, empty, luminous, aware, ecstatic, and free. As Saraha puts it in a famous verse, “the single seed of everything is mind, where existence and nirvāṇa both arise” (PD 41, Jackson 2004, 73). His assertion can be read several ways. The most obvious would be that all things are the result of the karma of beings, and karma, being mental “intention” (cetanā), is really “mind.” By another interpretation, Saraha is claiming – in the phenomenalist manner of Asaṅga, Vasubandhu, and other figures in the Yogācāra school – that everything in some sense is mind, or at least cannot be known outside mind. He sings, for instance, of how All Buddhas, sentient beings, and phenomena arise simultaneously with mind itself. (BT 48, Braitstein 2014, 135) More specifically, Mind, intellect, and the formed contents of mind are it, So too are the world and all that seems from it to differ, All things that can be sensed and the perceiver, Also dullness, aversion, desire, and awakening. (KD 33, Guenther 1969, 69) Saraha’s assertion that everything in saṃsāra and nirvāṇa  – even our delusions  – partakes of the primordial purity, emptiness, luminosity, awareness, ecstasy, and freedom of the ultimate raises several questions. First, how do we stray from our own original nature? How could something essentially pure ever be tainted by impurity? On one level, the answer should already be clear. Our senses, body, and intellect present reality to us in deeply misleading ways: 129

Roger R. Jackson

as multiple rather than single, defiled rather than pure, substantially existent rather than empty. To return to our original nature, then, seems to require transcending the body, senses, and intellect – for in doing so, we eliminate the main basis for fixation on self, and what remains is just the primordially pure and blissful mind itself. On a different and deeper level, however, we never have departed from our original nature since, as Saraha repeatedly assures us, all things – including the body, senses, and intellect, and even delusion and suffering – are “it,” or “that,” or “the real.” A further question therefore arises: what is the point of trying to regain our original nature when we have never lost it to begin with? If we all already are buddhas, why practice the path? And if everything already partakes of the ultimate, why restrain the body and senses, why seek to transcend intellect, why bother to behave ethically? A further question inevitably arises: if all things are empty, is the primordial nature insisted upon by Saraha – the pure, luminous, blissful mind – also empty or, as his language sometimes suggests, is it an exception, the one thing that truly exists? Saraha is equivocal on such matters, delighting in paradox and apparent contradiction, but we will briefly attempt to tease out some possible answers. In attempting to understand Saraha’s paradoxes, the key, as is often the case in Mahāyāna Buddhism, is an appreciation for the interplay between the two truths: the ultimate ( parmārtha) and the conventional or phenomenal (vyāvahārika, saṃvṛti). Although these are variously defined and distinguished by different thinkers, conventional truth typically corresponds to the more or less reliable everyday deliverances of the senses and intellect, while the ultimate denotes the real and abiding nature of persons and phenomena. Mahāyāna thinkers generally concurred that the ultimate nature of things is emptiness but disagreed over what emptiness is empty of. Philosophers associated with the Mādhyamika school, such as Nāgārjuna and Candrakīrti, typically described emptiness as a “non-affirming negation” ( prasajya-pratiṣedha): the lack – in any entity or concept – of intrinsic, self-sufficient existence, with emptiness itself being the ontological condition for the possibility of the functioning of the environments, things, persons, and minds that constitute the conventional world. A different view was taken by proponents of Yogācāra and the buddha–nature doctrine, who typically regarded the emptiness of worldly things to be their lack of difference from the mind perceiving them and the emptiness of buddha mind to be an “affirming negation” ( paryudāsa-pratiṣedha): its lack of anything other than buddha mind, which itself is replete with eternally perfect qualities.8 Philosophers may attempt to draw clear lines between the two truths, but many Mahāyāna masters – from the authors of the Perfection of Wisdom sūtras, to Nāgārjuna, to great tantric adepts like Saraha – switch between ultimate and conventional levels of discourse without notice, leaving readers to fend for themselves amidst the apparent contradictions in such statements as Rāhulabhadra’s celebration of the goddess Prajñāpāramitā: Seeing you, one is bound to suffering, not seeing you, one is also bound. Seeing you, one is freed from suffering, not seeing you, one is also freed. (verse 15, Strong 2002, 142) Returning to the previously mentioned questions posed to Saraha, we might note first that he gives no real explanation of how or why our original purity has been tainted by delusion: it is simply the way things are – a mystery, perhaps. As to whether we ought to bother practicing the path if we all already are buddhas, a possible answer is supplied through the concept of two truths: yes, we all are eternally in possession of perfect realization, but – for whatever reason – we have fallen asleep spiritually and do not recognize our original, undimmed enlightenment. Thus, while we may not really be deluded, we are conventionally in deep distress – and amidst 130

Saraha

our distress, conventional notions of bondage and liberation, pure and impure, and good and evil, must be applied, and we must be induced to recognize our original nature as if we had strayed from it. The concept of two truths also helps us understand what it means to say, as Saraha does, that ultimate reality is at once empty and real. If, as emphasized in Yogācāra and buddha–nature circles, the emptiness of the primordial mind is an “affirming negation,” then the mind’s emptiness need not contradict its eternal perfection, since emptiness here simply means the mind’s emptiness of anything other than its own perfection – a view that would be referred to in Tibetan Buddhism as “other-emptiness” (gzhan stong). Although Saraha seems naturally aligned with this perspective, proponents of the opposing “self-emptiness” (rang stong) position could and did read his verses as suggesting that the primordial mind is empty of self-existence, just as everything else is, and that his apparent assertion of the mind’s true existence must, like similar claims in buddha–nature literature, be taken as inspiring and important conventional claims, but conventional claims nevertheless.

The Path to Realization Given that, conventionally speaking, we must return to our original nature, how do we do so? From early on, Buddhists have discussed whether awakening is achieved suddenly (the “subitist” position) or in progressive stages (the “gradualist” position). When the discussion has become a debate – as in late first-millennium India, China, and Tibet – subitists have insisted that awakening depends solely or primarily on seeing reality directly, and that no amount of conventional philosophizing or gradual practice will bring it about, so we must simply drop concepts and realize our original nature at a single blow. Gradualists have countered that sudden awakening, while not impossible, requires sharp intelligence and prior training in this or a previous life, and that prematurely dropping such concepts as self and other, saṃsāra and nirvāṇa, and good and evil, is extremely dangerous for those who are unprepared – the vast majority of practitioners.

The Subitist Approach Saraha himself acknowledges that “a yogi desiring great gnosis/may proceed by stages or at once” (QD 80, Jackson 2012, 184), but at first glance, he seems partial to subitism. He declares, for instance, that “the great seal is instantaneous full awakening” (BT 71, Braitstein 2014, 139), and in a verse that is virtually a charter for meditation on the great seal, he sings, If mind points to mind, concepts are held in suspension; just as salt dissolves in water, so mind dissolves into its nature. ( PD 78b, Jackson 2004, 98) A key way for mind to see mind is to “make the mind like space” (PD 42a, Jackson 2004, 74) by dropping concepts and resting within it: “this mind so tightly bound – relax it, and you’re free, no doubt” (PD 42h–43, Jackson 2004, 78). Similarly, the free mind settles into its nature just as it’s always been: without need to label or examine. (QD 74, Jackson 2012, 183) For the most part, Saraha’s subitism is straightforward: do not think or do anything, he says, and do not meditate in the usual sense; just see your own mind – and everything else – as the 131

Roger R. Jackson

pure, empty, blissful reality they always have been. In several texts, however, he introduces a quasi-systematic element to his subitist instruction through a discussion of four symbolic terms: recollection, non-recollection, the unborn, and beyond-intellect. The first two are the most crucial, for they describe the way we misconstrue reality and the way we can overcome our error. Recollection, or mindfulness (smṛti), is a venerable term in accounts of Buddhist meditation: typically, it is re-cognition of a focal object when one has lost concentration. Here, it connotes any deliberate application of the intellect, which, as we have seen, is anathema for Saraha. The way we eliminate it is by “non-recollection” (asmṛti): simply dropping thought and letting our original nature shine forth. A more or less synonymous term used by Saraha is cognitive disengagement (amanasikāra), which negates the classical Buddhist mental factor of “cognitive engagement” (manasikāra), whereby we intentionally bring some object within the purview of cognition. As with non-recollection, cognitive disengagement points to our need to immediately and completely abandon intellect and enter directly into a state whereby we see reality as it is. The third and fourth symbolic terms are simply elaborations on the characteristics of this reality: “the unborn” indicates that our original nature is beyond either origination or cessation, while “beyond-intellect” indicates that this unarisen reality is beyond all analytical comprehension or verbal predication.

Gradual Paths The subitist approach seems simple, but it is deceptively so, and Saraha concedes as much by presenting a variety of more gradual practices that might lead to the same realization. The major context for these gradual practices is the “completion stage” (saṃpannakrama) techniques required by the esoteric and sometimes transgressive Mahāyoga and Yoginī tantras. Although Saraha mocked such practices because they could be misconstrued and misapplied by the unwary, they are in fact the key to realization for that majority of practitioners unsuited to subitism. Because completion-stage practices are carried out within the body, we may read Saraha’s body-positive verses – e.g., “what’s bodiless is hidden in the body;/know this, and you’re freed there” (PD 89, Jackson 2004, 104) – as implicitly referring to advanced yogic techniques involving the 72,000 channels (nāḍī) that interpenetrate our coarse physical body, the “hormonal” drops (bindu) and breath-related energies ( prāṇa) that move within the channels, and the energy centers (cakras) found where the outer channels intersect with (and typically constrict) the all-important central channel. The way we are freed within the (subtle) body is through practices that involve moving wind-related energies from the outer channels into the central channel, and then gaining proficiency in moving energies and drops up and down there, opening up the cakras, and inducing various states of gnostic bliss. Typically, the sequence culminates in the dissolution of the energies, drops, and awareness itself into a drop at the heart cakra, where we purify the subtlest mental and physical aspect of ourselves and create the buddha bodies that complete the Mahāyāna path. There are various ways of describing these culminating practices, and while Saraha does not lay out any particular system or sequence, he alludes to certain operations, such as making the mind like space and dissolving the winds and other elements within the subtle body until we attain “utmost great bliss” (PD 42b–c, Jackson 2004, 75). Elsewhere, he refers to techniques of sexual yoga that help induce the movement of energies into and within the central channel. In the People Dohās (verses 84–86, Jackson 2004, 101), he celebrates a mysterious yoginī who may be a goddess he worships or his flesh-andblood partner – or both. In his Queen Dohās, he details the qualities possessed by a proper consort (QD 41–46, Jackson 2012, 179) and specifies that the tantric yogī “who’s . . ./aware that 132

Saraha

all things are suchness. . . ./ . . . sees great bliss during sex” (QD 39–40, Jackson 2012, 178). Elsewhere, he refers to the consort as an “action seal” (karmamudrā), one of four seals – or confirmations – of spiritual practice described in the later tantric systems (BT 79–96, Braitstein 2014, 140–44).9 While Saraha accepts the necessity and efficacy of the action and other seals, it is clear that for him the great seal (mahāmudrā) – the unmediated, complete, and final realization of the nature of mind and reality – is supreme, and perhaps sufficient on its own (BT 81, ST 31, 34–35, Braitstein 2014, 141, 157, 158). In later tantric formulations, the completion stage can only be approached when one has mastered the “generation stage” (utpattikrama), a complex system of practices – a sādhana – in which we change how we see ourselves and the world, visualizing ourselves and others as divine buddha–deities, regarding all words and sounds as mantras, and imagining our environment as a pure land or maṇḍala. Saraha asserts that “the stages of generation . . ./are to the stage of completion like little stars to the moon and sun” (ST 32, Braitstein 2014, 157) and repeatedly dismisses the “yoga with signs” involved in the generation stage (e.g., ST 45, Braitstein 2014, 160), yet he often refers to mystic syllables, maṇḍala circles, and other elements of sādhana (e.g., PD 90, 90a, 98, Jackson 2004, 104, 105, 109), making it likely that his subtle-body yoga practice was indeed preceded by the re-envisioning required by the generation stage. To practice the generation stage, in turn, requires an esoteric initiation, or empowerment (abhiṣeka). Saraha rarely recommends, and occasionally denigrates, this linchpin of tantric practice, but because his main frame of reference remains the completion stage, which typically presupposes both initiation and sādhana practice, he assumes it in his discourse. And in fact, A person deprived of the precious initiations is like a lowly servant aspiring to be king. (QD 16, Jackson 2012, 178) The source of initiation – the sine qua non for receiving it – is the guru, whose importance as initiator, instructor, and object of devotion cannot be overstated. As Saraha asserts, “in your reverence for the guru, the discipline of the secret vehicle is complete” (BT 34, Braitstein 2014, 132). More fully, Through the guru, teachings, transmissions, commentaries, and special instructions are shown. You will attain your own intrinsic identity . . . through pure special instructions from the guru; if you worship the guru, you will obtain supreme connate bliss. (ST 5, Braitstein 2014, 194; cf. trans. 151) In the Queen Dohās, Saraha describes the guru as “the king of physicians,” “the sublime boat that delivers us/from saṃsāra’s deep, vast sea,” “the mighty boatman who’s gained great bliss/ . . . possessed of sun-like gnosis,” “the wheel-turning king who’s wise in the ways/of changing all dharmas to bliss as elixirs turn metals to gold,” and the one whose “mind, like a river, submerges dichotomous views” (QD 53–55, Jackson 2012, 180–81). In short, “the root of every attainment is the adamantine master” (QD 62, Jackson 2012, 181). Elsewhere, Saraha insists repeatedly that the only reliable basis for realization of our original nature is the guru’s “ambrosial” instruction. Without it, he says, we are lost; with it, we have access to “that,” the connate nature, the rootless real, our inmost nature, our inmost thought, our purified awareness (PD 17, 29, 37, 38, 39, 69, respectively). 133

Roger R. Jackson

In principle, a disciple should only accept a guru who is pure, insightful, compassionate, and experienced on the path. Conversely, a guru is only supposed to initiate or instruct a disciple who already possesses the qualities emphasized by non-tantric Buddhism: restraint of the senses, detachment from saṃsāra, a desire for spiritual liberation, an intellectual and contemplative appreciation of the nature of reality, and profound compassion. Saraha instructs his audience not to be bound by sensuous things, noting that the connate nature is found “where senses subside and self-nature cannot stand” (PD 71, 29, Jackson 2004, 94, 68); and in one of his most oft-quoted passages, he sings, Rejecting compassion, stuck in emptiness, you won’t gain the utmost path. Nurturing compassion all by itself, you’re stuck in rebirth and won’t win freedom. But if you’re able to join emptiness with compassion, you won’t remain in existence – or nirvāṇa. (PD 15a–b, Jackson 2004, 61) Thus, a disciple’s suitability for initiation is – at least in principle – dependent upon previous mastery of basic Buddhist virtues. The tantric path is difficult and potentially dangerous and, transgressive rhetoric aside, actually requires an even greater degree of renunciation, understanding, and compassion than non-tantric practices, lest the disciple be overwhelmed by the mental and physical forces evoked in the Vajrayāna. For Saraha, the guru is important above all because they point out our original nature, which we recognize either through receiving initiation and undertaking various tantric yogas or perhaps by the guru’s blessing and instruction alone. In this sense, the guru may open the way either to the gradual practices of the Vajrayāna path or to a sudden approach to awakening that short-circuits the slow accretion of virtues in favor of attaining the qualities of a buddha all at once.

The Fruits of the Path Whether approached suddenly or gradually, the final fruit of the path is buddhahood, which was subdivided by Mahāyānists into variously-enumerated “bodies” (kāya). Saraha admits four bodies: (1) a truth body (dharmakāya) consisting of a buddha’s pure and perfect mental qualities; (2) a celestial enjoyment body (sambhogakāya) that instructs high-level bodhisattvas in Mahāyāna theory and practice; (3) an emanation body (nirmāṇakāya) consisting of various manifestations for the sake of ordinary sentient beings; and (4) an essential body (svabhāvikakāya), which may signify either the unity of the first three bodies or the empty nature of the truth body. Saraha refers to these buddha bodies in the Queen Dohās: Non-dichotomy devoid of “is” and “isn’t” is the truth body, its essential bliss is the great-enjoyment body, its manifold guises for wandering beings are the emanation body, and the very gnosis that knows they’re indivisible is the essential self of them all. (QD 63–64, Jackson 2012, 182–83) Saraha does not mention the buddha bodies explicitly in the People Dohās, but expresses their nature and function metaphorically in a series of verses that bring together his poetic gifts and

134

Saraha

delight in paradox so as to emphasize the unity, emptiness, serviceability, and ineffability of the awakened state: The precious tree of nondual mind spreads through the triple world; it bears compassion flower and fruit, though there is no other or doing good. The precious tree of emptiness blooms compassion, various and many hued; another’s pleasure is its final fruit: mind intent on others’ joy. The precious tree of emptiness lacks compassion; it has no root or branches, so if you prattle about it, your troubles will only grow. It’s just one, just one tree – that’s why the fruit is one. Knowing it can’t be split, you’re free from existence – and from nirvāṇa. (PD 108–110, Jackson 2004, 114–15)

Conclusion: Saraha, Buddhism, and Philosophy The anti-philosophical strain in Saraha’s rhetoric might lend the impression that he has nothing to do with traditional Buddhist thought, but it should be evident by now that he is deeply indebted to the Buddhist thinkers preceding him. His critical spirit is actually quite akin to that of many Indian Buddhists, including Śākyamuni himself, Nāgārjuna, and many of his fellow great adepts. From a positive standpoint, he accepts conventionally such basic Buddhist notions as saṃsāra and nirvāṇa, the four noble truths, and the necessity to transcend self-grasping for the sake of liberation. He is equally beholden to Mahāyāna thought, sharing with Great Vehicle traditions such ideas as emptiness, non-duality, mind-only, buddha nature, and the bodhisattva’s need to master wisdom and compassion. More specifically, Saraha reflects the ideas and images developed by three major Mahāyāna philosophical traditions. Madhyamaka influence is evident from his repeated references to emptiness – however construed – as the nature of things and his use of paradox to undermine our certainty about the categories through which we encounter and attempt to control reality. Yogācāra influence is evident in his strong emphasis on the metaphysical primacy of mind and the importance of attaining a non-dichotomous understanding of its true nature. The influence of buddha-nature discourse is clear from his consistent assertion that all beings not only can become buddhas but in a basic sense already are buddhas because their minds are naturally pure, luminous, and empty of all afflictions. To these ideas derived from early and Mahāyāna Buddhism, Saraha adds distinctively Vajrayāna elements, including reliance upon the guru, celebration of the body and senses, an emphasis on idealized female figures, and techniques for attaining buddhahood within the subtle body. So, is Saraha rightly considered a great Indian Buddhist philosopher? It depends, of course, on what we mean by philosophy. If using conventional logical analysis and rational argumentation are the sole criteria, then the answer is no – his approach is a far cry from that of Nāgārjuna or Dharmakīrti. As Herbert Guenther notes, however, the tantric ethos forces us to reconceive philosophy as the attempt “to know reality directly, not by rumor or description . . . [but as] a way of life and quest for meaning [involving] the whole of man and not merely his brain” (Guenther 1963, 112, 113). This description easily could be applied to all of Buddhism – and indeed, to wisdom traditions everywhere. Such an approach – in which theory and practice are united in pursuit of human freedom – in turn suggests that figures like Nietzsche in the West and Saraha in India may help us to think about philosophy not as the allegedly dispassionate application of rational methods for the sake of arriving at some supposedly objective truth, but rather as the poetic articulation of a unified vision of reality and the

135

Roger R. Jackson

potential humans have for attaining ultimate wisdom, joy, and goodness within that reality. As we have seen, the reality Saraha sings is the pure, empty, luminous awareness that is our original nature, and in some sense the original nature of everything, with the consequence that the purpose of life is to realize – to reacquaint ourselves with – the basic nature we never have lost. This sense of philosophy may seem distant from the meaning of the term in recent AngloAmerican thought, but it finds intriguing echoes in the work of twentieth-century Continental European philosophers. Martin Heidegger (1889–1976), for instance, expressed an ongoing concern with the question of being and emphasized in his later works the importance of replacing “calculative thinking” with a “meditative thinking” that “contemplates the meaning in everything that is” through “releasement toward things and openness to the mystery” (Heidegger 1966, 46, 65). Pierre Hadot (1922–2010) undertook a radical reconsideration of the ancient Epicureans, Skeptics, and Stoics, seeing them not as mere logicians but as thinkers concerned with how to live a meaningful life through the practice of “spiritual exercises” aimed at selftransformation. He also argued that these thinkers’ use of rational argumentation was just one of many techniques conducive to humans’ attunement to themselves and the cosmos (Sharpe n.d.). And, if philosophy as a rational method is merely one of multiple ways to approach and identify reality, then perhaps poetry like that of Saraha (or Rumi, or Rilke, or Stevens) may be just as powerful – or even more powerful – a means to that end as reasoning, and just as “philosophical” as the syllogisms of learned scholars. Heidegger sets them on the same lofty plane in proclaiming that “the thinker utters being; the poet names what is holy” (Heidegger 1968, 391) – and in Saraha, we see the two roles coalesce so completely that we are forced to reconsider both poetry and philosophy, and ask whether, at the highest level, they really differ at all.

Notes 1 For modern scholarship on Saraha, see especially Guenther (1969, 1993); Jackson (2004, 2012), Schaeffer (2005); Shahidullah (2007); Braitstein (2014). 2 For editions, see Jackson (2004, 53–115) and Kvaerne (1977, 168, 222, 226), respectively. 3 These are the Performance Song Dohā Treasury (Dohākoṣa-nāma-caryāgīti) and Instruction Song Dohā Treasury (Dohākoṣopadeśagīti), translated in Jackson (2012, 173–84) and Guenther (1969, 63–71), respectively. 4 The full titles are A Body Treasury Called the Immortal Adamantine Song (Kāyakośāmṛtavajragīti), A Speech Treasury Called the Gentled Voiced Adamantine Song (Vākkośarucirasvaravajragīti), and the Unborn Treasury of Mind Adamantine Song (Cittakośājavajragīti); for editions and translations, see Braitstein (2014, 125–208). 5 In this and most other cases, I have slightly modified existing translations for the sake of terminological and stylistic consistency. 6 A parenthetical abbreviation followed by a number indicates the verse number in a given collection. 7 In citations from my translations of the Apabhraṃśa (but not the Tibetan) of Saraha’s People Dohās, I have restored the line length found in the original. 8 The terms for the two types of negation were not coined until later in the Indian tradition, but may reasonably be read back into the writings of early figures. 9 The four are the action seal (karmamudrā), pledge seal (samayamudrā), dharma seal (dharmamudrā), and great seal (mahāmudrā).

References Braitstein, Lara. 2014. The Adamantine Songs (Vajragīti) by Saraha. New York: American Institute of Buddhist Studies. Guenther, Herbert. 1963. The Life and Teaching of Nāropa. London: Oxford University Press. ———, trans. and annot. 1969. The Royal Song of Saraha: A Study in the History of Buddhist Thought. Seattle: University of Washington Press.

136

Saraha ———. 1993. Ecstatic Spontaneity: Saraha’s Three Cycles of Dohās. Berkeley: Asian Humanities Press. Heidegger, Martin. 1966. Discourse on Thinking. Translated by John M. Anderson and E. Hans Freund. New York: Harper and Row. ———. 1968. “What Is Metaphysics?” Translated by R. F. C. Hull and Alan Crick. In Existence and Being, edited by Werner Brock, 353–92. London: Vision. Jackson, Roger R. 2004. Tantric Treasures: Three Collections of Mystical Verse from Buddhist India. New York: Oxford University Press. ———. 2012. “Saraha’s Queen Dohās.” In Yoga in Practice, edited by David Gordon White, 162–84. Princeton: Princeton University Press. Kerouac, Jack. 1959. Mexico City Blues. New York: Grove Press. Kvaerne, Per. 1977. An Anthology of Buddhist Tantric Songs: A  Study of the Caryāgīti. Oslo: Universitetsforlaget. Schaeffer, Kurtis R. 2005. Dreaming the Great Brahmin: Tibetan Traditions of the Buddhist Poet-Saint Saraha. New York: Oxford University Press. Shahidullah, Muhammad. 2007 [1928]. The Mystic Songs of Kāṅha and Saraha: The Dohā-koṣa and the Caryā. Translated by Pranabesh Siṇha Ray. Biblioteca India no. 329. Kolkata: Asiatic Society. Sharpe, Matthew. n.d. “Pierre Hadot.” Internet Encyclopedia of Philosophy. Accessed March 19, 2022. https://iep.utm.edu/hadot/. Strong, John S. 2002. The Experience of Buddhism: Sources and Interpretations. 2nd ed. Belmont, CA: Wadsworth. Tāranātha, Jonang. 1983. The Seven Instruction Lineages. Translated and edited by David Templeman. Dharamsala: Library of Tibetan Works and Archives.

137

PART 3

Abhidharma Philosophers

ABHIDHARMA PHILOSOPHERS Introduction to Part 3

The Sanskrit term “Abhidharma” (“Abhidhamma” in Pāli) has multiple meanings. It often refers to a corpus of texts, one of the “three baskets” (tripiṭaka; tipiṭaka) of the Buddhist canon. In contrast to the discourses (Sūtra, Sutta) and the monastic code (Vinaya) – the other two “baskets” of the canon  – Abhidharma literature systematically analyzes forms of conscious experience and their objects in an explicitly Buddhist philosophical framework. One of the distinguishing features of the Abhidharma texts is the comprehensive account of mental and physical factors authors try to achieve, often in the form of lists. There are two complete canonical collections of Abhidharma texts extant today – one preserved in Pāli claimed by the Theravāda tradition, and one preserved in Chinese ascribed to the Sarvāstivāda tradition – but previously there were additional Abhidharma texts associated with other schools, some of which exist in fragments while others are completely lost. The term “Abhidharma” can also refer to commentaries on these canonical texts, as well as to other texts that are intended to provide a systematic analysis, metaphysics, phenomenology, and evaluation of the varieties of consciousness and its objects using the earlier Abhidharma categories. Abhidharma can thus be a way of characterizing a style and approach to doing philosophy in Buddhist traditions, one of the first manifestations of Buddhist scholasticism. What these diverse meanings of the term share is that Abhidharma takes up the teachings from Buddhist discourses and gives a more formal and methodical account of Buddhist doctrine. While Abhidharma developed in the early centuries of Buddhism in India, later thinkers, including those associated with the Mahāyāna traditions of Madhyamaka and Yogācāra, were working within an Abhidharma philosophical context, and many of them wrote Abhidharma texts. This is why Abhidharma can be seen as the backbone of so much Buddhist philosophy in India and remains a constant reference for Buddhist philosophers. In the first chapter of this part on Abhidharma philosophers, Maria Heim describes the Enumeration of Phenomena (Dhammasaṅgaṇī) and Analysis (Vibhaṅga), which constitute the first two books of the canonical Pāli Abhidhamma. Most traditional Theravāda commentators, as well as modern scholars, have interpreted the Pāli Abhidhamma as an ontology of ultimate reals. Heim defends an alternative interpretation for which she finds evidence in the Abhidhamma as well as in some postcanonical commentary, particularly that of Buddhaghosa. According to this interpretation, the Abhidhamma is a method of unending phenomenological analysis, a practice of continually exploring and then transforming experience. Heim uses the DOI: 10.4324/9781351030908-15

141

Abhidharma Philosophers

canonical Abhidhamma texts to demonstrate how they invite practitioners into a therapeutic and soteriological process of transformation through the endless exploration of experience that demonstrates that there is no final ontology, no final account of ultimate reality. The next chapter, by Rupert Gethin, focuses on Points of Discussion (Kathāvatthu), the fifth of the seven Abhidhamma texts collected in the Pāli Canon. The text is therefore regarded as “the word of the Buddha” (buddhavacana), even as it is also attributed to a particular figure, Moggaliputta Tissa, who is said to have lived in the third century BCE, because he is thought to be correctly explicating the Buddha’s teachings in the work. Through dialogues between one voice advocating the views of the text and another voice presenting the opponent’s views, Points of Discussion addresses more than 200 points of contention in early Buddhist doctrine. The dialogues follow a structure of reductio ad absurdum arguments, in which the opponent comes to understand that their beliefs are wrong because they are inconsistent. We are thus able to gain an understanding of both what some Buddhists were arguing about and the methods of argumentation they were employing. Gethin discusses these methods and then focuses on several particular points of doctrine debated in the text. As Buddhism entered new geographical locations in the Indian subcontinent in its early centuries, different groups formulated distinctive interpretations of the Buddha’s teachings. Abhidharma texts were composed in part to articulate and distinguish the doctrinal positions of diverse monastic communities. In his chapter, Bart Dessein discusses Kātyāyanīputra (ca. first century CE) and his Abhidharma Treatise in Eight Sections (*Abhidharmāṣṭagranthaśāstra or *Abhidharmāṣṭaskandhaśāstra), also known as the Source of Knowledge (Jn͂ ānaprasthāna). Kātyāyanīputra is often regarded as the founder of the Sarvāstivāda school of Buddhism, and his text played a central role in the formation of the Sarvāstivāda identity with its own philosophy and an accompanying understanding of the Buddhist path. In the Abhidharma Treatise, he developed a comprehensive analysis of the material world and defended the role of reason and insight on the Buddhist path. Dessein presents some of Kātyāyanīputra’s original contributions to Buddhist philosophy and traces two distinct developments that emerged from his work: the pedagogical digests that articulate the essence of Sarvāstivāda views compiled in Bactria and Gandhāra, and the commentaries on the Abhidharma Treatise composed by the Vaibhāṣikas of Kaśmīra. In the final chapter in this part, Peter Skilling discusses what he terms the “Vātsīputrīya/ Saṃmitīya” complex of schools – including the Bhadrayānīya, Dharmottarīya, Ṣaṇṇagarika, and Kaurukulla-Sāṃmitīya – which are often referred to by their opponents as “Pudgalavādins.” One of the many views these traditions held was that the person ( pudgala) was neither different from nor the same as the constituent elements of sentient beings, a position that many other Buddhists critiqued as an abandonment of the traditional Buddhist understanding of “non-self.” While the Vātsīputrīyas/Saṃmitīyas probably did not think of this one idea as defining their tradition – the Saṃmitīya metaphysical texts surviving in the original language do not address the question of the pudgala – their opponents did. The Vātsīputrīya/Saṃmitīya were absorbed by other traditions by the sixteenth century, and for many years they have been regarded as not genuinely Buddhist. In contrast, drawing on a variety of sources, Skilling argues that for well over a millennium the Vātsīputrīya/Saṃmitīya schools were in fact a dominant form of mainstream Buddhism in South Asia. With this in mind, he makes the broader point that we should avoid thinking of one or another Buddhist school as somehow original, orthodox, or heterodox, and instead recognize the dynamic changes that have always marked the development of Buddhist thought.

142

9 ˙ GANĪ AND THE DHAMMASAN ˙ GA ˙ VIBHAN The Perfectly Awakened Buddha and the First Abhidhammikas Maria Heim The Enumeration of Phenomena (Dhammasaṅgaṇī) and Analysis (Vibhaṅga) are the first two books of the Pāli Abhidhamma. They are attributed by the tradition to the Buddha (in an account that I will consider briefly in what follows) and, along with the other five books of the Abhidhamma, are considered to be canonical; all seven books comprise the third collection of early Buddhist texts in the Theravāda called the Abhidhamma Piṭaka (the other two piṭakas, or genres of Buddhist knowledge, are the Suttanta, the discourses, and the Vinaya, the monastic rules). There are several good summaries of all seven Abhidhamma books, and I will make no attempt to recap their lengthy contents here (see Nyanatiloka 2008). While traditional exegetes did not consider historical questions of development, holding as they did that all of this material came from the Buddha (even if mediated by disciples and other scholars), modern scholars have tended to see the Enumeration of Phenomena and Analysis as the earliest to emerge. Both texts also have extensive layers of exegesis on them, at both the commentary (aṭṭhakathā) and subcommentary (ṭīkā) layers, as well as medieval compendia (saṅgaha) that attempt to synthesize and summarize the Abhidhamma tradition as a whole. The Abhidhamma has been influential throughout Theravāda history, particularly in Burma/Myanmar, where a lively scholarly tradition on it continues to the present day. In this chapter, I focus on these first two books in order to concentrate on this earliest iteration of what became a long and rich philosophical tradition. My concerns are philosophical rather than historical, and I  will read these texts to ask, first, what they themselves tell us they are doing, and second, how their philosophical implications have been understood by traditional and modern scholars. There have been widely divergent views on what this early philosophical tradition is about, ranging from those who see it as an ontological system positing ultimate reals, to those who read it as a system of phenomenological analysis to investigate and transform experience. While I will show the nature of the scholarly arguments for both interpretations, my own view tends to the latter reading, and I demonstrate why. I also offer additional philosophical possibilities that take special notice of the potential contributions suggested by the early Abhidhamma as a modal analysis of experience. I should note that “Abhidhamma” means the “higher” or “further” Dhamma, or teaching, in that it offers additional development of the basic doctrines of Pāli Buddhism as they are articulated in the suttas, such as the Four Noble Truths, the five aggregates, the twelve-fold DOI: 10.4324/9781351030908-16

143

Maria Heim

dependent origination, the four foundations of mindfulness, etc. Readers new to Indian Buddhism should become familiar with these core doctrines before attempting to understand Abhidhamma and the interpretation of it that I advance here.

Some Preliminaries: Matrices and the Fecundity of Lists The Pāli tradition as represented by Buddhaghosa (the ascribed commentator of the Abhi‑ dhamma texts) holds that the Abhidhamma Piṭaka was discovered and taught by the “Perfectly Awakened Buddha” (Sammāsambuddha). It was discovered during the fourth week of his awakening as he attained omniscience and sat with unblinking eyes contemplating “endless and immeasurable” methods for interpreting experience.1 These endless methods (naya) became the Abhidhamma texts. But how could he teach endless methods? The teaching of them has various accounts, one of which is that the Abhidhamma, being coextensive with the Dhamma itself, was taught when the Wheel of the Dhamma was turned (the suttas contain the same teachings, but in them the teachings are given in the contexts of particular narratives when they are taught to particular interlocutors). Another account, one that suggests that the Abhidhamma can also be conceived as a distinct body of material, has it that the Buddha taught the Abhidhamma to his mother in heaven during a three-month period (where he began to teach her all the lists of phenomena as we see them in the Dhammasaṅgaṇī), while also coming down to earth periodically and teaching it to his disciple Venerable Sāriputta. When teaching to Sāriputta, “the foremost disciple renowned for analysis,” he gave the method “much like one gesturing to the ocean that is seen by stretching out one’s hand while standing at the shore.”2 That is to say, the Buddha initiated through a gesture the oceanic methods of the Abhidhamma to a disciple skilled in analysis who could expand on them. In this reading, the Abhidhamma is the beginning of a series of methods and listings of formulas for analyzing experience that is inherently generative, indeed “endless and immeasurable.” These considerations, while perhaps fanciful from a modern point of view, may be instructive as we begin to interpret the prolixity of lists in this genre of canonical teachings and how we might conceive of them as methods. Indeed, the Abhidhamma texts consist mostly of lists. A key term for understanding Abhi‑ dhamma lists is mātikā, matrix, though the use of mātikā was by no means exclusive to this genre. A mātikā is a table of contents, outline, or listing that can lead to further expository development of the items on it. In both Pāli and English, the word is derived from “mother” and retains the sense of being that from which something further issues (Gethin 1992b, 160–61). Much like a professor jotting down a brief outline of lecture notes which she will then expand in class, a Buddhist teacher would have ready (though without the jotting down since for centuries the tradition was transmitted orally) a list which she could expand via commentary. Of course, this can work conversely: mātikās also function as brief (saṅkitta) summaries or contractions of teachings that are elsewhere expansive (vitthāra), and thus offer the Dhamma in nutshells. Rupert Gethin has shown that mātikās in the Buddhist sources were used not only to flesh out the details of items on the original list, but also to generate further lists. Lists generate other lists until one begins to get a sense of the overall teaching in a highly systematic way. To show this, he starts with the familiar list of the Four Noble Truths. We begin with a list of four items (Suffering, Origin, Ceasing, and the Path), where three of these mention other lists: the Truth of Suffering mentions the five aggregates (form, feeling, perceiving, volitional constructions, and consciousness); the Truth of Origin mentions three types of craving; and the Truth of the Path is, of course, eightfold. Pressing further, each of the aggregates itself subsumes more lists (four kinds of form, three [or six] kinds of feeling, six kinds of perceiving, et cetera) and then each of the constituents of the Eightfold Path is parsed by further lists (right view contains 144

The Dhammasan˙ganī and Vibhan˙ga ˙

the Four Noble Truths [here the parts contain the whole], right thought is the three wholesome thoughts, right speech is of four kinds, et cetera). One could go on. Gethin pursues many of the possible lists (but not all) to which the Four Truths can lead to arrive eventually at all thirtyseven factors of awakening (and even that is an arbitrary stopping place), and an overall sense of core Buddhist teachings.3 The utility of lists in an oral culture relying on memory is obvious and often noted. But further, Gethin argues, this capacity of lists to subsume and generate other lists was invaluable for teaching and understanding in a way that goes beyond rote learning because “the lists help one learn the Dhamma with a view to its inner structure and dynamic” and can “act as a kind of flowchart for the composition of a discourse” (Gethin 1992b, 156). The matrices become guidelines or methods for recitation and composition within the oral performative culture in which teachings were taught and known. If any list can lead potentially to all other lists, the Abhidhamma specialist can speak to any feature of the Dhamma and indeed the overall patterns and connections of the entire system. These features lead Buddhaghosa to insist that only Abhidhamma specialists really know how to preach the Dhamma, because they do not get muddled.4 Another generative listing practice useful for both memorization and teaching is to order teachings according to the lists of “twos,” “threes,” “fours,” and so on. Lists by numbers, often in a question-and-answer format, is a very common practice in the suttas, as we see in the “Chanting Together” (Saṅgiti) and “Expanding Tens” (Dasuttara) suttas, and of course in the entire structure of the Aṅguttara Nikāya.5 This provided not only an orderly and systematic teaching, a memory aid, and a collective and performative style of preaching (how many twos can we remember and recite together?), but it also allows one to see new identifications and relationships and to improvise within the terms of the larger system. The “Great Questions” sutta, for example, is the improvisational expansion of the Buddha’s teaching by a pundit nun from Kajaṅgalā. She is asked by lay people to preach on a list of questions about ones, twos, threes, et cetera. Claiming that she had heard neither the Buddha nor the monks expand on these particular points, she offers to explain, in detail, how she understands these questions. She expands the lists in a way that uses the numerical schema to discuss core teachings. The lay people honor her for this and she invites them to check her answers with the Buddha, who confirms that he would have answered exactly as she did.6 Finally, if we are right in finding the kernel of Abhidhamma practices in some of the sutta literature, another purpose of lists was for the Buddha to teach his disciples the skills to analyze experience from multiple angles and in multiple ways. Such aspectual and modal interrogations of experience are important, both in the contemplative practices he advocates and in teaching. The latter can be shown in “The Many Types of Elements Sutta,” where the Buddha urges his disciples to train to become “learned” ( paṇḍita) and “inquisitive” (vīmaṃsaka). They do so by becoming “skilled” (kosala) in the elements, bases, and dependent origination (these fundamental teachings, some discussed in what follows, are not necessary to elaborate here for the purpose at hand). Ānanda takes this to heart so that as the Buddha begins to give numerical listings of the elements, Ānanda asks him, repeatedly, “but venerable sir, is there another mode of teaching ( pariyāya) whereby one can become skilled with the elements?” In each case, the Buddha says yes, and gives a different numerical listing. It is thus through Ānanda’s persistent questions that we get first eighteen elements, then a different listing of how the elements can be described by way of six, and then a different list of six, then yet another list of six, then a breakdown of them into three, then into two.7 This exercise suggests two main points to bear in mind as we enter this literature. First, lists of phenomena can be variable: there is no single list of the elements because they can be divided up differently according to different methods 145

Maria Heim

of looking at them. Second, there is a valued skill in questioning to arrive at different modes of breaking down and teaching about experience. A skillful inquisitor does not stop with one list.

The Enumeration of Phenomena This first text of the Abhidhamma Piṭaka, the Enumeration of Phenomena (hereafter, Enumeration) is structured into three books. Book I begins with a matrix of twenty-two triplets (that is, lists of threes) and one hundred pairs (lists of twos). It then takes up the first list of triplets which are three questions: “which are the good (kusala) dhammas? Which are the bad (akusala) dhammas? Which are the indeterminate dhammas?”8 The text itself does not define dhamma or kusala, but at this stage in the literature, dhamma seems to mean “object of experience” that one can observe in contemplative introspection.9 “Phenomenon” will serve as a translation. For its part, the commentary adds the nuance that these are not essentialist or living entities.10 As for kusala, the commentary defines it as “salutary, blameless, competent, and producing happy results,”11 a range which I attempt to capture quite generally as “good.” Akusala, “bad,” is the opposite of this, and indeterminate are undetermined or neutral. Answering these three questions of this first matrix takes up half of the entire Enumeration, as it proceeds to ask about specific types of moments of awareness (citta). For example, it begins by taking up a type of good awareness that is associated with sensual desire and accompanied by joy and knowledge. What are the good phenomena? On whatever occasion there is the arising of a good awareness of the sensory realm accompanied by joy and knowledge, having as its object a (visual) form, sound, smell, taste, touch, or mental phenomenon, then at that occasion there are: contact, feeling, perceiving, intention, awareness, initial thinking, sustained thinking, joy, pleasure, oneness of mind, faculty of faith, faculty of energy, faculty of mindfulness, faculty of concentration, faculty of wisdom, mental faculty, faculty of happiness, faculty of vitality, right view, right thought, right effort, right mindfulness, right concentration, power of faith, power of energy, power of mindfulness, power of concentration, power of wisdom, power of shame, power of apprehension, non-greed, non-hatred, non-delusion, non-covetousness, non-malice, right view, shame, apprehension, tranquility of body, tranquility of mind, lightness of body, lightness of mind, softness of body, softness of mind, workableness of body, workableness of mind, proficiency of body, proficiency of mind, uprightness of body, uprightness of mind, mindfulness, meta-attention, calmness, insight, exertion, balance, and whatever other dependently-arisen formless dhammas occur on that occasion – these are good phenomena.12 This particular occasion of experience can include at least these fifty-six phenomena (dhammas), though it need not include all of them; elsewhere, these types of dhammas are referred to as cetasikas, psychological phenomena. The rest of Book I lists the cetasika dhammas occurring in various types of good moments of awareness, then in the various occasions of bad moments of awareness, and then in occasions of the indeterminate, and in every case working these classifications of experience through the distinctions that occur in the other triplets and pairs, as well as many others. There are several things to notice about this first list of fifty-six dhammas. First, most of its first few items have precedent in the “One-by-One Sutta,” which describes Sāriputta listing the phenomena he experienced and identified “one by one” during his meditation experiences. His 146

The Dhammasan˙ganī and Vibhan˙ga ˙

introspection revealed that, among other phenomena, contact, feeling, perceiving, intention, awareness, initiative, resolve, energy, mindfulness, equanimity, and attention arose, endured, and faded away in his jhānic meditation.13 While initiative (chando), decision (adhimokkho), and attention (manasikāro) enter into these Abhidhamma lists only at the postcanonical level, Sāriputta’s observations might be the seed of this method of cataloging the phenomena present through introspective practice. Second, one should notice that some of the items repeat under different modes. For example, “mindfulness” (sati) occurs no fewer than four times: as a faculty, as a power, as “right mindfulness” (that is, one of the Eightfold Path factors), and listed on its own. This might be seen as unnecessarily redundant: should not a single mention of mindfulness (sati) be enough? Buddhaghosa resists forcefully the idea that the list is baggy or ill-considered because each of the different groupings and modalities (mindfulness as a power or a faculty or a path factor) speaks to different aspects or workings of these phenomena. They are like artisans who work in different guilds in which their functions and roles vary: the “same” artisan can work as a carpenter, plumber, et cetera, for the king.14 We might say that mindfulness might be present in a particularly strong way in a particular occasion of this type of experience, whereby it operates as a faculty governing the experience as a whole. Or in a different instance of a similar occasion of experience, mindfulness might not function as a ruling faculty but rather as a mild trace of a path factor. These differences suggest that these various classifications of how a phenomenon can present were intended to be captured in the list itself. The modern interpreter Nyanaponika Thera argues similarly that these repetitions under different aspects are not dispensable (as the later summarizations of the tradition suggest) but rather indicate “the different functions and ways of application of a single quality,” and are essentially practical in orientation for the psychological and transformative purposes at the heart of the whole business. He cites a modern psychologist to indicate the importance of this – “in psychology a difference in aspects is a difference in things” – and Nyanaponika explores with great nuance the implications of this aspectual or modal method of analysis.15 Another key feature of the list of fifty-six phenomena in this occasion of good awareness is that it ends with an “et cetera.”16 The list is not described as final or complete. Indeed, all lists of phenomena occurring in particular occasions of awareness (bad and indeterminate, too) in this text (and there are a good many) end with “et cetera” in this way. Finally, note that by the end of this list, these phenomena are described as “the good phenomena” (ime dhammā kusalā). This is important because many of the same items (contact, feeling, perceiving, intention, and so on) will appear again when the text begins to list the varieties of bad and indeterminate moments of awareness;17 there they will change normative valence and become bad and indeterminate phenomena. Again, this feature has been noted by Nyanaponika, who argues for the importance of these groupings for indicating how items in them are “open” and changeable according to what else occurs on the occasion in which they arise. This builds into the system a resistance to seeing dhammas as single, discrete, self-contained units because their functions, intensities, roles, and valences change according to whatever else occurs in the occasion of experience being considered (Nyanaponika 1998, 40–41). Book II takes up matrices for analyzing form (rūpa). Here it is important to note that form is not “matter” in the sense of the physical stuff of the world, rendered in a reductionist attempt to get at atoms, as it is sometimes understood. The text’s many matrices defining form labor in much the same vein of enumerations that classify and elaborate it according to single definitions, pairs, triplets, and so on up to lists of eleven. Throughout, rūpa consists of dhammas occurring always and only with other dhammas (of the other four aggregates, often grouped together as that part of our phenomenality associated with naming [nāma] experience), and 147

Maria Heim

is analyzed as it is experienced by our six senses, karmic results, and other psychological factors and features of experience. For his part, Buddhaghosa interprets rūpa as that part of human phenomenality that is impacted (literally “molested,” ruppana) by sensory contact. For example, when we begin to analyze our experience of forms, we find that the “earth element” is experienced in terms of its hardness, how it acts as a foundation, and how it receives other things; and the “water element” is analyzed by its flowing, spreading, accumulating, and so on.18 This is not a physics of matter and processes that occur “out there” independent of an observer, but rather an analysis of the human experience of them – how they are for us. Book III is a summary, but instead of reducing or tidying up the numerous analytical matrices and practices in the first two-thirds of the text, it offers yet further matrices and classifications that consider dhammas in terms of yet further modalities and aspects. These groupings consider various kinds of phenomena in terms of their association with the causes (hetu), conditions ( paccaya), hindrances (nīvaraṇa), fetters (saṃyojana), oozings (āsava), views (diṭṭhi), forms of clinging (upādāna), defilements (kilesa), path factors (magga), and so on. In other words, it appears that every core teaching, classification, and distinction given in the Suttanta finds a place in the Enumeration as a mode of analysis of dhammas.

Analysis Our second text, Analysis (Vibhaṅga), is structured rather differently than Enumeration, but it interacts with it in interesting ways. It consists of eighteen chapters that treat, with great analytical depth, core teachings found in the suttas. These are: aggregates, bases, elements, the Four Truths, the faculties, dependent origination, the four foundations of mindfulness, right striving, the bases for magical power, awakening factors, path factors, the jhānas, the immeasurables (that is, the divine abidings), the five moral precepts, the four kinds of discriminating analysis, understanding, miscellaneous items, and an analysis of the “heart of the Dhamma.”19 These lists function much as we have described previously with the help of Gethin as lists leading to further lists in a highly generative way that can develop one’s understanding of both the intricacies of the details and the overarching structure of the teachings. The lists also develop a modal treatment of experience as it is interpreted by those teachings. The text is very systematic in that each chapter has three parts: Suttanta Analysis, Abhi‑ dhamma Analysis, and Lines of Questioning. The first is an analysis that picks up on classifications and distinctions mentioned in the Suttanta, but according to Buddhaghosa, this is only a partial analysis. The fuller analysis is given in the Abhidhamma Analysis which is more elaborate, looking at the category under analysis from “every side,”20 by offering matrices of single definitions, pairs, triplets, and so on. The Lines of Questioning will be familiar from the questions of the Enumeration beginning with which items are good, which are bad, and which are indeterminate, and going on from there to offer additional matrices of the twenty-two triplets and one hundred pairs used in that text. So in this generative way, Analysis puts to creative use for its groups and items the same method that the Enumeration uses to treat dhammas. Of course, many of the items in Analysis’s teachings are themselves dhammas. There is space only for a brief example of the first chapter, the analysis of the aggregates. This offers first a relatively short Suttanta Analysis of each of the five aggregates: form (rūpa), feeling (vedanā), perceiving (saññā), volitional constructions (saṅkhāra), and consciousness (viññāṇa). The five aggregates doctrine is of paramount importance because it breaks down and groups the entire experience or phenomenality of a human being into these five clusters of phenomena. Beginning with form, the Suttanta Analysis breaks it down further into what is experienced as past, present, and future – according to whether it is experienced subjectively 148

The Dhammasan˙ganī and Vibhan˙ga ˙

or by others, whether it is gross or subtle, inferior or superior, and distant or proximate. (Notice that these analyses of rūpa, like those in Enumerations, do not itemize the physical stuff in the world, but rather list modalities of human experiences with this cluster of phenomena; again, this is because rūpa refers not to matter but to one side of our phenomenality.) The Abhi­ dhamma Analysis piles on more analyses according to single ways of describing form, and matrices of pairs, triplets, and up to elevens; the Lines of Questioning runs rūpa through the twenty-two triplets and one hundred pairs. Then the chapter takes up feeling, the second of the aggregates, and analyzes it by the same modes in all three parts that it did with the treatment of form. And so on for the other three aggregates. And this is only the first chapter of Analysis – there are seventeen more. Each chapter deploys somewhat different modalities as are appropriate to it, and anyone with an interest in any of the eighteen teachings would be well served by exploring these chapters that analyze it exhaustively (or exhaustingly, as the case may be). The final chapter, the “Heart of the Dhamma,” circles back to many of the categories the whole text has covered, to ask about further types and classifications of the aggregates, bases, elements, and so on. As in the Enumeration’s final book, something that looks like it might summarize winds up elaborating further.

Philosophical Interpretations Abhidhamma as an Ontology of Ultimate Reals To what end, all this analysis? There is a substantial body of scholarship that treats the Pāli Abhidhamma as, at bottom, a metaphysics or ontology of ultimate reality, and in fact we can say that this is the dominant position in both traditional Theravāda scholarship and modern Buddhist studies. To begin with representatives of the latter, contemporary scholars of Buddhist philosophy have not always made much of a distinction between the Pāli Abhidhamma and the Sarvāstivāda Abhidharma on this point, and some have assumed that what holds for the latter also describes the former. For example, Jan Westerhoff asserts that while there emerged different Abhidharma canonical traditions “all were united by a common core of philosophical principles,” which include a notion of “primary existent objects” that are “ultimately real.” Because these ultimately real objects, dharmas, are said to have an “intrinsic nature” (svabhāva), they “exist no matter what, without depending on the existence of any other dharma or on any conceptualizing mind” (Westerhoff 2013, 130). Other texts make the same move in conflating the Abhidharma/Abhidhamma systems and characterizing them as offering, as The Princeton Dictionary of Buddhism puts it in its single entry on both, “an objective, impersonal, and highly technical description of the specific characteristics of reality” (my italics) (Buswell and Lopez 2014, 4); when this dictionary turns to the Enumeration it treats its purpose in the same vein as “a systematic analysis of all the elements of reality” (241). While Paul Williams distinguishes between the Sarvāstivādin tradition and the Theravāda, and suggests that the latter might be characterized as an “event ontology” rather than a “substance ontology,” he is insistent that Abhidhamma analysis involves “seeing things as they are, and that is a matter of ontology” (Williams and Tribe 2000, 92). For many scholars, the matter is settled. As we turn to scholars trained within the Pāli tradition, we find a long history of an ontological reading of the Abhidhamma dating at least to the medieval compendium literature, and likely initiated in some of the layers of exegetical tradition that precede it. Far and away the most significant development after the first commentaries is the Compendium on the Meaning of the Abhidhamma (Abhidhammatthasaṅgaha), attributed to the eleventh-century monk, Ācariya Anuruddha. This text is the most widely known Abhidhamma text and has almost entirely 149

Maria Heim

supplanted the canonical literature in both monastic curricula in Theravāda countries and modern Western scholarship. Indeed, it is this text’s system of Abhidhamma that is regularly cited and discussed as the Pāli Abhidhamma in textbooks and scholarly works, despite the fact that it differs in several highly significant respects from the canonical literature (a point that is itself seldom noted). For example, textbooks that cite a standard enumeration of Abhidhamma categories of eighty-nine classes of cittas and fifty-two cetasikas are drawing from the Compendium rather than the Enumeration or any other canonical listing, where the lists do not correspond exactly with this system.21 The Compendium owes its reputation in part to its concise and highly systematic presentation of what had grown to be a huge and unruly body of material when we consider not just the seven books of the Abhidhamma but the many layers of commentary and subcommentary that followed in their wake. The Compendium is only about fifty pages, making it a handy and useful summary and primer. While this is not the place to compare systematically this text with the Enumeration, I can mention several features of it that differ from its canonical predecessor, and these features suggest a far more ontological orientation. As we have seen, the Enumeration lists its phenomena always in reference to questions about what occurs in particular occasions of momentary awareness; for example, the list we began with previously describes a certain specific type of momentary awareness, an “occasion” in which some combination of fifty-six or more dhammas could occur. Other lists describe bad or indeterminate moments of awareness, different occasions of experience. In other words, the text does not attempt a single generic list of items of experience abstracted from the occasions in which they occur. But the Compendium tidies all of this up to offer a single abstract list in an arrangement of eighty-nine cittas and fifty-two cetasikas. The Compendium also adds distinctions not present in the canon, such as a category of “beautiful cittas,” and a distinction between universally present and occasional cetasikas. Further, when we considered the first kind of awareness in the Enumeration, we noted that some fifty-six cetasikas (and counting) could be present in that particular type of good awareness and that some of the items could be said to be repeating under different aspects (mindfulness, for example, appears four times). In the Compendium’s generic list of fifty-two cetasikas, all such apparent repetitions have been pruned away, and the list is presented without the possibility that it could be supplemented; it appears to be closed and total. Moreover, even a cursory glance at the two lists reveals that a not insignificant number of the cetasikas differ from the lists of cetasikas occurring in the Enumeration, as for example: the Compendium does not list citta itself (it may have seemed redundant to treat citta as both a cetasika and a moment of awareness comprised of cetasikas), calm meditation (samatha), and vipassanā. Faith, mindfulness, energy, and others are listed once as such, but not in their aspects as powers and faculties. Among akusala dhammas, missing in the Compendium but present in the Enumeration are wrong thought, wrong effort, wrong concentration, and some of the dhammas acting as powers and faculties (such as the power of energy). In addition, the Compendium includes items that are not in any of the Enumeration’s listings, and came into the Abhidhamma tradition only with the aṭṭhakathā, including attention, resolve, initiative, and among bad cetasikas, conceit, envy, avarice, rigidity, and sluggishness.22 Doubtless, the field would benefit from a more sustained comparative treatment of these two texts and the implications of their differences than I can offer here, but I have tried to suggest enough of these differences to suggest that we are quite possibly dealing with different systems. Of course, some may view these differences as simply the Compendium’s more rational, simplified, and elegant rendering of the same basic text. To be sure, it lists many of the Enumeration’s distinctions and classifications that may capture its complexity when put into action ramifying these classifications with its basic list of dhammas. And we have no standing to require that the “tradition” should adhere to the canonical formulations rather than what 150

The Dhammasan˙ganī and Vibhan˙ga ˙

it may consider to be a more concise maturation of it. But the differences in both content and form are significant enough that they suggest that modern scholars may wish to avoid speaking of a single Abhidhamma theory, represented by this medieval summary of it. In terms of content, the Compendium presents new and different dhammas and new distinctions and categories. Perhaps even more significantly, in terms of formal presentation it presents a complete, abstract, “view-from-nowhere” list, and many of the features of the modal quality of presentation (such as how a particular phenomenon occurs under different aspects and operating in different modes, such as powers and faculties) are no longer evident. These formal features suggest not just differences in the details, but a substantial development of the tradition over time away from what appeared as open-ended sets of modal analytical practices to a more reductive, closed, and complete system. The differences and tensions between a total and totalizing system and an open-ended one deserve further scholarly scrutiny. Perhaps the Compendium’s most significant ontological move is the use of language that was not present in the Enumeration but that came to be used and then invested with ontological significance in later exegetical layers. Two critical terms in this regard are paramattha (ultimate meaning) and sabhāva (the particular nature of a thing). Neither is present in the Enumeration or Analysis.23 We can begin with paramattha. The Compendium frames its project in terms of listing the four main types of phenomena from an “ultimate” standpoint: citta, cetasika, rūpa, and nibbāna, which then receive further elaboration as they come to structure the whole system. In earlier commentarial literature, paramattha just meant a kind of ultimate, analytic, “furthest-sense” language, but there is reason to believe that by this medieval text (and likely much before in the ṭīkā literature), it had come to refer to ultimate reality – the things or events that really exist. We have traveled considerably from the early usage of dhamma as merely an object of experience. The Compendium says that “in this way the Tathāgatas reveal what is ultimate as four: awareness, psychological phenomena, form, and nibbāna.”24 This assertion suggests that Tathāgathas are not just speaking in furthest-sense language, but rather are revealing ultimate reality. The text also speaks of paramattha in terms of what exists when it contrasts dhammas, such as form and feeling, with what is not found to exist in this sense, such as earth and mountains.25 This suggests a distinction about how our concepts capture what really exists, and the terms that capture what really exists are dhammas like rūpa and vedanā. It now becomes possible to refer to “ultimate dhammas” ( paramatthadhamma) as the commentaries on the Compendium go on to do.26 In his commentary on the Compendium, modern interpreter Ledi Sayadaw takes himself to be offering “an explanation of the ultimate” (Sayadaw 1913–1914, 129; see Braun 2015, 20–21). The ontological interpretation is of course a very powerful religious and philosophical claim: the Buddha, in this highest form of his teaching, declared and revealed in the most direct terms what really exists. Another crucial term in the ontological shift is sabhāva, which is not mentioned by any of the canonical Abhidhamma texts. It enters the Pāli tradition in the aṭṭhakathā as referring to the particularity or specific characteristic a thing has that makes it different from other things, and it can be used very broadly in this way. But after Buddhaghosa, the term sabhāva came to be freighted with maximalist ontological significance to mean the “intrinsic nature” of something (much as the Sanskrit term svabhāva came to be understood by the Madhyamaka critics of the Abhidharma traditions). Further, dhamma comes to be identified with sabhāva, so dhammas come to be that which have their own, intrinsic nature.27 While the Compendium is terse enough to not always be entirely clear about what it means by such terms, it does indicate that all of the fifty-two cetasikas, for example, are established “by way of their particular nature.”28 The project has come to involve listing phenomena according to their intrinsic natures, and so we come to find modern accounts that speak of dhammas as “things in themselves, ultimates.”29 Whereas 151

Maria Heim

the Enumeration began to list phenomena that can occur in particular occasions of awareness and that must be analyzed in a modal way, the Compendium simply lists the bare phenomena themselves, recruiting the notion of paramattha to indicate ultimacy and the notion of sabhāva to indicate an intrinsic nature. Y. Karunadasa characterizes the “dhamma theory” that has at this point emerged in strongly ontological terms, although he acknowledges that this theory was “not precisely articulated” in the canonical texts.30 Combined with the ontological sense of paramattha, the theory is now seen to be positing “ultimate existents” that can be said to have “objective existence,” and the ṭīkās begin to speak of “ultimate intrinsic realities” ( paramatthasabhāva).31 A dhamma now represents “a fact having an objective counterpart” and “an actual datum of objective experience” (Karunadasa 1996, 18). For Karunadasa, dhammas are the “ultimate irreducible data” of cognition and objective existence, and the “ultimate elements of existence”; they are “not amenable to further analysis” (Karunadasa 1996, 5). He further suggests that the tradition can be considered a kind of “critical realism” in that it “recognizes the distinctness of the world from the experiencing subject” and that “the dhammas do not exist in dependence on the operation of the mind” (Karunadasa 1996, 28). While I think that much of this language considerably overstates what we actually see in the literature (especially the idea that the dhammas are somehow independent of experience and not amenable to further analysis), I think he is right to say that the ṭīkā literature, as represented here primarily by the Compendium, has generated an ontology of ultimate reals as the highest teaching that is the Abhidhamma.

Abhidhamma as Endless Phenomenological Analysis While the preponderance of tradition and scholarly weight falls on the side of the ontological reading, we can identify an entirely different set of choices going as far back as the aṭṭhakathā (the earliest postcanonical exegetical layer that we have). This material – codified in the fifth century, likely by a team of scholars headed by Buddhaghosa – emphasizes not ontology but analysis in its reading of the Enumeration and Analysis. Before turning to Buddhaghosa’s reading of the Abhidhamma, it may be useful to recap several features of the canonical texts themselves that suggest that assuming them to be fundamentally ontological in nature and purpose may be misleading. As we have seen, Abhidhamma matrices, which generate other lists to expound the teaching and its structures, continue ever more finely grained analyses and questioning “from all sides” in what Analysis calls the Abhidhamma Analysis. I hope that even in my brief synopses of both books, it is clear that analysis begets analysis without ever landing on a final, single, ultimate list of phenomena as the point or outcome of all the many modes of analysis these books deploy. The point rather seems to be to generate analyses that consider and reconsider teachings about experience in a modal and aspectual way. I have also mentioned the practical nature of these methods of analyzing in contemplative practice experience. They were used not to canvass all of psychological life and the material world, but rather to examine further the particular teachings the Buddha gave in the suttas to observe and change one’s experience for therapeutic and soteriological purposes. The distinction between conventional teachings and ultimate teachings occurs first in the Kathāvatthu (and is not mentioned in any of the other six canonical books) and then is picked up at the aṭṭhakathā layer, where it refers not to different kinds of truth or reality, but to different registers of the Buddha’s teachings. In the Kathāvatthu the distinction can be read not as describing ultimate reality per se, but as indicating how one arrives at different terms: “does one arrive at ‘person’ by means of the furthest-sense, by what is realized?”32 The answer is no, 152

The Dhammasan˙ganī and Vibhan˙ga ˙

because “person” does not survive the analytic dismantling of it into parts, such as the aggregates, so it is not a paramattha term. By contrast, the technical language of the Abhi­dhamma, such as aggregates, dhammas, form, feeling, et cetera, are arrived at via furthest‑sense ( paramattha) analysis. (This does not entail, however, that any mention of “person” ( puggala) is inappropriate,33 at least not when we consider that an entire book of the Abhidhamma, the Puggalapaññatti, explores the nature of the person, and not just reductively, but also socially, morally, in terms of progress, and so on). For Buddhaghosa, paramattha teachings are given in a register that is technical and analytical – language in the “furthest sense.” This is contrasted with the Buddha’s more ordinary language that is given in conventional (sammuti) or customary (vohara) teachings. He is insistent that the distinction refers to different types of the Buddha’s language (bhāsā) and teachings (desanā), and not to different kinds of truth (sacca).34 One would not say that ultimate language is true and conventional language is false, because both refer to the Buddha’s teachings and are thus (for him) both incontrovertibly and unqualifiedly true. It is not an ontological distinction, and the mention of paramattha is not an occasion to describe reality. It may seem like a short step to say that something that cannot be analyzed further is what actually exists, but this is precisely what the canonical materials, and Buddhaghosa, nowhere say (and indeed, they never grant that these categories cannot be analyzed further). Buddhaghosa mentions sabhāva but does not imbue it with ontological significance, and he uses the term quite broadly. For example, he describes the different sabhāvas or particularities of each of the decaying corpses in the “meditations on the disgusting” in the Visuddhimagga. When he does so, he is not mounting an argument that these different types of decaying corpses are final units of analysis and thus ultimate reals; rather, this is part of a contemplative practice identifying particular features of a meditation object. Dhammas, too, can be defined in terms of their particularities because it can be useful to define a thing by its specific characteristic, as we see here: “dhammas have particularities (sabhāva); alternatively, dhammas are brought about by conditions, or dhammas are brought about by their particularities.”35 The word dhamma can be understood in terms of sabhāva, but for him, it does not mean “essence” or “intrinsic nature,” but only a thing’s particularity whereby it can be defined as different than something else. The claim that dhammas can be “brought about” by either their own particularities or by conditions itself rules out that they are self-existent, sui generis phenomena. It says that they can be understood or grasped by either their particularities or by how they are conditioned by other phenomena. Further, Buddhaghosa emphasizes that dhammas are “empty.” Discussing the “Section on Emptiness” (suññatavaro) in the Enumeration, he says this: “These are just dhammas” is mentioned to show that, due to their emptiness, “they are only dhammas, without essence, without a leader.” Therefore, the meaning is to be understood in this way: on an occasion that the first main good awareness arises in the realm of sense desires, on that occasion fifty-six dhammas arise by virtue of being constituents of that awareness, in the sense of their particularities.36 Dhammas have particularities but are without essences due to their emptiness. Of course, emptiness as a mode of analysis had long been part of the canonical Abhidhamma and the Paṭisambhidāmagga (which explicitly discusses the “emptiness of sabhāva” in the case of each of the five aggregates), and Buddhaghosa follows suit.37 While Buddhaghosa’s interpretation of these two terms, paramattha and sabhāva, begins to show that he reads the tradition differently than do his successors, there are other features of his 153

Maria Heim

interpretation relevant to the point. As we have seen, he reads the Abhidhamma texts as an “endless and immeasurable” series of methods and practices that begin to give room for the Buddha’s active and dynamic omniscient practices of knowing things without obstacle. He identifies his school of commentary as the Vibhajjavāda, which among other things entails that one steers clear from launching one’s own view and instead “takes up the meaning of a text and then returns again to that meaning by explaining it with different methods ( pariyāya).”38 We have already had occasion to note the “Many Types of Elements Sutta,” in which training in being inquisitive about multiple modes of teaching and numerous lists is a fundamental skill for a teacher; the “Many Types of Feeling Sutta” does similar work with feeling (vedanā), showing that this dhamma may be sliced and diced variously and is hardly irreducible (see Heim 2021). This is the spirit and style in which Buddhaghosa read the canonical texts and carried out his own analyses in the Visuddhimagga. I have elsewhere argued that when we take the time to work through Buddhaghosa’s theory of scripture and the interpretative practices he engages in, we see him articulating a view of the Abhidhamma Piṭaka as a kind of “phenomenological analysis,” a modal examination of experience through methods designed to explore experience and change it (Heim 2018; Heim and Ram-Prasad 2018). He is not engaged in metaphysical or epistemological arguments. Among modern scholars, as may be apparent, Nyanaponika Thera shares much of this view in an elegant and insightful discussion focusing on the Enumeration. He describes the canonical Abhidhamma as a whole as a phenomenology in that it deals with “phenomena, that is, with the world of internal and external experience,” which he contrasts with “ontology, or metaphysics, that inquires into the existence and nature of an essence, or ultimate principle, underlying the phenomenal world” (Nyanaponika 1998, 190). He modulates this to some extent, however, by suggesting that the practices of analysis in the Abhidhamma do in fact aim at contributing to “ontological problems,” namely that it shows that in reality there no abiding essence to be found in the world of experience (Nyanaponika 1998, 21); and he does slip back into talking about “reality” from time to time (though he is careful to say that it deals with “actuality from an exclusively ethical and psychological viewpoint and with a definite practical purpose”) (Nyanaponika 1998, 2, 46). Notably, Bhikkhu Bodhi, in his introduction to Nyanaponika’s book, modulates the phenomenological reading further, and works it back to ontology, asserting that “the Abhidhamma draws up a list of ontological actualities” and does in fact discriminate between what is real and what is only apparently real (Nyanaponika 1998, xvi – xviii). It is important to many that the Abhi­ dhamma be seen to be the Buddha offering an account of ultimate reality, the way things really are. My own reading is to take seriously the possibility that the intellectual purpose of the canonical Abhidhamma is to offer a complex set of analytic practices and methods that allow the practitioner to explore the many – perhaps infinite – facets of experience without ever landing on a final, single, essential list of the contents or aspects of it. The lists aim not to arrive at a single, irreducible account of the nature of our experience, but rather to engage in the methods that would resist such a final account (and in this sense, the later tradition contravenes this spirit). The lists do not seek to provide an account of reality, even the reality of experience, but rather to provide methods and practices of analysis as part of the contemplative and analytical purpose to inquire into and transform experience. This is philosophy of a different sort than metaphysical assertion and argument. It is a practice of exploring and transforming experience within the therapeutic and soteriological aims of Buddhist dogma. Part of how I have arrived at this view is that I find much of the modern ontological account unconvincing, either as a reasonable account of the canonical texts or in terms of how it could work philosophically. We may begin with some of the textual considerations. If the Enumeration was attempting to arrive at a final list of irreducible dhammas, why did it not simply and clearly state, at some point, what these are? Given that it emerged in the ancient Indian context 154

The Dhammasan˙ganī and Vibhan˙ga ˙

in which sophisticated ontological systems were ubiquitous, including other traditions of Buddhist thought, why did it not avail itself of such ontological terminology? For that matter, why not define dhamma in terms that would make clear that it is an element of ultimate reality, instead of assiduously avoiding language that could be (and later was) construed in ontological ways (sabhāva, paramattha)? It was either innocent of such ideas, or it deliberately avoided them. If we are to be persuaded that the later medieval tradition seamlessly represents the canonical tradition, we need to see how exactly this works in a treatment that first reckons with the obvious differences in the texts. Philosophically, I find some of the readings we have explored about dhammas either undefended or incoherent. We can begin with widespread slippage in the secondary scholarship that begins by noting that the dhammas are objects of experience as they are conveyed by core teachings, but then slides into asserting that they refer to elements of reality, without ever explaining why this must be. It is worth recalling that the original matrices provided lists of doctrines, and doctrines are always practices – the Four Noble Truths is a practice of diagnosing experience and understanding its conditions; the five aggregates is a practice of dismantling the idea of person; dependent origination is a practice of understanding mutual causes and conditions in analyzing experience. The items that come to be treated as dhammas are taken from these lists of practices: things like feeling, perceiving, mindfulness, path factors, and so on. These come to be objects of experience in introspective methods designed to recognize them “one by one,” to see their conditionality, and to transform the practitioner’s phenomenality. This makes it most unclear what it would mean to say that dhammas exist, as we saw previously, “objectively” or that they exist “no matter what, without depending on the existence of any other dharma or on any conceptualizing mind” as “impersonal,” untethered “from any experiencing subject.” Nor is it clear how dhammas can be “irreducible” or “not amenable to further analysis” when the canonical texts can always be counted on to provide another list that analyzes, in another mode, each of the phenomena it takes up – elements can be eighteen, or six, or a different six, or a still different six, or three, or two; feelings can be carved up variously and reconsidered by lists of two, three, five, six, eighteen, thirty-six, one hundred and eight, and so on (Heim 2021). Rather, the question is: where, exactly, do these analyses end? How do we find a stopping point that is not arbitrary? I think Gethin is closer to the mark when he notes the danger that comes from breaking things up into parts in that “we might then take the parts as real and begin to reify the world again, if in a different way.” It seems to me that the early Abhidhamma authors sought to avoid precisely this same danger through the elaboration of the various mātikās. Try to grasp the world of the Dhammasaṅgaṇi, of the Paṭṭhāna, and it runs through one’s fingers.39 He goes on to emphasize the use of the mātikās in “practical psychology,” in a system “primarily concerned to distinguish states and processes of mind on the basis of actual observation.” And he is right, I think, to note that one often encounters details that are beautiful and profound (though other tasks at hand have precluded space for me to begin to do justice to these). Finally, I remain deeply intrigued by the philosophical potential of a modal and aspectual exploration of experience. I am persuaded by Buddhaghosa and Nyanaponika Thera that proliferating groupings and classifications of phenomena can help resist essentialist accounts of what we experience. Our feelings and sensations and moral dispositions do seem, phenomenologically, to be open and changeable in their workings, valences, and intensities to whatever else occurs with them in any given occasion of awareness. Is “pleasure” (to take at random just one dhamma explored in the matrices) really the same thing in every instance in which it occurs? 155

Maria Heim

Do “right view” and “right thought” (to take two more) have single uninflected unchanging natures across all moments of experience in which they might occur? The abstractions that we use to get at our experience – whether we are talking about the solidity we experience in the things we touch or the hedonic impact we encounter when we feel – are always going to be at some remove from the singular particularity of our actual moments of experience. But it could be that those abstractions do a more useful job of describing experience when they are subjected to numerous, perhaps even endless, modes of inquisitive analysis that classify and reclassify, divide and redivide, consider and reconsider, than when they aim to land on a single, final, highly abstract, essentialist, objective, and ultimate “own nature” that they must be. The early Abhidhamma texts teach us that the philosophy of experience can be an “endless and immeasurable” application of methods useful for taking up concrete moments of experience. This suggests more of a process of analytic work than a final description of reality. Scholars of the early Western phenomenological tradition might see affinities in their approach and this repeated application of phenomenological questions that remain open to further analysis. In the Buddhist tradition, however, this philosophy as practice is, of course, ultimately therapeutic as its methods provide programmatic ways to examine and then fundamentally change the patterns of experience.

Notes 1 See Atthasālinī (As) 14–17 (Tin and Rhys Davids 1976, 16) for the details in this paragraph; see also Heim (2018, 40–44, 154–63) on the authorship and teaching of the Abhidhamma as Buddhaghosa understood these. “Endless and immeasurable” is ananto aparimāṇo. While all translations are my own unless otherwise noted, I also cite the available English translation. All abbreviations follow the Pāli Text Society. 2 As 16 (Tin and Rhys Davids 1976, 20): Evaṃ sammāsambuddhe nayaṃ dente paṭisambhidāppattassa aggasāvakassa velante ṭhatvā hatthaṃ pasāretvā dassitasamuddasadisaṃ nayadānaṃ hoti. 3 Gethin 1992b, 150–56. My translations are slightly different than Gethin’s, most notably, “perceiving” for saññā instead of “recognition” (though he is of course right that it involves identifying and naming what one perceives). 4 Atthasālinī 29 (Tin and Rhys Davids 1976, 37). 5 There is substantial scholarship discussing the origins of the Abhidhamma in terms of suttas that are similar to, or precursors for, some of what we see in the Abhidhamma Piṭaka; for some of it see Anālayo (2014) and Cousins (2015). Both are also very helpful on the oral and performative aspects of these practices. 6 A v.54–59 (Bodhi 2012, 1376–79). This the second of two suttas called Mahāpañhā. At the conclusion of Kijaṅgalā’s teaching, the Buddha calls her “paṇḍitā,” a learned pundit. 7 Bahudhātuka Sutta (M iii.61–67; Ñāṇamoli and Bodhi 1995, 925–30). Briefly, the eighteen elements (dhātu) are each of the six senses (mind is a sense) parsed into three components of sensing; the first six elements are: earth, water, fire, air, space, and consciousness; the next six are: pleasure, pain, joy, distress, equanimity, and ignorance; the next six are: sensual desire, renunciation, ill will, non-ill will, cruelty, and non-cruelty; the three are sense-sphere, form-sphere, and formless-sphere elements; the two are: conditioned and unconditioned. Another sutta that refuses to settle on a single list of phenomena (in its case, feeling) so we get lists of two, three, five, sixteen, thirty-six, and one hundred and eight (“The Many Types of Feeling Sutta,” Bahuvedanīya Sutta at M i.396–400; cf. S iv.223–28, 231–32 see Heim 2021). 8 Dhs 1: [katame] kusalā dhammā akusalā dhammā abyākatā dhammā (Rhys Davids 1975, 1–2). 9 In the suttas, dhammas in this technical sense are phenomena observed in meditation, as in the Mahāsātipaṭṭhāna Sutta (D ii.290) and the Anupada Sutta (discussed in what follows). 10 As 38: nissattanijjīvatā (Tin and Rhys Davids 1976, 49). 11 As 38: ārogyaanavajjachekasukhavipākesu (Tin and Rhys Davids 1976, 48). 12 Dhs 8–9 (Rhys Davids 1975, 1–5). 13 Anupada Sutta, M iii.25–29 (Ñāṇamoli and Bodhi 1995, 899–902).

156

The Dhammasan˙ganī and Vibhan˙ga ˙ 14 As 135–136 (Tin and Rhys Davids 1976, 178–80). Gethin discusses the implications of listing the “same” phenomena as powers and faculties (1992a, 143). 15 Nyanaponika (1998, 88–89; 37–41), citing James Ward. These are points he emphasizes throughout ch. IV. “Modal” and “aspectual” are my terms, however. 16 Ye vā pana tasmiṃ samaye aññepi atth paṭiccasamuppannā arūpino dhammā: that is, literally, “and whatever other dependently-arisen formless dhammas occur on that occasion” (Dhs 8, 75, 87, et cetera). 17 As, for example, Dhs 75, 87. 18 Visuddhimagga XI.93, 96. Gethin makes the same point about rūpa: the analyses of rūpa “focus on the physical world as experienced by a sentient being” (1986, 36). See also Heim and Ram-Prasad (2018) for a systematic demonstration of what Buddhaghosa (and I think the Abhidhamma on which he is relying) means by nāma and rūpa. 19 The Vibhaṅga is translated by U. Thiṭṭila (1969), and his introduction provides a useful overview. 20 As 2–3 (Tin and Rhys Davids 1976, 4–5). 21 It is standard to cite the Compendium’s system, as for example, Williams and Tribe (2000, 90–91), Gethin (1998, 210–15), Lamotte (1988, 594–97), and Nyanatiloka (2008, Appendix II); following Nyanatiloka, the system is given in charts in appendices in Ñāṇamoli’s translation of the Visuddhimagga (Ñāṇamoli 1999), even though the Visuddhimagga itself nowhere articulates it as such. 22 Compare Dhs 8 and Dhs 75 to the Compendium’s chapter II. The latter’s main commentary (the Abhidhammatthavibhāvinī) occasionally notes a discrepancy between the Enumeration and the Compendium, as for example, the way that the “heart base” (hadayavatthu) is not mentioned in the former but warrants listing among rūpa dhammas by the Compendium since it is “in accordance with the scriptural tradition and reasonable” (Wijeratne and Gethin 2007, 221). 23 It is surprising the extent to which scholars read these terms back into the canonical texts, however (as, for example, Nyanatiloka 2008, 16). 24 Saṅgaha VI.65; Wijeratne and Gethin 2007, 255. My translation. 25 Saṅgaha VIII.42; Wijeratne and Gethin 2007, 322. 26 For example, Abhidhammatthavibhāvinīṭīkā §2 (see Wijeratne and Gethin 2007, 7) and Ledi Sayadaw’s Paramatthadīpanī 332. 27 Karunadasa (1996, 15 and 42n37), citing the Abhidhamma Mūlaṭīkā (121): dhammo ti sabhāvo. 28 Saṅgaha III.1; Wijeratne and Gethin (2007, 88). 29 U. Thiṭṭila (1969, xxv). This kind of reading comes to be read back into the canonical texts, as in this instance. 30 Karunadasa (1996, 2). Noa Ronkin, who frames the Pāli Abhidhamma in strongly metaphysical and ontological terms, also acknowledges, drawing on Gethin, that the canonical tradition did not posit irreducible elements and is more open-ended than is usually recognized (2005, 109). 31 Karunadasa (1996, 19 citing the Abhidhammārthasaṁgraha-sannaya); the use of paramatthasabhāva is also found in the Vibhaṅga-Mūlaṭīkā and the Mahāṭīkā on the Visuddhimagga. 32 Kvu 1: puggalo upalabbhati saccikaṭṭhaparamatthenāti. 33 Many scholars assume that this distinction about language is really about reality: Karunadasa says that the passage denies “the reality of a person” (1996, 11), and that this concerns “two levels of reality” (1996, 12). Ledi Sayadaw asserts that from the standpoint of ultimate reality, the notion of a personal entity is “just an erroneous view” (1913–1914, 129). 34 Papañcasūdanī i.137 and Manorathapūraṇi i.94. See Heim (2018, 85–94) for translations and discussions of the relevant passages from Buddhaghosa. 35 As 39: Attano pana sabhāvaṃ dhārentīti dhammā. Dhāriyanti vā paccayehi, dhārīyanti vā yathāsabhāvatoti dhammā. It is not easy to translate dhārenti and dhārīyanti – to wear, have, possess, bear, bring. 36 As 155: Dhammāva ete dhammamattā asārā apariṇāyakāti imissā suññatāya dīpanatthaṃ vuttā. Tasmā evamettha attho veditabbo  – yasmiṃ samaye kāmāvacaraṃ paṭhamaṃ mahākusalacittaṃ uppajjati, tasmiṃ samaye cittaṅgavasena uppannā atirekapaṇṇāsadhammā sabhāvaṭṭhena dhammā eva honti; this is on Dhs 121. 37 Paṭisambhidāmagga II.178 (sabhāvena suññaṃ). For a translation of its chapter on emptiness see Ñāṇamoli (1982, 356–61). The Paṭisambhidāmagga is an Abhidhamma-like text that was placed in the Suttanta Piṭaka; Buddhaghosa relies on it heavily. For another example of Buddhaghosa noting the emptiness of dhammas, note how he sees the twelve parts of dependent origination as empty in Vism XVII.273 and 283.

157

Maria Heim 38 Vism XVII.25 and Sammohavinodaṇī 130: atthaṃ saṅgāhentena tamevatthaṃ punarāvattetvā aparehipi pariyāyantarehi niddisantena. 39 Gethin (1992b, 165). It should be noted that Gethin from time to time slides into a metaphysical treatment of dhammas and the whole system; as for example, dhammas are physical and mental events that are “the way things are” and very much like “atoms” (1998, 209).

Bibliography Anālayo. 2014. The Dawn of Abhidharma. Hamburg: Hamburg University Press. Bodhi, Bhikkhu, ed. 2007. A Comprehensive Manual of Abhidhamma: The Abhidhammattha Saṅgaha of Ācariya Anuruddha. Kandy: Buddhist Publication Society. Bodhi, Bhikkhu, trans. 2012. The Numerical Discourses of the Buddha: A Translation of the Aṅguttara Nikāya. Boston: Wisdom Publications. Brahmavamso, Ajahn. 1995. Chaṭṭhasaṅgāyana Saṅgāyana Tipiṭaka. Igatpuri: Vipassana Research Insitute. Braun, Erik. 2015. “The Great War of the Commentaries: The Abhidhamma and Social Change in Colonial Burma.” History of Religions 55 (1): 1–40. Bronkhorst, Johannes. 1985. “Dharma and Abhidharma.” Bulletin of the School of Oriental and African Studies 48: 305–20. Buswell, Robert E., and Donald S. Lopez, eds. 2014. The Princeton Dictionary of Buddhism. Princeton: Princeton University Press. Cousins, Lance S. 1981. “The Paṭṭhāna and the Development of the Theravādin Abhidhamma.” Journal of the Pāli Text Society 9: 22–46. ———. 2014. “The Case of the Abhidhamma Commentary.” Journal of the International Association of Buddhist Studies 36–37: 389–422. ———. 2015. “Abhidhamma Studies III: Origins of the Canonical Abhidha(r)mma Literature.” Journal of the Oxford Centre for Buddhist Studies 8: 96–145. Gethin, Rupert. 1986. “The Five Khandhas: Their Treatment in the Nikāyas and Early Abhidhamma.” Journal of Indian Philosophy 14 (1): 35–53. ———. 1992a. The Buddhist Path to Awakening: A Study of the Bodhi-Pakkhiyā Dhammā. Leiden: E. J. Brill. ———. 1992b. “The Mātikās: Memorization, Mindfulness and the List.” In In the Mirror of Memory: Reflections on Mindfulness and Remembrance in Indian and Tibetan Buddhism, edited by Janet Gyatso, 149–72. Albany: State University of New York. ———. 1998. The Foundations of Buddhism. New York: Oxford University Press. ———. 2009. “He Who Sees Dhamma Sees Dhammas: Dhamma in Early Buddhism.” In Dharma: Studies in its Semantic, Cultural, and Religious History, edited by Patrick Olivelle, 91–120. New Delhi: Motilal Banarsidass. Heim, Maria. 2018. Voice of the Buddha: Buddhaghosa on the Immeasurable Words. New York: Oxford University Press. Heim, Maria, and Chakravarthi, Ram-Prasad. 2018. “In a Double Way: Nāma-Rūpa in Buddhaghosa’s Phenomenology.” Philosophy East and West 68 (4): 1085–115. Heim, Maria. 2021. “Some Analyses of Feeling.” In The Bloomsbury Research Handbook of Emotions in Classical Indian Philosophy, edited by Maria Heim, Ram-Prasad Chakravarthi, and Roy Tzohar, 87–105. London: Bloomsbury. Karunadasa, Y. 1996. The Dhamma Theory: Philosophical Cornerstone of the Abhidhamma. Kandy: Buddhist Publication Society. Lamotte, Étienne. 1988. History of Indian Buddhism From the Origins to the Śaka Era. Translated by Sarah Webb-Boin. Louvain: Institut Orientaliste. Ñāṇamoli, trans. 1982. The Path of Discrimination (Paṭisambhidāmagga). London: The Pāli Text Society. ———. 1999. The Path of Purification: Visuddhimagga By Bhadantācariya Buddhaghosa. Kandy, Sri Lanka: Buddhist Publication Society. Ñāṇamoli, and Bhikkhu Bodhi, trans. 1995. The Middle Length Discourses of the Buddha: A Translation of the Majjhima Nikāya. Somerville, MA: Wisdom Publications. Nyanaponika Thera. 1998. Abhidhamma Studies: Buddhist Explorations of Consciousness and Time. Edited by Bhikkhu Bodhi. Boston: Wisdom.

158

The Dhammasan˙ganī and Vibhan˙ga ˙ Nyanatiloka. 2008. Guide Through the Abhidhamma Piṭaka: A Synopsis of the Philosophical Collection Belonging to the Buddhist Pāli Canon. Kandy: Buddhist Publication Society. Rhys Davids, C. A. F., trans. 1975. A Buddhist Manual of Psychological Ethics (Dhammasaṅgaṇī). London: Pāli Text Society. Ronkin, Noa. 2005. Early Buddhist Metaphysics: The Making of a Philosophical Tradition. New York: RoutledgeCurzon. Sayadaw, Ledi. 1913–1914. “Some Points in Buddhist Doctrine.” Journal of the Pāli Text Society 7: 115–69. Thiṭṭila, U., trans. 1969. The Book of Analysis (Vibhaṅga). Bristol: Pāli Text Society. Tin, Pe Maung, and C. A. F. Rhys Davids. 1976. The Expositor (Atthasālinī). London: Pāli Text Society. Westerhoff, Jan. 2013. “Metaphysical Issues in Indian Buddhist Thought.” In A Companion to Buddhist Philosophy, edited by Steven Emmanuel. Oxford: Wiley-Blackwell. Wijeratne, R. P., and Rupert Gethin, trans. 2007. Summary of the Topics of Abhidhamma (Abhidhammatthasaṅgaha) by Anuruddha and Exposition of the Topics of Abhidhamma (Abhidhammatthavibhāvinī) by Sumaṅgala. Oxford: Pāli Text Society. Williams, Paul, and Anthony Tribe. 2000. Buddhist Thought: A Complete Introduction to the Indian Tradition. London: Routledge.

159

10 MOGGALIPUTTA TISSA’S POINTS OF DISCUSSION (KATHĀVATTHU) Reasoning and Debate in Early Buddhist Thought1 Rupert Gethin Introduction The Points of Discussion (Kathāvatthu) is a collection of early Buddhist debates about disputed points of Buddhist thought. It thus provides insight into the philosophical issues that preoccupied early Buddhists and more especially their methods of argument and reasoning. It is counted the fifth of seven works that constitute the Abhidhamma-piṭaka, the third division of the Pali canon of the Theravāda school of Buddhism. As such, it is awarded by followers of the school the status of ‘the word of the Buddha’ (buddhavacana). Yet like two other works of the Pali canon, it is traditionally attributed to a named individual. The Path of Discrimination (Paṭisambhidāmagga) and Explanation (Niddesa) of the Khuddaka-nikāya are attributed to one of the Buddha’s immediate disciples, Śāriputra. The Points of Discussion, however, is associated with an individual who is acknowledged to have lived several generations after the death of the Buddha (ca. 400 BCE), during the reign of the emperor Aśoka in the middle of the third century BCE, the elder Moggaliputta Tissa. The Theravāda exegetical tradition acknowledges that some might object to regarding a work composed a century or more after his death as ‘the word of the Buddha.’ However, it justifies doing so by using a strategy derived from the discourses of the Sutta-piṭaka. In a number of discourses, we find the Buddha giving a brief outline that is subsequently explained in full by a disciple; questioned about this, the Buddha confirms that he would have given a full explanation in exactly the same terms as the disciple has done. In a similar vein, Theravāda tradition claims that the Buddha, having foreseen Moggaliputta’s situation, established the basic headings of the Kathāvatthu, and on the basis of these Moggaliputta later expounded the full text not by his own knowledge but according to a method and scheme received from the Buddha (As 4–5; Kv-a 1–2). In this matter, it seems we must side with the ancient skeptics: that the contents and method of the Points of Discussion derive directly from the Buddha is hardly possible. The Points of Discussion assumes the development of attempts to set out the Buddha’s teaching systematically and in detail on the basis of an established collection of discourses of the Buddha; it assumes the development of different schools of Buddhist systematic thought 160

DOI: 10.4324/9781351030908-17

Moggaliputta Tissa’s Points of Discussion (Kathāvatthu)

(Abhidharma) with their own positions on a variety of issues. In short, the Points of Discussion reflects the Buddhist intellectual world of the third–first centuries BCE (Cousins 1991, 34). The Points of Discussion itself provides no explicit indication of its authorship or of the circumstances of its composition. The text is a collection of discussions (kathā) of various points (vatthu) of Buddhist doctrine. Each discussion takes the form of a dialogue between representatives who take opposing positions on the points in question. The oldest extant commentary (Kv-a, composed probably at the beginning of the fifth century by an unnamed monk at the invitation of Buddhaghosa) tells us that in each case, these two opposing positions are those of the advocate (sakavādin), who presents the position of the author’s own school, the Theravāda school of Lanka, and an opponent ( paravādin), who presents the position of another school. The text provides no details of which schools of Buddhism these opponents might represent. According to the details provided by the commentary, the opponents are representatives of various Buddhist schools. This seems to be corroborated by what we know of the positions on points of doctrine taken by different Buddhist schools from other sources (see Bareau 2013). The discussions are arranged in twenty-three chapters (vagga) each comprising between five and twelve discussions. The chapters are simply numbered sequentially and not given titles, and while discussions of related points are sometimes grouped, there appears to be no overall topical method to the arrangement. In all, the Points of Discussion consists of between 216 and 226 topics of discussion.2 The text is difficult to date as a whole since its structure clearly lends itself to the addition of further points of discussion as they might arise. The lack of an overall method of ordering suggests that this is, in fact, how the text developed. Tradition tells us that Moggaliputta Tissa was born into a brahman family and educated in Vedic tradition. He was induced to join the Buddhist order by a monk supported by his family. He subsequently mastered the Buddhist textual tradition, achieved awakening, and became a monk of great renown. As the story has it, the ascendancy of the Buddhist monastic order during the reign of Aśoka attracted spurious monks. Ascetics belonging to non-Buddhist schools infiltrated the order and continued with their own practices while neglecting the rules of the Buddhist order. The king himself became involved and sought the help of a respected monk, namely Moggaliputta Tissa. With his help, the spurious monks were expelled and the order declared pure and unified. Moggaliputta Tissa then convened a Buddhist council at which he recited the Points of Discussion, defeating any opposing views (Jayawickrama 1962, 33–55). This account of the origin of the text seems ill fitted to its nature: a text that is concerned with the nuances and finer points of Buddhist doctrine seems largely irrelevant to these circumstances, and it is likely that different issues – keeping monastic discipline, guarding against schismatic tendencies, disagreement on points of doctrine, establishing an authoritative Buddhist textual corpus – have been conflated.

The method of argument The Points of Discussion is a difficult text. There is no critical edition; the only available translation into a modern European language is a paraphrase (Aung and Rhys Davids 1915, li–lii). A full study remains a scholarly desideratum; Warder’s 1963 article remains the most reliable guide to the overall method of argument. In order to begin to navigate the Points of Discussion, we must rely in good measure on the oldest commentary. While there exists an English translation of this commentary (Law 1940), it too must be regarded as provisional in nature. The literary form of the Points of Discussion is that of a series of dialogues between an advocate of the text’s own positions and an opponent. The text begins abruptly with the first dialogue and continues without ever providing any indication of who is speaking which lines. Although who (Theravādin or opponent) must be saying what is – for the most part – apparent, 161

Rupert Gethin

it is not always so (Aung and Rhys Davids 1915, lii–liii), which makes following an argument’s twists and turns sometimes difficult. Each dialogue begins with the Theravādin advocate putting the disputed point to the opponent in the form of a proposition and asking if this is the position the opponent holds. The opponent assents to the proposition. The advocate then puts further propositions to the opponent, which – in nearly all cases – the opponent denies. The Points of Discussion’s advocate then suggests that the first proposition to which the opponent has assented implies these further propositions, so in denying these, the opponent should acknowledge that his initial position must be false. As we shall see, the method in part adumbrates the prasaṅga method of reductio developed by Nāgārjuna (second century CE) and his Mādhyamika followers. In principle, a complete dialogue for a single disputed point follows the same pattern (see Table 10.1). First, there is what the commentary calls ‘the way of setting out the thesis’ (vādayutti),3 a set of eight exchanges (A) termed ‘refutations’ (niggaha) in which the opponent’s position is shown to be incoherent. This is followed by a series of further discussions (B) addressing various propositions supplementary to the main proposition and aimed at further demonstrating the unwanted implications of the opponent’s position; often, these supplementary discussions end with both sides citing canonical sūtra sources (suttāharaṇa) in support of their respective positions. In the case of the first disputed point, the series of supplementary discussions takes up some sixty pages (Kv 11–69, §§ 17–245); as the text progresses, the number of supplementary points tends to become fewer. The initial set of eight refutations (A) follows a formalized pattern that provides a template for demonstrating the contradiction involved in the opponent’s assenting to an initial proposition while denying a second proposition. This template is intended to be adapted and applied as appropriate to the supplementary propositions (B) and also to all other disputes in the Points of Discussion. Each of the eight refutations consists of five sections: (i) an initial demonstration in five steps of the contradiction involved in affirming a thesis; (ii) a rejoinder in four steps; (iii) a refutation in four steps based on the rejoinder; (iv) the application of the refutation in four steps; and (v) a conclusion in four steps asserting the success of the rejoinder.4 In the first section of the first of the eight refutations (initial demonstration), the Theravādin advocate sets out the refutation of the opponent’s position; the opponent then responds in the second section by fallaciously turning the Theravādin’s method of reduction back on him and apparently winning the first round. In the first section of the second refutation, the opponent repeats his counter to the Theravādin’s position, who then responds in the second section by repeating his refutation of the opponent’s position and so winning the debate. As Warder points out, these first two refutations are considered basic; the subsequent six refutations can be applied as appropriate. In effect, they form three pairs that parallel the initial pair. Thus, in the third and sixth refutations, the question is whether the opponent’s position applies in all circumstances; in the fourth and seventh, whether it applies always; in the fifth and eighth, whether it applies to everything. But even in the first two refutations, it is apparent that much – from the perspective of the logic of the argument – is redundant. This is in practice acknowledged by the text itself insofar as the eight refutations are set out in full only in the case of the first disputed point (Kv 2–11, §§ 1–16); in all subsequent debates, only the initial part of the first exchange is given, with the rest indicated by ‘and so on’ ( peyyāla). This is possible since the first section (i) – the straightforward direct statement of the argument in five steps – of the first refutation contains the essential reasoning that informs the argument in the next seven refutations. The remaining sections and refutations can in principle be constructed for each disputed point on its basis. 162

Moggaliputta Tissa’s Points of Discussion (Kathāvatthu) Table 10.1  The structure of the points of discussion. First point of discussion A. The way of setting out the thesis (vāda-yutti): 8 ‘refutations’ (niggaha) 1. Refutation in 5 sections i. Direct (anuloma): exchange initiated by Theravādin advocate (5 steps) ii. Rejoinder ( paṭikamma): exchange initiated by opponent (4 steps) iii. Refutation (niggaha): statement by opponent (4 steps) iv. Application (upanayana): statement by opponent (4 steps) v. Conclusion (niggamana): statement by opponent (4 steps) 2. Refutation in 5 sections i. Counter ( paccanīka): exchange initiated by opponent (5 steps) = 1.ii ii. Rejoinder: exchange led by Theravādin (4 steps) = 1.i iii. Refutation: statement by Theravādin (4 steps) iv. Application: statement by Theravādin (4 steps) v. Conclusion: statement by Theravādin (4 steps) [1–2 = 10 sections; 42 steps] 3. Refutation in 5 sections: everywhere (sabbattha) – paired with 6 4. Refutation in 5 sections: always (sabbadā) – paired with 7 5. Refutation in 5 sections: in everything (sabbesu) – paired with 8 6. Refutation in 5 sections: everywhere – paired with 3 7. Refutation in 5 sections: always – paired with 4 8. Refutation in 5 sections: in everything – paired with 5 B. Supplementary applications (saṃsandana) 1. Additional discussions using the template (abbreviated) set out in A 2. Citing authoritative sūtra sources (suttāharaṇa) Remaining 216/226 points of discussion A. The way of setting out the thesis: 8 ‘refutations’ 1. Refutation in 5 sections (only section i. stated) i. Direct exchange initiated by Theravādin (5 steps) ‘And so on. . .’ ( peyyāla) B. Supplementary applications 1. Additional discussions 2. Citing authoritative sūtra sources

Given the significance of this first section, it is worth examining its steps in detail. The commentary points out that the Theravādin demonstrates his refutation in two ways: direct (anuloma) and reverse ( paṭiloma). The first disputed point concerns the existence of the person: the opponent is a Buddhist who asserts the reality, in some sense, of the person, a pudgalavādin.5 Its first section is as follows (with names of the speakers, the distinction between the direct and reverse ways of refutation, and the five steps as indicated by the commentary added): Theravādin: The person exists in a real and ultimate sense? Opponent: Correct. Theravādin: The person exists in a real and ultimate sense in the same way as something which is real and ultimate? Opponent: No, this cannot be said. Theravādin: Understand this refutation: 163

Rupert Gethin





[Direct (anuloma)] – If the person exists in a real and ultimate sense, (1) then clearly it should be said that the person exists in a real and ultimate sense in the same way as something which is real and ultimate. – What you say is false, namely that (2) it can be said that the person exists in a real and ultimate sense, but it cannot be said that the person exists in a real and ultimate sense in the same way as something which is real and ultimate. [Reverse ( paṭiloma)] – But (3) if it cannot be said that the person exists in a real and ultimate sense in the same way as something which is real and ultimate, (4) then clearly it cannot be said that the person exists in a real and ultimate sense. – What you say is false, namely, (5) that it can be said that the person exists in a real and ultimate sense, but it cannot be said that the person exists in a real and ultimate sense in the same way as something which is real and ultimate.

To use the terminology of the commentary, it is apparent that the first proposition (the opponent’s position) takes the form of the protasis or antecedent (ṭhapanā) of a conditional statement, and the second proposition its apodosis or consequent ( pāpanā). The logical steps involved in this initial argument have been analyzed by Aung (Aung and Rhys Davids 1915, xlviii–l), Schayer (1933, 91–92; 2001, 28), Bocheński (1961, 421–23), Warder (1963, 64, 67), Watanabe (1983, 157–59), and Matilal (1998, 35–37), though without complete agreement. In the first place, there is disagreement as to whether this argument is better expressed by way of the logic of terms or the logic of propositions. Schayer (1933, 90–91; 2001, 28) is clear that we should use the latter; Bocheński (1961, 422–23) disagrees; Warder (1963, 64, 67), without reference to the issue – or to Schayer and Bocheński – uses both; Matilal (1998, 37) leaves the issue open. Expressing the debate by way of the logic of terms gives the following: Theravādin: Opponent: Theravādin: Opponent: Theravādin:

Is A B? Yes. Is A C? No. [Direct (anuloma)] – If A is B, then A is C. – Therefore not: (A is B) and not (A is C). [Reverse ( paṭiloma)] – If A is not C, then A is not B. – Therefore not: (A is B) and not (A is C).

Expressing it by way of the logic of propositions gives the following: Theravādin: Opponent: Theravādin: Opponent: Theravādin:

p? Yes. [⊦ p] q? No. [∼q]

164

Moggaliputta Tissa’s Points of Discussion (Kathāvatthu)



[Direct (anuloma)] – If p then q. [p ⊃ q] – That p and not q, is false. [∼ ( p. ∼q)] [Reverse ( paṭiloma)] – If not q, then not p. [∼q ⊃ ∼p] – That p and not q, is false. [∼ ( p. ∼q)]

As Warder (1963, 63–64) points out, not all disputed points are of the form: Is A B? Is A C? He distinguishes seven basic forms on the basis of an analysis of the first twenty-one debates. It would seem that at least some disputes (e.g. II.11, ‘there are two cessations’) could only be expressed as propositions. There is also some disagreement as to whether the argument is better reduced to the logical form of modus tollens – if it is true that p, then it is true that q; it is not true that q, therefore it is not true that p (Schayer 1933, 91, 2001, 28) – or modus ponens – if it is not true that q, then it is not true that p; it is not true that q, therefore it is not true that p (Matilal 1998, 37). But a preoccupation with the logical form of the argument may in part miss the point. The Points of Discussion is concerned with more than mere syllogism. I have set out in full just the first section of the first refutation, which, as noted, contains the essentials of the argument. Warder (1963, 64–65, 67–68) and Watanabe (1983, 163–67) analyze and set out the logical steps involved in each of the five sections of the first and second refutations; Watanabe (1983, 167–74) continues by analyzing the initial sections of the third–eighth refutations. But even if we take just the first and second refutations as constituting a full debate, we have ten sections with a total of forty-two steps. As Warder (1963, 66) notes, this seems ‘unnecessarily long-winded’ and clearly can never have been intended as a neat syllogism. The reasons for the redundancy of much of the argument from a logical point of view remain unclear. Repetitions that appear redundant are a feature of early Buddhist literature and, besides their mnemonic purpose in a literature that was composed and transmitted orally, they seem also to have served a meditative purpose. Certainly to construct and recite, on the basis of the template of the first point of discussion, the sections of the first two refutations for subsequent points would demand sustained attention and considerable mental agility in order to follow the inversion of statements and swapping of terms. Moreover, the proliferation of layers of repetition appears to be a deliberate literary device employed in the canonical Abhidhamma works, which, as the commentaries point out, are supposed to be infinite in extent when expanded in full and thus beyond our grasp (Gethin 2020). There is also perhaps a rhetorical purpose in the redundant arguments: in each debate, the opponent is shown to resort to fallacious sophistry (chala) and so is thoroughly defeated. And yet this does not mean that the opponent is not allowed some good arguments, for a basic purpose of the Points of Discussion is also to probe the implications and coherence of the teaching of the Buddha as set out in the received discourses of the Buddha, which both the Theravādin and opponents accept as authoritative. Nonetheless, despite the logical redundancy of the series of refutations, this formalized system of debate and stereotyped analysis appear to mark a significant step in the development of Indian logic.6 Schayer thus suggests that the basic form of argument reveals that the author of the Points of Discussion understood the ‘definition of implication’ (namely, the equivalence of ‘if p then q’ and ‘that p and not q, is false’) and the law of transposition (the equivalence of ‘if p then q’ and ‘if not q, then not p’) (1933, 92, 2001, 28).

The application of the method: two examples Having considered the basic method of argument of the Points of Discussion, I shall briefly consider its application to two different disputed points: one purely conceptual and abstract 165

Rupert Gethin

(that past, future, and present realities exist), the other with more immediate and practical implications since it concerns something that might be adduced as evidence of the attainment of awakening or not (that arhats may have seminal emissions).

Past, present, and future realities exist The sixth disputed point of the first chapter is concerned with the position that fundamental and irreducible realities (dharmas) – the mental and physical ‘events’ that make up the world – of the present, past, and future exist (Kv 115–143). This is, of course, the position held by the Sarvāstivāda school. This point of discussion is of particular interest in the present context since we possess an account of the same dispute in a text belonging to a different school that must be more or less contemporary with the Points of Discussion. The Consciousness Group (Vijñānakāya) – a text of the Sarvāstivāda Abhidharma-piṭaka that survives only in Chinese translation (T. 1539) – opens with an account of a debate between an advocate of the Sarvāstivāda position and an opponent named Maudgalyāyana (Muganlian). It seems likely, though it cannot be proved, that this opponent should be identified with Moggaliputta Tissa (La Vallée Poussin 1925, 344). In the Points of Discussion, the debate begins with a simple statement of the Sarvāstivāda position: ‘Everything exists?’ The opponent assents. In trying to refute such a view, the advocate of Theravāda interrogates the opponent by placing a series of propositions and questions, which should be understood as debated according to the template set out previously. This method is such that it does not allow for an initial account of why some Buddhist thinkers might want to hold such a position, though at the close of the debate, the opponent is allowed to quote some sūtra sources that give some grounds for the position. The assumption must be that those who come across this discussion already know what it is about. I will return to the grounds for such a position in what follows after first considering the attempt by the Theravādin advocate to demonstrate its incoherence. The initial sections seem to aim at confirming that the opponent does indeed wish to maintain the counterintuitive position that things – the five bundles (skandha) of material form, feeling, perception, volitional forces, and consciousness  – that have ceased (the definition for what is past) or have not yet arisen (the definition for what is future) nonetheless exist (Kv 115–119, §§ 1–4). The argument is that this is self-evidently incoherent: ‘ceased’ means ceased to exist, and ‘not yet arisen’ means not yet existing. The Theravādin advocate’s next move is to argue that the opponent’s position must surely compromise the doctrine that all irreducible realities, all dharmas (apart from nirvāṇa), are impermanent, which is a core principle of Buddhist doctrine. Indeed, according to the opponent’s position, material realities must retain their intrinsic material quality in the future, in the present, and in the past, given that they are said to ‘exist in all times’ (the meaning of the school’s name). Further, there seems to be an internal incoherence since, while they lose the qualities of being future and present, dharmas never lose their quality of being past (Kv 120–125, §§ 7–20). Following this, the Theravādin suggests that the opponent’s position has absurd consequences: it leads to a position where we must say that the eye that exists in the past sees a visible form that is past with a cognition that is past; similarly, a future eye must see a future visible form and a present eye a present visible form (Kv 126–129, §§ 23–28). The argument then shifts to the situation of the noble ones, such as arhats, who, by definition, have eradicated greed, hatred, and delusion: if their past greed, hatred, and delusion exist, then why do they not affect them? For the opponent accepts that an ordinary person’s past greed, hatred, and delusion still affect them, since he accepts them as being existent in all times (Kv 131–136, §§ 35–46). 166

Moggaliputta Tissa’s Points of Discussion (Kathāvatthu)

The Points of Discussion continues by spelling out further absurd consequences of the opponent’s position: that if past hands, and so on, exist, then all the actions performed by those hands must exist (Kv 136–137, §§ 47–49); that there must exist three instances of everything, one in the future, one in the present, one in the past (Kv 137–138, §§50–54). Then once more it returns to what seems to be regarded as the fundamental point that when we ordinarily say of things that they are ‘past’ or they are ‘future,’ we must mean that they do not exist. Yet asked if what is past exists, the opponent replies yes; asked if what exists is past, the opponent replies that what exists may be past or it may not be past, since he holds the position that things that are present and future also exist. But this, the Theravādin suggests, is to say that what is past is not past (since it exists), and what is not past is past (since both are alike in existing). This applies also to the future and the present (Kv 138–140, §§ 55–58). The debate ends with the Theravādin and opponent each citing sūtra sources that apparently lend support to their respective positions (Kv 140–143, §§ 59–64). It is possible to distinguish three lines of argument in the overall discussion. The first concerns linguistic usage (to say of things they are past or future can only mean they do not exist). The second concerns the implications of the disputed position for Buddhist doctrine (e.g., it compromises the fundamental doctrine that things are impermanent). The third attempts to demonstrate that the position has absurd consequences (there must be three of everything). Turning to The Consciousness Group, which presents Sarvāstivādin arguments from their own perspective, the basic method of argument is similar. The Sarvāstivādin repeatedly puts his opponent Maudgalyāyana’s position to him: the past and future do not exist; only the present and what is beyond conditions (nirvāṇa) exist. When Maudgalyāyana assents, the Sarvāstivādin confronts him with a statement to which he must say yes or no: initially, in this debate, whether it is possible to know that greed, hatred, and delusion are ethically unwholesome (akuśala). Maudgalyāyana agrees that it is. He must then answer whether the greed, hatred, and delusion that are the object of such a cognition are past, future, or present. If he says past or future, he is defeated, because they must nonetheless exist as the objects of the cognition. If he says present, he must admit that a person can have two cognitions at once: (i) an unwholesome cognition associated with greed and, at the same time, (ii) a wholesome cognition that takes the first cognition as its object and understands the greed to be ethically unwholesome. But, according to an accepted Abhidharma principle, two simultaneous cognitions are not possible. The Consciousness Group proceeds by adducing other circumstances that seem to demand that something past or future must exist in some sense (La Vallée Poussin 1925, 346–49; T. 1539, 531a27–c2). Bronkhorst (1993) has suggested that a comparison of this disputed point in the Points of Discussion and The Consciousness Group reveals that the author of the Points of Discussion was unaware of the Sarvāstivādin arguments for their position. I am not so sure. Bronkhorst claims (on the basis of Kv 127–29, §§ 1.6.23–28) that the position of the Points of Discussion is that only present cognitions of present objects with a present sense organ are possible; this shows that the author of the Points of Discussion was unaware of the Consciousness Group’s arguments based on the cognition of past objects. But I think Bronkhorst misreads the Points of Discussion’s attempts at reductio of the Sarvāstivādin position. First, it is quite clear that the Theravāda Abhidhamma did not deny that past material and mental dharmas might be the objects of present cognition (Dhs 187, §§ 1041–3; 241, §§ 1417–8); they simply denied that such objects should be said to exist. Whether their position is coherent is not the issue in the present context. The point is that neither the Points of Discussion nor The Consciousness Group shows explicit awareness of their opponent’s arguments for their positions. But this does not mean they were unaware of them. On the contrary, the authors of these treatises assume a certain familiarity with the background to these disputed points. The strategy of both 167

Rupert Gethin

is to bypass the opponent’s specific arguments and to identify and attack his position’s weak points directly by drawing out its implications with reference to shared axioms of Buddhist doctrine – and thereby demonstrate its incoherence. Thus, in the first disputed point of the Points of Discussion, the opponent does not attempt to set out the positive philosophical arguments for the position that the person exists, nor does the Theravādin set out the positive philosophical arguments for not-self. Rather, the debate is an intra-Buddhist affair that hinges on a single point: if you want to say the person exists, then you can only meaningfully make this claim if you mean that the person exists in precisely the same way that systematic Buddhist thought, in the form of the Abhidharma, states that other things (namely, dharmas) exist. The argument is only effective because the opponent, as a Buddhist, must avoid saying the self exists in these terms.

The arhat has impure emissions of semen The second chapter of the Points of Discussion opens with a discussion of the proposition that an arhat (one who is awakened and has eradicated all greed, hatred, and delusion) may have impure seminal emissions. The commentary suggests that such a position is maintained by those who have noticed that some whom they believe to be arhats have seminal emissions; according to the Theravādins, though, these are not genuine arhats  – they claim arhatship through self-delusion, conceit, or deceit (Kv-a 55). That an arhat might have seminal emissions is one of five points (all found in the Points of Discussion, though not grouped as five points) referred to in some later Buddhist sources as ‘the five points of Mahādeva,’ which all focus on what it means to be an arhat (Cousins 1991). The opponent assents to the proposition that an arhat has impure emissions of semen. The Theravādin suggests that this is incompatible with other propositions that the opponent also accepts, namely (i) that an arhat has no desire; and (ii) that it is ordinary men with desire who have seminal emissions. The opponent continues to maintain his initial thesis according to which the arhat, despite having no desire, has seminal emissions (Kv 163–164, §§ 1–2). The Theravādin then switches to a different line of questioning: if not through desire, how precisely do arhats have seminal emissions? The opponent claims there are mischievous gods belonging to the retinue of Māra (the god who, in Buddhist texts, continually tempts with the objects of the five senses, attacking the Buddha on the night of his awakening and offering him his daughters): these gods provide the arhat’s seminal emission. The Theravādin then asks if these gods bring their own seminal emissions, that of others, or the arhat’s, or if the emission exudes from the pores or the skin. The opponent denies all these possibilities, and the Theravāda suggests a contradiction (Kv 164–165, §§ 3–4). Again, the Theravāda switches to a different line of questioning: why do the gods provide seminal emissions? The opponent suggests that it is to induce doubt in the arhat. The Theravādin asks if arhats have doubt. If the opponent answers no, he must acknowledge the contradiction; if he answers yes, then the Theravādin asks whether the doubt is in the Buddha, the Dharma, the Sangha, and so on. The opponent denies that it is doubt of this sort. Again, there is an implied contradiction (Kv 165–166, §§ 5–6). The Theravādin now switches to physiological matters. First, where does the seminal emission come from? The opponent replies that it issues from food and drink. The Theravāda asks whether all those who eat and drink, including women and eunuchs, have seminal emissions. The opponent admits that this is not so. Further, if semen is produced by food and drink, there should be somewhere in the body where it is stored, as there is for feces and urine. The opponent again denies this (Kv 166–167, §§ 7–8). 168

Moggaliputta Tissa’s Points of Discussion (Kathāvatthu)

The Theravādin now returns to matters related to the initial line of questioning. If the arhat has seminal emissions, is he interested in such things as sex, having a family, perfumes, and money, as ordinary men who have seminal emissions are? The opponent denies this. The Theravādin asks whether the arhat has abandoned greed, hatred, delusion, and other afflictions; whether he has developed the constituents of awakening and the path and understood the four truths. The opponent agrees he has. The Theravādin suggests this is not compatible with having seminal emissions (Kv 167–170, §§ 9–13). At this point, the opponent changes tack. When asked if an arhat has seminal emissions, instead of responding simply yes, he replies that an arhat ‘skilled in his own practice’ (sadhammakusala) does, but an arhat ‘skilled in superior practice’ ( paradhammakusala) does not. This terminology only occurs in the Points of Discussion. We can do little better than follow the commentarial explanation of the distinction: the former is an arhat who has reached awakening without developing full concentration (samādhi) based on the four absorptions ( jhāna) and four formless meditation attainments (samāpatti), while the latter is one who has developed all eight. The significant point is that the opponent has introduced a distinction between arhats that the Theravādin now argues is beside the point. He does so by revisiting previous lines of question: Are the two types of arhat alike in having abandoned greed, hatred, and delusion and other afflictions? Are they alike in having developed the constituents of awakening and the path? The opponent agrees they are. The Theravāda once again suggests the opponent’s position is incoherent. (Kv 170–172, §§ 14–20). In response to the opponent’s restating his position, the Theravādin now cites a canonical source (Vin I 295) that states explicitly that there is no occasion on which an arhat’s semen might be released (Kv 172, § 21). The opponent acknowledges the source; the Theravādin suggests this is a final contradiction. But this is not quite the end of the discussion. The opponent takes the initiative and asks if the Theravādin denies that an arhat can have seminal emissions. The Theravādin denies this. The opponent returns to the question of providing things for an arhat: can people provide robes and the other monastic requisites for arhats? The Theravādin agrees they can. The opponent now asks whether the Theravādin agrees that, since others can provide arhats with robes, there is provision of things by others. The Theravādin agrees (Kv 172, §§ 22–23). The discussion thus ends with the opponent claiming that since provision of things by others is admitted, the Theravādin should not deny that an arhat can have seminal emissions. The point is presumably that if gods can provide the material requisites for arhats, those of Māra’s ilk might theoretically provide unwanted things, such as semen (Cousins 1991, 42). The overall argument in the case of this disputed point revolves around three principal issues: (i) the connection between seminal emissions and sexual desire; (ii) the physiology of the production and storage of semen; and (iii) whether a relevant distinction can be made between types of arhat.

Conclusions The Points of Discussion exemplifies a particular style of debate and method of argumentation focused on disputed points of Buddhist doctrine. The method of argumentation works by demonstrating the incompatibility of two propositions. The first is a particular position taken on the disputed point; the second may be an accepted point of Buddhist doctrine, something generally accepted to be true by ordinary human beings, or some unwanted implication of the first. The incompatibility is demonstrated using what is essentially a logical formula, although it is elaborated rhetorically. The opponent is in this way cornered, having either to deny essential points of Buddhist doctrine or admit absurd consequences in order to maintain the position. 169

Rupert Gethin

To what extent we might view the style of debate, the method of argumentation, and the compilation of a set of disputed points as the work of a particular author (call him Moggaliputta Tissa) is unclear. What is clear is that the set of disputed points is easily added to; the same is true of the supplementary points that are introduced in the course of debating the principal point. The parallels to some parts of the Points of Discussion that are found in The Consciousness Group suggest a situation in which Buddhist thinkers attempting to articulate the finer points of Buddhist doctrine were engaged in debate with each other. Such debates would have been a context for the exchange and development of methods of argumentation. That Moggali (of Theravāda sources)/Maudgalyāyana (of Sarvāstivāda sources) was a significant figure at some point in this process seems quite possible. The Middle Indian diction of certain of the repeated phrases of the Points of Discussion is distinctly eastern, which suggests that something of its core has relatively early origins in Magadha close to the ancient Mauryan capital of Pāṭaliputra. The question of the valid means for establishing knowledge ( pramāṇa), which became central to Buddhist philosophy from the fifth century CE, is not explicit in the Points of Discussion, but two valid means of knowledge are implicitly accepted in the text: the authority of scripture (sūtra) and reason ( yukti). These are put to use in a particular way. Denied the opportunity to articulate their positions at length, opponents generally must answer yes or no, and are thereby continually confronted with the unacceptable and absurd consequences of their positions. It seems not unreasonable to suggest that such a method of argumentation paves the way for someone like Nāgārjuna, who exploits such a method to its full: whatever position you adopt categorically, there will be unwanted consequences and incoherence.

Abbreviations As Dhs Dhs-mṭ Kv Kv-a T Vin

Atthasālinī. Müller, Edward, ed. 1897. London: Pali Text Society. Dhammasaṅgaṇi. Müller, Edward, ed. 1885. London: Pali Text Society. Dhammasaṅgaṇi-mūlaṭīkā. 1960. Rangoon: Buddhasāsana-samiti. Kathāvatthu. Taylor, Arnold Charles, ed. (1894, 1897) 1979. London: Pali Text Society. Kathāvatthuppakaraṇa-aṭṭhakathā. Jayawickrama, N. A., ed. 1979. London: Pali Text Society. Taishō Shinshū Daizōkyō. Junjirō Takakusu and Kaigyoku Watanabe, ed. 1924– 1932. Tokyo: Taishō Issaikyō Kankōkai Vinayapiṭaka. Vols. I–V Oldenberg, Hermann, ed. 1879–1883. London: Pali Text Society.

Notes 1 I am grateful to Pierre-Julien Harter for his careful reading and his comments and suggestions. 2 There is some variation in the editions on what counts as a separate point of discussion, especially in chapters nine and twenty-three. The PTS edition counts 216 points, the Burmese edition 226. 3 Kv-a 2.2–3; explained at Dhs-mṭ 15 as yuttī ti upāyo, vādassa yutti vādayutti, vādappavattanassa upāyo ti attho. 4 The number of steps in each section is stated in the text itself, but how to count those steps is taken from the commentary. As Warder (1963, 66) notes, it is not clear why the first section is said to consist of five steps and the second four, since they both have the same form. 5 The commentary suggests that those affirming the existence of the person include Buddhists and non-Buddhists: ‘Who affirms the person? Within the teaching [of the Buddha], the Vajjiputtakas and

170

Moggaliputta Tissa’s Points of Discussion (Kathāvatthu) Sammitiyas, and outside it, many followers of other systems of thought.’ (Kv-a 9.16–18: ke pana puggalavādino ti. sāsane Vajjiputtakā c’ eva Sammitiyā ca bahiddhā ca bahū aññatitthiyā.) 6 Bocheński (1961, 422) criticizes Randle’s’ assessment (1930, 13–14) of the Kathāvatthu as exhibiting disregard for logic.

Bibliography Aung, Shwe Zan, and Caroline A. F. Rhys Davids. 1915. Points of Controversy or Subjects of Discourse: Being a Translation of the Kathā-Vatthu. London: Pali Text Society. Bareau, André. (1955) 2013. The Buddhist Schools of the Small Vehicle. Honolulu: University of Hawai’i Press. Bocheński, Józef Maria. 1961. A History of Formal Logic. Notre Dame: University of Notre Dame Press. Bronkhorst, Johannes. 1993. “Kathāvatthu and Vijñānakāya.” In Premier Colloque Étienne Lamotte: Bruxelles et Liège, 24–27 Septembre 1989, 57–61. Louvain-la-Neuve: Université catholique de Louvain, Institut orientaliste. Cousins, Lance Selwyn. 1991. “The ‘Five Points’ and the Origins of the Buddhist Schools.” The Buddhist Forum 2: 27–60. Frauwallner, Erich. 1995. Studies in Abhidharma Literature and the Origins of Buddhist Philosophical Systems. Albany, NY: State University of New York Press. Gethin, Rupert. 2020. “Reading Repetitions in the Saṃyutta-nikāya and Early Abhidhamma: From the Mahā-vagga to the Dhammasaṅgaṇi.” In Research on the Saṃyukta-āgama, edited by Dhammadinnā, 109–169. Taipei: Dharma Drum Corporation. Jayawickrama, N. A. 1962. The Inception of Discipline and the Vinaya Nidāna: Being a Translation and Edition of the Bāhiranidāna of Buddhaghosa’s Samantapāsādikā, the Vinaya Commentary. London: Luzac. La Vallée Poussin, Louis de. 1925. “La controverse du temps et du pudgala dans le Vijñānakāya.” In Études Asiatiques publiées à l’occasion du vingt-cinquième anniversaire de l’École française d’Extrême-Orient, 343–76. Paris: G. Van Oest. Law, Bimala Churn. 1940. The Debates Commentary (Kathāvatthuppakaraṇa-aṭṭhakathā). London: Pali Text Society. Matilal, Bimal. 1998. The Character of Logic in India. Albany, NY: State University of New York Press. Randle, H. N. 1930. Indian Logic in the Early Schools. London: Humphrey Milford. Schayer, Stanislaw. 1933. “Studien zur indischen Logik. II. Altindische Antizipationen der Aussagenlogik.” Bulletin International de l’Académie Polonaise des Sciences et des Lettres, Classe de Philologie (1–6): 90–96. ———. 2001. “Studies on Indian Logic. II. Ancient Indian Anticipations of Sentential Logic.” Translated by Piotr Balcerowicz. In Materials of the International Seminar: Argument and Reason in Indian Logic, edited by Piotr Balcerowicz and Marek Mejor, 27–33. Warsaw: Instytut Orientalistyczny, Uniwersytet Warszawsk. www.balcerowicz.eu/indology/schayer2001.pdf. Warder, Anthony Kennedy. 1963. “The Earliest Indian Logic.” In Trudy Dvadtsat’ Piatogo Mezhdunarodnogo Kongressa Vostokovedov, Moskva 9–16 Avgusta 1960 [Proceedings of the Twenty-Fifth International Congress of Orientalists], vol. IV, edited by B. G. Gafurova, 56–68. Moscow: Izd-vo vostochnoi lit-ry. Watanabe, Fumiro. 1983. Philosophy and its Development in the Nikāyas and Abhidhamma. New Delhi: Motilal Banarsidass.

171

11 KĀTYĀYANĪPUTRA AND THE LARGE COMMENTARY (MAHĀVIBHĀS․Ā)

The Development of Abhidharma Literature and of a Sarvāstivāda Self-identity Bart Dessein As Buddhist doctrines spread across the Indian subcontinent, during and immediately preceding the Mauryan period, local monastic communities formed in the early stages of Buddhist history adapted to the local circumstances of the regions where they settled. Distinctive monastic schools (nikāya) began to develop in the fourth–second centuries BCE, each based upon the acceptance of particular interpretations of monastic discipline (Bechert 1973, 8). This period is often referred to as Nikāya Buddhism, and is characterized by textual reorganization: whereas the sacred texts of Buddhism were, at first, grouped together according to their style, this initial classification gradually gave way to the well-known “three baskets” (tripiṭaka) schema that groups texts according to content. Texts that describe the life of the historical Buddha and his sermons were collected in the Sūtra literature. Texts relating to monastic life were grouped in the Vinaya literature. Philosophical treatises were grouped in the Abhidharma collection. Buddhist historiography traces the origin of the Sūtra and the Vinaya literature to two direct disciples of the Śākyamuni Buddha: Ānanda and Upāli, respectively. Abhidharma texts postdate Sūtra and Vinaya texts. They typically cull lists of doctrinal items (dharma) and matrices (mātṛkās) from the Sūtra and Vinaya texts and bring them into a more or less organized philosophical system. It is very likely that the term “Abhidharma” refers to such lists of doctrinal items. One possible interpretation of the term “Abhidharma” is thus “concerning the components (dharma) of existence,” a term that also reflects the ontologies developed in Buddhist philosophy.1 The use of numerical lists in Sūtra, Vinaya, and Abhidharma literature points to the overall oral/aural Indian literary context in which Buddhism originated. These numerical lists hereby served as mnemotechnical devices in the oral recitation of texts. The monastic code of the Sarvāstivādins refers to both the oral origin of the Abhidharma and to the numerical lists as the backbone of these texts (in this case, a series of numerical lists comprising five elements each) with the following words that are attributed to Ānanda, a direct disciple of the Buddha (T.23.1435, 449a19 ff.): Ānanda was further asked: “Where did the Buddha start to preach the Abhidharma?” Ānanda answered: “Thus have I  heard: At one time, the Buddha was staying in 172

DOI: 10.4324/9781351030908-18

Kātyāyanīputra and Mahāvibhās∙ā

Śrāvastī. At that moment, the Buddha told all bhikṣus (monks): ‘When talking about the five . . ., five . . ., five . . ., five. . . .’ ” As the historical Buddha and his early disciples faded into the past, discussions between different monastic groups and subgroups about the correct interpretation of the often enigmatic numerical lists around which the Buddha had formulated his sermons increased. This contributed to the development of doctrinal sects within existing monastic communities. The Abhidharma literature is therefore also the literature of so-called sectarian Buddhism. Buddhist historiography refers to this phase of sectarian Buddhism as the eighteen schools of the Buddhism of the Disciple (Śrāvakayāna). A more common, albeit denigrating, name to refer to these eighteen schools is Lesser Vehicle Buddhism (Hīnayāna). This name contrasts early Buddhism with Mahāyāna or Greater Vehicle Buddhism that gradually developed from within the different schools and sects of early Buddhism. While the Abhidharma tradition may have had an oral origin, the texts as we possess them now are the product of an expanding written tradition. This is the logical outcome of both the increasing complexity of philosophical debate and the need for sectarian self-identification and concomitant recognition of specific philosophical texts as peculiar to one’s own group. It is to the self-identification of the Sarvāstivādins as a distinct group – a process that postdates the reign of the famous king Aśoka the Maurya (r. 268–232 BCE) and the committing to writing of philosophical texts  – that the historical figure of Kātyāyanīputra, a Buddhist scholar born to a Brahman family around the beginning of the common era, is connected.2 The commentator Paramārtha (499–569 CE) acknowledges this distinctive role of Kātyāyanīputra in his Compilation of the Investigation into the Secrets of the Subtle Meaning of the Three Treatises (San lun xuanyi jian you ji 三論玄義檢幽集) when stating that: In the beginning of the three hundred years after the demise of the Buddha, Kātyāyanīputra left this world, having invoked the creation of two schools. . . . Having come to the middle of the three hundred years after the demise of the Buddha, two schools were formed because of minor reasons: the first are the Sarvāstivādins who are also called the Hetuvādins, and the second are the Haimavatas who are also called the Sthaviravādins. (T.70.2300, 363a8–12) Tradition has it that it was on the occasion of a council that was convened by the Kuṣāṇa king Kaniṣka (c. 127–150 CE) that the Sarvāstivādins committed their canon to writing. Tarānātha’s History of Buddhism (Chos ’byung) more precisely informs us that at this council, the Sūtras and the Abhidharma that had not yet been committed to writing were written down, and those texts that had already been written down were purified. The reason given for committing these texts to writing was that ordinary people recited the scriptures incorrectly, making omissions and interpolations (Obermiller 1931–1932, vol. 2, 101). This statement implies that the oral medium of transmitting texts must – at least for some period – have coexisted with a written transmission. This also renders plausible the possibility that some texts were transmitted in both ways simultaneously. The importance of this for Kātyāyanīputra’s text is that it is very likely that two versions existed simultaneously: the Abhidharma Treatise in Eight Sections (*Abhidharmāṣṭagranthaśāstra or *Abhidharmāṣṭaskandhaśāstra) and the Source of Knowledge (Jñānaprasthāna). The self-identification of the Sarvāstivādins as a distinct philosophical group hinges on their standpoint that all entities (dharma) exist in the past, present, and future. That is to say, through the dynamics of causality, it is impossible for any entity to exist in the present without 173

Bart Dessein

the causes that shaped it but that, at present, already belong to the past. All present entities cast forth their future existence through the continuous operation of causality. In this sense, all present entities will actually also exist in the past, when they will have become the cause of that as of yet still future entity that will, at that moment, become a present entity. It is this doctrine that has given its name to this school: sarvam asti, meaning “all exists.” This highly ontological trait of Sarvāstivāda thinking is also representative of a more general development within Buddhism from a stage that focused on an individual’s liberation from their psychological suffering that was prominent in early Sūtra texts toward a doctrine that engaged with issues like the world in a temporal and a spatial sense and an individual’s position within this world.3 Within the Sarvāstivāda monastic community, there may have been agreement on the fact that “all exists.” However, ongoing philosophical discussions within this particular monastic group gradually led to the formation of two prominent doctrinal subgroups: the Sarvāstivādins of Bactria and Gandhāra, and the Vaibhāṣikas of Kaśmīra. Though each subgroup took Kātyāyanīputra’s text as a foundation, they engaged with different aspects of his works and formulated their ideas in different types of texts: while the more conservative group of Bactria and Gandhāra compiled pedagogical digests that organized and systematized the materials Kātyāyanīputra had selected in his Abhidharma Treatise in Eight Sections, the Sarvāstivāda subgroup that moved to Kaśmīra declared the Source of Knowledge to be the primary body (śarīra) of Sarvāstivāda thinking, identifying six Abhidharma texts that predated the self-identification of the Sarvāstivādins as a separate doctrinal group as the support or feet ( pāda).4 They compiled extensive and polemical commentaries on this Source of Knowledge version. One of these commentaries is the Large Commentary (Mahāvibhāṣā), a text that also contains Mahāyāna elements.

Kātyāyanῑputra and the Abhidharma Treatise in Eight Sections/Source of Knowledge Authorship and Extant Versions of the Text While the Sanskrit and the Tibetan traditions attribute the Abhidharma work that is entitled Source of Knowledge (Jñānaprasthāna) to Kātyāyanīputra,5 the Chinese tradition takes Kātyāyanῑputra as the creator (Chinese zao 造) of an Abhidharma work that is either called Apidamo fazhi lun 阿毘達磨發智論 (Source of Knowledge) or (Jñānaprasthāna) Apitan ba jiandu lun阿毘曇八犍度論 (Abhidharma Treatise in Eight Sections) (*Abhidharmāṣṭagranthaśāstra or *Abhidharmāṣṭaskandhaśāstra) (T.26.1543, 771b16; T.26.1544, 918a5). Kātyāyanῑputra is characterized as the compiler rather than the composer of the Source of Knowledge in the Large Commentary (on the Abhidharma) (*[Abhidharma]mahāvibhāṣā[śāstra], Apidamo da piposha lun 阿毘達磨大毘婆沙論), where we read that Kātyāyanῑputra “received the lectures” (shouchi yanshuo受持演說) the Buddha gave in a question-and-answer format and compiled them “in order to transmit them widely” (guang ling liubu廣令流布) (T.27.1545, 1a11–19). This suggests that the concept of authorship for Abhidharma works is not simply attribution of the text to an original figure who composed it; rather, Abhidharma texts used doctrinal materials that were transmitted over time. It also suggests that these materials were shaped into philosophical constructs that conformed to the developing interpretations of the words of the historical Buddha. Furthermore, it makes it clear that the argument of authority was an important issue for Abhidharma texts, as the compilers presented the Abhi‑ dharma texts to be the result of lectures received from the Buddha.6 The history of the transmission of the Abhidharma Treatise in Eight Sections reveals that there were several translations

174

Kātyāyanīputra and Mahāvibhās∙ā

in Chinese from presumably different “original” texts written in different languages (Middle Indic and Sanskrit), which makes it difficult to speak of one original text identifiable as the work of Kātyāyanīputra (Demiéville 1961, 472 n. 5; Takakusu 1904–1905, 84). Giyū Nishi (1934) analyzed the doctrinal positions associated with different doctrinal groups in the extant vibhāṣā compendia and concluded that the Source of Knowledge and the Abhidharma Treatise in Eight Sections represent different lineages of transmission within the larger Sarvāstivāda group. According to Ryōgen Fukuhara (1965, 218–19), the transliterations in the Abhidharma Treatise in Eight Sections indicate that the work may have been the Gandhāran version of the text, while the Source of Knowledge represents the recension of the work by the Kāśmῑri Vaibhāṣikas. It therefore is not unlikely that Kātyāyanῑputra compiled his Abhidharma Treatise in Eight Sections in central India, before the Sarvāstivāda school began to flourish in Kaśmῑra, and that the Source of Knowledge is a reworked version of that text. This corroborates the preceding hypothesis that the Abhidharma Treatise in Eight Sections, which Kātyāyanῑputra compiled by placing into the mouth of the historical Buddha doctrinal materials that had been transmitted over time, was later referred to as the Source of Knowledge. This hypothesis is relevant to the importance of each of these texts – the Abhidharma Treatise in Eight Sections and the Source of Knowledge – in the development of two new types of texts, each connected to a different doctrinal subgroup that developed within the larger Sarvāstivāda monastic community. The subgroup of the Sarvāstivādins of Bactria and Gandhāra, later called “Sautrāntika” (literally, “those following the Sūtras”), formulated their interpretations of Buddhist doctrine in summary pedagogical digests in which the essential elements of the doctrine were given in stanzas (kārikā), followed by an auto-commentary. This helps us understand why the titles of their texts often use the term hṛdaya (“heart,” “essence”; Chinese xin 心). The Sarvāstivāda subgroup of Kaśmīra, on the other hand, formulated their interpretations of the words of the Buddha in the form of extensive and polemical prose commentaries on the Source of Knowledge. It is these prose commentaries, called vibhāṣā (lit. extensive commentary) in Sanskrit, from which the Sarvāstivāda subgroup takes its name. The Bactrian and Gandhāran subgroup objected to the Kāśmīri Sarvāstivādin developments and claimed to adhere to the word of the Buddha more strictly; thus, they were called Sautrāntika to indicate that their doctrinal position was closer to the original words of the Buddha (as preserved in the Sūtras) than that of the Kāśmīri Sarvāstivādins.

Doctrinal Innovations of the Text: The Path Structure, the Material World, and the Liberating Power of Understanding Like the Abhidharma authors before him, Kātyāyanīputra selected numerical lists of doctrinal items (dharma) from Sūtra and Vinaya texts. The name Abhidharma Treatise in Eight Sections indicates that this material was organized into eight main sections. These eight are further subdivided into forty-four subsections. The Source of Knowledge takes over this general division into eight sections and forty-four subsections as outlined in Table 11.1. Kātyāyanīputra states that the “highest mundane factor” (laukikāgradharma) – i.e., the transition point from an ordinary being ( pṛthagjana) to being a noble (ārya) person – is “abandoning the affairs of an ordinary being and obtaining the doctrine of the Buddha; abandoning heterodox affairs and obtaining the true doctrine” (T.26.1543, 771c15–16). Moreover, he argues, all heterodox views are to be abandoned through one of the four noble truths that encapsulate Buddhist doctrine: through insight into the truth of suffering, the truth of the origin of suffering, the truth of the extinction of the origin of suffering, or the truth of the path leading to the extinction of the origin of suffering (T.26.1543, 772c9–20). Thus, Kātyāyanīputra

175

Bart Dessein Table 11.1 Organization of Abhidharma materials in the Abhidharma Treatise in Eight Sections and the Source of Knowledge. Main Section

Subsection

Main Topics

1. Miscellany (*saṃkῑrṇa) T.26.1543, 771b21–784b25 T.26.1544, 918a7–929b2 2. Fetters (*saṃyojana) T.26.1543, 784c7–812a12 T.26.1544, 929b10-951a6 3. Knowledge ( jñāna) T.26.1543, 812a21–841a27 T.26.1544, 951a6–972a8 4. Action (karman) T.26.1543, 841b7–854a1 T.26.1544, 972a16–981c1 5. The four fundamental material elements (mahābhūta) T.26.1543, 854a21–867a5 T.26.1544, 981c9–991b17 6. Faculties (indriya) T.26.1543, 867a16–887a16 T.26.1544, 991b18–1007c23 7. Concentration (samādhi) T.26.1543, 887b7–905a16 T.26.1544, 1008a6–1027b11 8. Views (dṛṣṭi) T.26.1543, 905a27–917b8 T.26.1544, 1027b12–1031c28

1–8

Structure of the path and foundational topics in Buddhist philosophy Three fetters responsible for the arising of ninety-eight contaminants Knowledge, wisdom, insight, and the types of persons who possess them Virtuous and unvirtuous actions, their causes, and their results

9–12 13–17 18–22 23–26

27–33

The four material elements and their derivatives, and the relationship of the fetters to materiality Relationship of the faculties to the material world and to the fetters

34–38

Forms of concentration obtained in various cosmic realms

39–44

Range of Buddhist and nonBuddhist views

introduces the innovative element of a “path structure” – i.e., the path of vision (darśanamārga) that is the instrument to annihilate obstructions to attaining nirvāṇa – as a foundational topic of Sarvāstivāda philosophy. The name “path of vision” refers to the idea that it is through attaining vision and realization of the four noble truths that defilements (kleśa) are gradually eliminated and eventually will no longer arise, thus leading to the attainment of nirvāṇa. It is noteworthy, however, that the concept of the “highest mundane factor” and the “path of vision” are the first topics discussed in Kātyāyanīputra’s work. It thus appears that this text was not meant for ordinary beings who would first need to be instructed on the nature of “ordinariness” and who, once convinced of the loathsomeness of ordinary life, would start to follow the Buddhist path, but for adepts (śrāvaka) who are already convinced of the necessity to develop the path of vision (Armelin 1978, 15). This observation is corroborated by the fact that the remaining seven subsections of this first main section, “Miscellany,” discuss highly complex doctrinal topics such as dependent origination ( pratῑtyasamutpāda) and the characteristic marks (lakṣaṇa) of conditioned factors (saṃskṛta dharma). The issue of dependent origination concerns the list of the so-called twelve members of dependent origination that explains that karmic activity is the force that continues the cycle of transmigration (saṃsāra). This is a list of predominantly psychic concepts such as ignorance regarding the Buddhist doctrine, the primary perceptual consciousness of a newborn baby, contact (sparśa) of the senses with the outside world, the feelings that result from these contacts, the craving for continuation or discontinuation of 176

Kātyāyanīputra and Mahāvibhās∙ā

these feelings, and concepts such as existence, birth, and decay and dying that directly relate to a human being’s physical existence.7 This list is related to the following famous teaching of the Buddha that denies the existence of a “self ” and that is to be read in the Vinaya: Then the Lord addressed the group of the five monks, saying: “Body, monks, is not self.  .  .  . Feeling is not self.  .  .  . Perception is not self.  .  .  . Consciousness is not self. . . . What do you think about this, monks? Is body permanent or impermanent?” “Impermanent, Lord.” “But is that which is impermanent painful or pleasurable?” “Painful, Lord.” . . . “Seeing in this way, monks, the instructed disciple of the ariyans disregards body and he disregards feeling and he disregards perception and he disregards the habitual tendencies and he disregards consciousness; disregarding he is dispassionate; through dispassion he is freed; in freedom the knowledge comes to be: ‘I am freed,’ and he knows: ‘Destroyed is birth, lived is the Brahma-faring, done is what was to be done, there is no more of being such or such.’ ”8 The twelve links in the chain of dependent origination may explain how transmigration is caused by karmic activity (for example, craving results from sense contact). However, it does not explain how karmic activity itself functions. For this reason, a list of four conditions ( pratyaya) and a redundant list of six causes (hetu) from the Sūtra literature were brought into the philosophical discussion.9 As the concept of “causes and conditions” is less directly linked to the concept of a human being’s individual life than the concept of the twelve members of dependent origination is, it gradually took over the role of the twelve members to explain the general principle of causality. Meanwhile, the list of twelve members was restricted to explaining the rebirth of a human being as one feature of the general process of causality. It is important to note here that Kātyāyanīputra explains the concept of dependent origination in the subsection on human beings of the first main section of his text, and tradition has it that when the Abhidharma Treatise in Eight Sections was translated into Chinese, a section on “causes and conditions” (hetupratyaya) was at first forgotten and added later. This may suggest that: (1) Kātyāyanīputra had intended his text for the adept who envisaged his personal liberation, not for a larger audience of ordinary human beings; and (2) that it was only after the compilation of the Abhidharma Treatise in Eight Sections that the concept of “causes and conditions” gained prominence. Intrinsically linked to the concept of “dependent origination” are the concepts “birth” (utpāda), change in continuance (sthityanyathātva), and passing away (vyava) as characteristic marks of the conditioned. In the subsection that deals with these characteristic marks of conditioned factors, the question of whether entities that transgress through time because of the dynamics of causality exist as actual entities in the three time periods of past, present, and future is addressed. While all present entities are the karmic result of a cause that, in the present moment of time, already belongs to the past, they cannot exist without this past cause. Given the dynamics of causality, all present entities continuously form a future result. The momentariness of time implies that all present entities instantaneously become entities of the past and will live on as causes for the future entities that are now being formed. Causality and momentariness thus make it impossible that any conditioned factor would possess an enduring self-nature (svabhāva). Commenting on the Source of Knowledge, the Large Commentary (on the Abhidharma) states: Furthermore, another reason why the Source of Knowledge presents this topic is that there are some who are deluded with regard to the self-nature of the entities of the three time periods, denying the existence of the past and future entities, and who 177

Bart Dessein

maintain that the present entities are unconditioned. To repudiate their position, it is shown that the essence and the characteristic marks of the past and future entities truly exist, and that the present entities are conditioned. (T.27.1545, 393a18–20) It is because the Buddhist adept knows that all conditioned factors (including human beings as conditioned factors) are without self-nature that he turns away from worldly life, seeking to annihilate all volitional actions that would form fetters that would bind him to this world. The introduction of concepts such as a path of vision, dependent origination, and factors that arise in dependency upon causes (hetu) and conditions ( pratyaya) thus all relate to the ultimate aim of Buddhist practice: the removal of defilements that is liberation. In the four subsections of the second main section, “Fetters,” Kātyāyanīputra analyzes defilements in more detail. This section explains how three fetters (saṃyojana) are responsible for the arising of ninety-eight contaminants (anuśaya), i.e., proclivities that result in the karmic activities of human beings. The text states (T.26.1543: 785c19–20): “How many of these three fetters have a retribution (savipāka) and how many do not have a retribution? Answer: All that is unvirtuous (akuśala) has a retribution.” These three fetters are: (1) the view that there is a self (satkāyadṛṣṭi); (2) doubt (vicikitsā) regarding whether defilements can be annihilated through vision of the four noble truths; and (3) the view that merely adhering to precepts and vows will lead to liberation (śῑlavrataparāmarśa). The view that there is a self can be regarded as the fundamentally erroneous conviction that causes a human being to perform volitional actions for the benefit of maintaining human life and making it pleasurable.10 Each of these volitional actions relates to craving (tṛṣṇā) and is incited by a proclivity. All volitional actions have a karmic result and lead to rebirth and suffering in a future life. Knowledge of the non-self of all conditioned factors and the conviction that insight in the four Buddhist truths can annihilate these proclivities, both of which are addressed in the first main section of the text, will bring the cycle of rebirths to a halt. Conversely, doubt regarding the efficacy of insight into the noble truths in eliminating defilements may put the working of the proclivities back in operation. As Kātyāyanīputra writes (T.26.1543, 785a23), “In the sense realm (kāmadhātu) of ordinary existence, doubt is unvirtuous.” While these first two fetters are related to an adept’s knowledge and understanding, the third fetter is more related to his sincerity: it is not the keeping of the monastic precepts and vows as such that will determine whether or not an adept will attain liberation, but rather the degree of his conviction. This final fetter suggests that Kātyāyanīputra indeed intended a readership of monastics. This section also explains the annihilation of the contaminants in its relation to the three realms of the Buddhist cosmology: the sense realm (kāmadhātu) of ordinary existence; the material realm (rūpadhātu) where, through meditation, sensual desire for the material is brought to a halt; and the etheric realm of formlessness (ārūpyadhātu) that pertains to the domain of higher forms of meditation. This content is presented in a more systematic way in the Heart of the Abhidharma (*Abhidharmahṛdaya), a text that will be discussed in the following section. The third main section, “Knowledge,” deals with the types of knowledge (  jñāna) the adept gains in the different stages that lead to becoming an arhat, i.e., a liberated person. Numerical lists of types of knowledge are hereby discussed according to their mutual relations (T.26.1543, 821a17–20). Although it is clear from this discussion that the types of knowledge are related to the four noble truths and to the different cosmic realms – conventional knowledge relating to the sense realm, and subsequent knowledge relating to the material realm

178

Kātyāyanīputra and Mahāvibhās∙ā

and the realm of formlessness – the precise nature of this relationship is not systematically explained. The fourth main section, “Action,” discusses the difference between virtuous, or skillful (kuśala), and unvirtuous, or unskillful (akuśala), actions, their causes, and their karmic result in a particular form of rebirth. As with the treatment of the types of knowledge, a multitude of doctrinal lists are mutually interlinked. The question of which type of wrong behavior is the gravest offense is answered in a way that again points to the monastic readership of Kātyāyanīputra’s text (T.27.1543, 843a26–27): “Dividing the Saṃgha (Buddhist community) through lies (mṛṣāvāda) is the gravest offence. Because of this behavior, one will be retributed with a kalpa-long life in the Avīci hell (the deepest of all hells).” While the first four main sections of the text thus share their concern with the issues of defilement, karmic consequences, and the method to reach liberation, the fifth main section, “The four fundamental material elements,” stands out as the first occurrence in Sarvāstivāda Buddhism of a discussion of the material aspect of the cosmos. Indeed, while the second main section of the text brought in the concept of the three cosmic realms, materiality as such was not yet discussed. This fifth main section explores the four fundamental material elements (mahābhūta): earth ( pṛthivī), water (ap), fire (tejas), and wind (vāyu), as well as the different forms of material elements derived therefrom. This materiality is distinguished from the mental domain of thoughts (citta), from thought-concomitants (caitasika) such as mindfulness, and from those formations that are dissociated from thoughts but cannot be considered matter either (cittaviprayuktasaṃskāra).11 Toward the end of the section, these material elements are brought into relation with the “fetters” discussed in the second main section, as it is more precisely explained how a human being is bound to the material world through fetters that are the result of sensual activity. From here, it is a logical step to the sixth main section, “Faculties.” The faculties are also discussed in their relation to the items of other doctrinal lists. One of the relations addressed is between the faculties and the material world, leading to defilements and their karmic consequences. The seventh main section, “Concentration,” goes on to discuss the different forms of concentration that are attained in various cosmic realms. Special attention is paid to the states of the nonreturner (anāgāmin) and the once-returner (sakṛdāgāmin), two stages preceding the attainment of arhat-ship. Finally, the last main section, “Views,” mentions some erroneous views (mithyādṛṣṭi): the view that the present and the future do not exist (T.26.1543, 913a17); the view that there is permanence (T.26.1543, 913a26); and the view that human beings have defilements that have no causes and conditions (T.26.1543, 913b5). These critiques mark the self-identification of Sarvāstivādins as a distinct doctrinal group within the larger Buddhist community. In these eight sections, Kātyāyanīputra combined concepts enumerated in doctrinal lists, from time to time introducing innovative interpretations. He did not, however, construct a philosophical system. This gives the text the impression of a scholastic patchwork. Indumati Armelin thus characterizes Kātyātanīputra’s text as merely “a testimony of an effort to systematization” (1978, 6). The lack of system is justified in the Large Commentary (on the Abhidharma) as follows: “The true nature (bhāva) and characteristic marks (lakṣaṇa) of the entities (dharma) within the Abhidharma should be sought, not their order of presentation” (T.27.1545, 1c23–24). As with earlier Abhidharma texts, then, Kātyāyanīputra’s work is not a fully elaborated philosophical system. Nevertheless, his presentation of doctrinal concepts such as the principle of causality; the three time periods of past, present, and future; and the possibility of

179

Bart Dessein

annihilating defilements through applying insight into the four noble truths and thereby bringing an end to the dynamics of causality all became fundamental for later Sarvāstivāda philosophy. This explains why Kātyāyanīputra is often seen as the founder of the Sarvāstivāda school. Moreover, while Kātyāyanīputra’s text may focus on the adept’s individual liberation, by bringing in concepts such as “causes and conditions” and the “three time periods,” it also testifies to an ontological development within Buddhism. This general ontological approach began to replace Buddhist doctrine as a method for personal, psychological healing, and came to characterize the Sarvāstivādins as a self-identified doctrinal group.

The Abhidharma Treatise in Eight Sections: The Source for New Doctrinal and Textual Traditions The Heart Treatises: Sarvāstivāda Pedagogical Compendia The doctrinal elements that were presented in an often unsystematized manner in the Abhi‑ dharma Treatise in Eight Sections/Source of Knowledge were for the first time brought into a well-structured philosophical treatise in the Tocharian *Dharmaśreṣṭhin’s Heart of the Abhidharma (*Abhidharmahṛdayaśāstra). The work is preserved in a Chinese translation (T.28.1550) completed in 391 CE and seems to be intended to be a widely accessible pedagogical digest aimed at a non-Buddhist readership. The Heart of the Abhidharma offers a systematic account of what must be understood and practiced in order to follow the path that leads from the condition of an ordinary human being to that of an arhat. This path is outlined in Table 11.2. The first chapter, “Elements,” sets out from an analysis of the world in which ordinary human beings live. This chapter presents reality according to five general categories ( pañcavastuka): aggregates (skandha), sense(field)s (āyatana), elements (dhātu), thoughts (citta), and formations dissociated from thoughts (cittaviprayukta saṃskāra). The aggregates are a list of five concepts that  – according to the Buddhist doctrine  – constitute a human being: matter (rūpa), feelings (vedanā), conceptualizations (saṃjñā), conditioning factors (saṃskāra), and consciousness (vijñāna). The sense(fields)s are the faculties of sight, hearing, smelling, tasting, touching, and thinking, and their respective object-fields (matter, sound, smell, taste, the tangible, and ideas and concepts). The elements are the same sense faculties and their respective object-fields, with the addition of the six types of consciousness that are the result of the contact of faculties and objects. *Dharmaśreṣṭhin thus starts his doctrinal outline with an explanation of the nature of a human being (the aggregates), the interaction of a human being with the outside world (the sense[field]s that constitute a human being as a psychological being [the elements]), and his life as a conscious being in the sensual world (thoughts and formations dissociated from thoughts). *Dharmaśreṣṭhin also connects these different categories through the concept of causes and conditions by explaining that because of the dynamics of causality, everything in the mundane world is devoid of a self-nature (svabhāva). After a description of the material world in which human beings live, the way a human being, through her faculties, interacts with this material world is described in the second chapter, “Conditioning Factors.” This chapter addresses all the psychic formations of a human being. Hence, these first two chapters can be regarded as corresponding to the first of the Buddhist four noble truths, the truth of suffering. From a pedagogical perspective, it is necessary to understand the true nature of human life before one can become convinced of the necessity of following the Buddhist path.

180

Kātyāyanīputra and Mahāvibhās∙ā Table 11.2  The path of vision and the spiritual path. Spiritual Path Arhat Non-Returner (has annihilated all nine types of proclivities of the sense realm) Once-Returner (has annihilated eight types of proclivities of the sense realm) Stream-Enterer (has annihilated five types of proclivities of the sense realm) Path of Vision Sense Realm (14) Knowledge regarding the doctrine of the path: path of liberation (13) Forbearance regarding the doctrine of the path: immediate path (10) Knowledge regarding the doctrine of cessation: path of liberation (9) Forbearance regarding the doctrine of cessation: immediate path (6) Knowledge regarding the doctrine of the origin: path of liberation (5) Forbearance regarding the doctrine of the origin: immediate path (2) Knowledge regarding the doctrine of suffering: path of liberation (1) Forbearance regarding the doctrine of suffering: immediate path

Realm of Form Realm of Formlessness (16) Subsequent knowledge regarding the doctrine of the path: path of liberation (15) Subsequent forbearance regarding the doctrine of the path: immediate path (12) Subsequent knowledge regarding the doctrine of cessation: path of liberation (11) Subsequent forbearance regarding the doctrine of cessation: immediate path (8) Subsequent knowledge regarding the doctrine of the origin: path of liberation (7) Subsequent forbearance regarding the doctrine of the origin: immediate path (4) Subsequent knowledge regarding the doctrine of suffering: path of liberation (3) Subsequent forbearance regarding the doctrine of suffering: immediate path

Four Elements That Are Conducive to Insight – Highest mundane factor – Forbearance – Summit – Warmth Four Applications of Mindfulness – Application of mindfulness on entities – Application of mindfulness on thoughts – Application of mindfulness on feelings – Application of mindfulness on the body Three Preparatory Exercises – Development of the analysis of the entities – Mindfulness of breathing in and breathing out – Contemplation of the repulsive

181

Bart Dessein

Knowledge of the true nature of the mundane world reveals that the cause of rebirth and suffering are volitional actions. This is explained in the third chapter, “Actions,” in which volitional actions are differentiated into unwholesome corporeal, verbal, and mental actions. This is then followed in the fourth chapter by an analysis of the basic moods, i.e., the proclivities (anuśaya) that are the fundamental reason why human beings perform these actions. There is a basic set of ten proclivities: five false views, attachment, repugnance, pride, doubt, and ignorance. Actions and proclivities correspond to the second Buddhist truth, the truth of the origin of suffering. The analysis of actions and proclivities enables the adept to eliminate the proclivities through progressing on the path to liberation, which is the subject of the fifth chapter, “Nobility.” Subsequently, the types of knowledge and the types of concentration a Buddhist adept attains while progressing on the path to liberation are discussed in the sixth and seventh chapters, “Knowledge” and “Concentration.” These chapters thus correspond to the third and fourth Buddhist truths, the truths of the extinction of the origin of suffering and the path leading to the extinction of the origin of suffering. As noted by Frauwallner (1971, 123), the fourth and fifth chapters of the Heart of the Abhidharma should be seen as the real core of the doctrine: “Proclivities” are the dynamic force that lead a human being to perform actions and, hence, are the ultimate cause of rebirth, and “Nobility” is the description of the different stages one goes through as a result of eliminating the proclivities on the path of vision. The central importance of the fifth chapter explains why it is in this chapter that the text transitions from theoretical approach to practice. This chapter explains with more precision how three preparatory exercises that were briefly mentioned in the first chapter lead to the conviction that life is only characterized by suffering and that the Buddhist path is the only path that can bring this suffering to a halt: contemplation of the repulsive (aśubhabhāvanā), mindfulness on breathing in and breathing out (ānāpānasmṛti), and the development of the analysis of the entities (dhātuvyavasthā). Contemplation of the repulsive consists of nine subsequent stages in which the repulsiveness of the human body is contemplated. Mindfulness on breathing in and breathing out consists of concentration on the stream of air when breathing, and on how this stream of air manifests itself in the body. Analysis of the entities is the analysis of six elements: earth ( pṛthivī), water (ap), fire (tejas), wind (vāyu), space (ākāśa), and consciousness (vijñāna). Following these three preparatory exercises, there are four applications of mindfulness: on the body (kāyasmṛtyupasthāna), on feelings (vedanāsmṛtyupasthāna), on thoughts (cittasmṛtyupasthāna), and on entities (dharmasmṛtyupasthāna). These four are a logical continuation of the realization that the body is subject to suffering. Contemplation on the body is the remedy against the erroneous view that what is unclean (aśubha), i.e., the body, is seen as clean (śubha); contemplation on feelings serves as remedy against the wrong conceptualization (mithyāsaṃjñā) of suffering (duḥkha) as happiness (sukha); contemplation on thoughts is a remedy for the conceptualization of what is impermanent (anitya) as permanent (nitya); and contemplation on the entities serves as remedy against the conceptualization that what is without self-nature (anātmaka) has a self-nature (ātman). Four elements that are conducive to insight (nirvedhabhāgīya) follow. In the first stage, “warmth” (ūṣman), the four noble truths are understood. Because of warmth, correct effort (samyagvyāyāma) and correct deliberation (samyaksaṃkalpa) of the adept increase, and these two attitudes are accompanied by virtuous actions. This brings the adept to the second mental state conducive to insight, called “summit” (mūrdhan), in which the four noble truths are fully understood. In the third stage, “forbearance” (kṣānti), the adept enjoys the four noble truths. The most extreme form of forbearance is the condition that is a direct antecedent of the highest mundane factor (laukikāgradharma), which is the highest quality an ordinary being can attain before she becomes an arhat. In this stage, the adept is able to contemplate the unsatisfactoriness pertaining to the mundane realm 182

Kātyāyanīputra and Mahāvibhās∙ā

in one moment only. The fact that “the highest mundane factor” appears only at this stage of the text again indicates that it was oriented toward a different readership than the text by Kātyāyanīputra. Once the adept has developed the four elements that are conducive to insight, he enters the actual noble path (āryamārga) that has two components: the path of vision (darśanamārga) and the spiritual path (bhāvanāmārga). The path of vision has sixteen moments: eight forms of forbearance and eight forms of knowledge. The first moment of the path of vision is forbearance regarding the doctrine of suffering (duḥkhe dharmakṣānti), which is the moment when the adept gains insight in the doctrine of suffering. Forbearance regarding the doctrine of suffering is called an immediate path (ānantaryamārga), because as soon as the adept has this insight, proclivities bound to the sense realm are immediately annihilated. The second moment of the path of vision produces a form of knowledge ( jñāna) that has the same object, the doctrine of suffering (duḥkhe dharmajñāna). It annihilates the proclivities of the sense realm through insight into the truth of suffering, ensuring that the adept will remain free of this type of defilement. While the first two moments were concerned with the sense realm, the third and fourth moments of the path of vision are, respectively, the subsequent forbearance regarding the doctrine of suffering (duḥkhe’nvayakṣānti) and the subsequent knowledge regarding the doctrine of suffering (duḥkhe’nvayajñāna), which are directed toward the suffering of the higher two realms of existence. These four types of forbearance and knowledge that regard suffering ensure the complete annihilation of that part of the proclivities that is to be annihilated through insight in the truth of suffering. The same structure of four moments applies successively to the insights of the origin of suffering, the extinction of the origin of suffering, and the path that leads to the extinction of the origin of suffering. But even after the path of vision has been completed, proclivities are not fully eliminated – a fifth method of elimination, belonging to the spiritual path, is necessary. The sixteenth moment of the path of vision is thus considered to be the first moment of the spiritual path. Through spiritual practice, one also has to annihilate those proclivities that have not been completely annihilated on the path of vision. As soon as the adept enters on the spiritual path, he is on his way to attaining four successive fruits of the seeker (śrāmanyaphala), culminating in arhat-ship. The arhat is no longer reborn because he has stopped all karmic actions. *Dharmaśreṣṭhin’s text became the fundamental work of the Sarvāstivāda subgroup that resided in Bactria and Gandhāra and that formulated its interpretation of the Buddhist doctrine in a series of pedagogical digests, i.e., texts explaining the essence or heart (hṛdaya) of the Abhidharma. As mentioned earlier, this subgroup of the Sarvāstivāda monastic community claimed that its interpretation of the doctrine was closer to the original sermons of the Buddha. *Dharmaśreṣṭhin’s text informed later Sarvāstivāda developments in Gandhāra. These developments in turn informed Vasubandhu’s Treasury of the Abhidharma (Abhidharmokośa). That Vasubandhu, when requested by the masters of Kaśmῑra to write commentaries on the stanzas (kārikā) of his Treasury of the Abhidharma, criticized the Vaibhāṣika viewpoints attests to the different doctrinal interpretations that existed within the larger Sarvāstivāda monastic group and to the knowledge of Vaibhāṣika viewpoints in Gandhāra. The Treasury of the Abhidharma is regarded as the final pedagogical digest of Sarvāstivāda thought, and also as anticipating Yogācāra thought.12

The Vibhāsā Commentaries As mentioned, with the vibhāṣā commentarial works, we touch upon the distinctive development of Kaśmīra in contrast to Gandhāran Buddhism. This contrast can be traced to the second council of Pāṭaliputra (contemporary Patna in northeastern India), which, according to tradition, 183

Bart Dessein

took place at the beginning of the second century after the historical Buddha’s demise. This synod was supposed to settle a dispute between the Sthaviravādins, presided over by Tissa Moggaliputta (Maudgaliputra), and the Sarvāstivādins, who were headed by Kātyāyanīputra.13 When the debate was judged in favor of Tissa Moggaliputta, some of the Sarvāstivādins are reported to have moved to Kaśmῑra. They considered the Source of Knowledge to be the definitive text of Sarvāstivāda philosophy and they referred to this text as the “body” (śarῑra) of Sarvāstivādin doctrine that is sustained by six earlier (central Indian) Sarvāstivāda works to which they referred as “feet” ( pāda). They also composed commentarial works (vibhāṣās) on Kātyāyanῑputra’s work. These vibhāṣās gave their name to the Vaibhāṣika subgroup of Sarvāstivādins. Those Sarvāstivādins who stayed behind emphasized *Dharmaśreṣṭhin’s Heart of the Abhidharma, a work that gave rise to a series of works culminating in Vasubandhu’s Treasury of the Abhidharma. That the Sarvāstivādins of Kāśmῑra valued the Source of Knowledge over the Abhidharma Treatise in Eight Sections can be traced back to a further council that was held in Kaśmῑra. In his biography of Vasubandhu, Paramārtha informs us that 500  years after the Buddha’s parinirvāṇa, Kātyāyanῑputra, who had gone forth in the school of the Sarvāstivādins, convened a synod in Kaśmῑra, which was attended by five hundred arhats and five hundred bodhisattvas. It is at this synod that all sayings concerning the Sarvāstivāda Abhidharma were collected and compiled (zhuan ji sapoduobu apidamo 撰集薩婆多部阿毘達磨) and made (zhi 製) into a work that consists of eight sections: the Source of Knowledge (T.50.2049, 189a1–26). We may also reiterate here that Nāgārjuna, in the Instruction on the Great Perfection of Wisdom, claimed that the text that Kātyāyanῑpitra composed in eight sections serves to “initiate knowledge.” The Life of Vasubandhu also claims that the Kuṣāṇa king Kaniṣka, who appears in the account of the synod, prohibited the vibhāṣā that was written on the Source of Knowledge from being taken out of the country (T.50.2049, 189a26–b24). This suggests that the synod in Kaśmῑra was meant not to settle a dispute with rival schools but rather to establish Kaśmῑra, i.e., Vaibhāṣika, self-identity.14 There is good reason to date the extant Chinese version of the Large Commentary (on the Abhidharma) later than Kaniṣka and, more precisely, in a period in which the philosophy of the Mahāyāna was already well developed in Kaśmῑra. The vibhāṣā compendia were not only commentaries on earlier Abhidharma texts, but some of the categories that figure prominently in the Large Commentary (on the Abhidharma) – the analysis of materiality (rūpa), the concept of the three vehicles (the vehicle of the disciples, the vehicle of the solitary buddhas, and the vehicle of the bodhisattva), and the bodhisattva’s spiritual development in ten stages (bhūmi), for example – would also figure prominently in Yogācāra Buddhism. As Robert E. Buswell and Padmanabh S. Jaini suggest, the Large Commentary (on the Abhidharma) is therefore also “a major source of material on early Mahāyāna developments” (Buswell and Jaini 1996, 119).

Abbreviations Abhidh-k-vy Chos ’byung T. Tāranāthae

Sphuṭārthā Abhidharmakośavyākhyā by Yaśomitra Chos ’byung by Bu ston Taishō shinshū daizōkyō大正新修大藏經 Tāranāthae de Doctrinae Buddhicae in India Propagatione Narratio contextum Tibeticum

184

Kātyāyanīputra and Mahāvibhās∙ā

Notes 1 For a discussion of the development of Abhidharma texts from mātṛkās, see Dessein (2013); for a discussion of the term “Abhidharma,” see Willemen, Dessein, and Cox (1998, 13–14). 2 A date of the second century BCE for the self-identification of the Sarvāstivādins has been suggested by Hirakawa (1974, vol. 1, 143), and a date of the first century BCE by Shizutani (1978, 48 ff). 3 In early Buddhist texts, it is sometimes stated that practicing the items of the list of thirty-seven “limbs of enlightenment” is the way to make an end to defilement (āsrava). See Dessein (2013, 33). For the transition from what has been referred to as the period of “psychological Buddhism” as it was succinctly expressed in the list of thirty-seven “limbs of enlightenment” to the period of “ontological Buddhism,” see Bronkhorst (1985). 4 See Abhidh-k-vy 9.11–12. These six are the T.26.1536, T.26.1537, T.26.1538, T.26.1539, T.26.1540, and T.26.1541 (also translated in T.26.1542). 5 For the Sanskrit tradition, see Abhidh-k-vy 9.12–14 and 12.4 ff.; 11.24–29. For the Tibetan tradition, see Chos ’byung, vol. 1, 49. 6 For an in-depth study of these elements, see Dessein (2012). 7 The standard list of twelve mutually interlinked members of dependent origination appears to be a compilation of smaller series, the two most prominent of which are a seven-membered (starting with ignorance and leading to feeling) and a five-membered formula (starting with craving and leading to decay-and-dying). See Wayman (1970–1971, 186). 8 Mahāvagga of the Pali Vinaya. Edition Oldenberg (1880, 13–14); translation Horner (1962, 20–21). 9 For a discussion of the four conditions and six causes, and the distinction between conditions and causes, see Stcherbatsky (1958, vol. 2, 138). 10 Note that according to the Buddhist doctrine, only volitional actions have a karmic result. 11 The formations dissociated from thoughts are a mixed list of concepts such as types of meditative attainment (samāpatti) and the linguistic concepts of syllables, words, and sentences. 12 There are two translations into Chinese of Vasubandhu’s Treasury of the Abhidharma: one done by Paramārtha in 565 CE, entitled Apidamo jushe shi lun阿毘達磨俱舍釋論 (T.29.1559), and one done by Xuanzang in 653 CE, entitled Apidamo jushe lun 阿毘達磨俱舍論 (T.29.1558). The version by Xuanzang was translated into French by Louis de La Vallée Poussin ([1923–1931] 1971) and into English by Lodrö Sangpo (2012). The French translation by Louis de La Vallée Poussin was translated into English by Leo Pruden (1991). For more on Vasubandhu, see Chapter 16 in this volume. 13 See Bareau (1955, 168–69). 14 For debates on the purpose of this synod, see Roerich (1949, 25); Obermiller (1931–1932, 96–97); Tāranāthae 47, 7 ff; Schiefner (1869, 58–61); Frauwallner (1952, 252); Bareau (1955, 132).

Bibliography Primary Sources Chos ’byung. Obermiller, Evgeniy Evgenevitsj, trans. 1931–1932. History of Buddhism by Bu-Ston, Part 1: The Jewelry of Scripture; Part 2: The History of Buddhism in India and Tibet. Materialen zur Kunde des Buddhismus 18–19. Heidelberg: In Kommission bei O. Harrassowitz, Leipzig. Sphuṭārthā Abhidharmakośavyākhyā. Wogihara, Unrai, ed. 1932. Sphuṭārthā Abhidharmakośavyākhyā: The Work of Yaśomitra. Tokyo: The Publishing Association of the Abhidharmakośavyākhyā. Taishō shinshū daizōkyō大正新脩大藏經. Takakusu, Junjirō 高楠順次朗 and Watanabe, Kaigyoku 渡邊 海旭, ed. 1924–1934. Taishō shinshū daizōkyō 大正新脩大藏經. Tokyo: Taishō issaikyō kankōkai. ———. T.23.1435: Daśādhyāyavinaya, Shi song lü 十誦律, Puṇyatara, Dharmaruci, Kumārajīva. ———. T.26.1536: Mahākauṣṭhila (Skt.), Śāriputra (Ch.), [Abhidharma]saṃgῑtiparyāya[pādaśāstra], Apidamo jiyimen zulun 阿毘達磨集異門足論, Xuanzang玄奘. ———. T.26.1537: Śāriputra (Skt.), Mahāmaudgalyāyana (Ch.), [Abhidharma]dharmaskandha[pāda­ śāstra], Apidamo fayun zulun 阿毘達磨法蘊足論, Xuanzang玄奘. ———. T.26.1538: Maudgalyāyana (Skt.), Mahākātyāyana (Ch.), Prajñaptiśāstra, Shishe lun 施設論, Dharmapāla, Weijing 惟淨.

185

Bart Dessein ———. T.26.1539: Devaśarman, [Abhidharma]vijñānakāya[pādaśāstra], Apidamo shishen zulun阿毘 達磨識身足論, Xuanzang玄奘. ———. T.26.1540: Pūrṇa (Skt.), Vasumitra (Ch.), [Abhidharma]dhātukāya[pādaśāstra], Apidamo jieshen zulun阿毘達磨界身足論, Xuanzang玄奘. ———. T.26.1541: Vasumitra, [Abhidharma]prakaranapāda[śāstra], Zhongshifen apitan lun 眾事分阿 毘曇論, Guṇabhadra and Bodhiyaśas ———. T.26.1542: Vasumitra, [Abhidharma]prakaranapāda[śāstra], Apidamo pinlei zulun阿毘達磨品 類足論, Xuanzang玄奘. ———. T.26.1543: Kātyāyanῑputra, *Abhidharmāṣṭagrantha[śāstra]/*Abhidharmāṣṭaskandha[śāstra], Apitan ba jiandu lun 阿毘曇八犍度論, Saṃghadeva and Zhu Fonian 竺佛念. ———. T.26.1544: Kātyāyanῑputra, Jñānaprasthāna, Apidamo fazhi lun 阿毘達磨發智論, Xuanzang 玄奘. ———. T.27.1545: 500 arhats, *[Abhidharma]mahāvibhāṣā[śāstra], Apidamo da piposha lun 阿毘達磨 大毘婆沙論, Xuanzang 玄奘. ———. T.28.1550: Dharmaśreṣṭhin, *Abhidharmahṛdaya[śāstra], Apitan xin lun 阿毘曇心論, Saṃghadeva, Huiyuan 慧遠. ———. T.29.1558: Vasubandhu, Abhidharmakośaśāstra, Apidamo jushe lun 阿毘達磨俱舍論. Xuanzang 玄奘. ———. T.29.1559: Vasubandhu, Abhidharmakośaśāstra, Apidamo jushe shi lun阿毘達磨俱舍釋論. Paramārtha. ______. T.50.2049: Paramārtha, Posoupandou fashi zhuan 婆藪槃豆法師傳. ———. T.70.2300: Paramārtha, San lun xuanyi jian you ji 三論玄義檢幽集. Tāranāthae de Doctrinae Buddhicae in India Propagatione Narratio contextum Tibeticum. Schiefner, Anton, ed. 1868. Petropoli: E. Codicibus Petropolitanis, Eggers.

Secondary Sources Armelin, Indumati. 1978. Le Coeur de la Loi Suprême. Traité de Fa-cheng. Abhidharmahṛdayaśāstra de Dharmaśrῑ. Paris: Librairie Orientaliste Paul Geuthner. Bareau, André. 1955. Les sectes bouddhiques du petit véhicule. Paris: École française d’Extrême-orient. Bechert, Heinz. 1973. “Notes on the Formation of Buddhist Sects and the Origins of Mahāyāna.” In German Scholars on India. Contributions to Indian Studies, edited by Cultural Department of the Embassy of the Federal Republic of Germany, 6–18. Varanasi: Chowkhamba Sanskrit Series Office. Bronkhorst, Johannes. 1985. “Dharma and Abhidharma.” Bulletin of the School of Oriental and African Studies 47 (2): 305–20. Buswell, Robert E., and Padmanabh S. Jaini. 1996. “The Development of Abhidharma Philosophy.” In Encyclopedia of Indian Philosophies. Volume VII. Abhidharma Buddhism to 150 A.D., edited by Karl H. Potter, 73–119. New Delhi: Motilal Banarsidass. Demiéville, Paul. 1961. “Un fragment Sanskrit de l’Abhidharma des Sarvāstivādin.” Journal Asiatique 249: 461–65. Dessein, Bart. 2012. “ ‘Thus Have I Heard’ and Other Claims to Authenticity: Development of Rhetorical Devices in the Sarvāstivāda Ṣaṭpādābhidharma Texts.” In Zen Buddhist Rhetoric in China, Korea, and Japan, edited by Christoph Anderl, 121–62. Leiden and Boston: Brill. ———. 2013. “Lists as Dynamic Devices in Early Buddhist Doctrine and Textual Tradition.” Antiquorum Philosophia 7: 29–48. Frauwallner, Erich. 1952. “Die buddhistischen Konzile.” Zeitschrift der Deutschen Morgenländischen Gesellschaft 102: 240–61. ———. 1971. Die Entstehung der buddhistischen Systeme. Göttingen: Vandenhoeck & Ruprecht. Fukuhara, Ryōgen福原亮厳. 1965. Ubu Abidatsumaronjo no hattatsu 有部 阿毘達磨論書の発達. Kyoto: Nagata Bunshōdō. Hirakawa, Akira 平川彰. 1974. Indo Bukkō shi インド佛教史, 2 vols. Tokyo: Shunjūsha. Horner, Isaline Blew. 1962. The Book of the Discipline (Vinaya-Pitaka). Volume IV. Mahāvagga. London: Luzac. La Vallée Poussin, Louis de. (1923–1931) 1971. L’Abhidharmakośa de Vasubandhu, 6 vols. Bruxelles: Institut belge des Hautes Études chinoises.

186

Kātyāyanīputra and Mahāvibhās∙ā Lodrö Sangpo, Gelong. 2012. Abhidharmakośa-Bhāṣya of Vasubandhu. The Treasury of the Abhidharma and Its Commentary, 4 vols. New Delhi: Motilal Banarsidass. Nishi, Giyū 西義雄. 1934. “Ubushūnai ni okeru hotchi hi-hotchi kei nado no shoshu no gakusetsu oyobi gakutō no kenkyū 有部宗內に於ける發智系‧非常發智系等の諸種の學說及學統の研究.” Shūkyō kenkyū 宗教研究11 (4): 564–79; 11 (5): 768–89. Oldenberg, Hermann. 1880. The Vinaya Piṭakaṃ: One of the principal Buddhist holy scriptures in the Pāli language. Vol. II: The Cullavagga. Pali Text Society. London: Williams & Norgate. Pruden, Leo. 1991. Abhidharmakośa of Ācārya Vasubandhu, 4 vols. Berkeley: Asian Humanities Press. Roerich, G. N. 1949. The Blue Annals, 2 vols. Royal Society of Bengal, Monograph Series, vol. VII. Calcutta: The Royal Asiatic Society of Bengal. Schiefner, Anton, trans. 1869. Tāranātha’s Geschichte des Buddhismus in Indien – aus dem Tibetischen übersetzt. Eggers, St. Petersburg: Kaiserliche Akademie der Wissenschaften. Shizutani, Masao 静谷正雄. 1978. Shōjō Bukkyōshi no kenkyū 小乘仏教史の研究. Kyoto: Hyakkaen. Stcherbatsky, Theodore. 1958. Buddhist Logic, 2 vols. The Hague: Mouton. Takakusu, Junjirō. 1904–1905. “On the Abhidharma Literature of the Sarvāstivādins.” Journal of the Pali Text Society: 67–146. Wayman, Alex. 1970–1971. “Buddhist Dependent Origination.” History of Religions 10: 185–203. Willemen, Charles, Bart Dessein, and Collett Cox. 1998. Sarvāstivāda Buddhist Scholasticism. Handbook of Oriental Studies 2/11. Leiden, New York and Köln: Brill.

187

12 THE VĀTSĪPUTRĪYA/ SĀM . MITĪYA

Buddhist Personalism as a Mainstream School of Thought Peter Skilling Introduction1 The self is life’s perpetual puzzle, and the nature of the self is one of the fundamental questions of philosophy. Its scope ranges from the psychological, intellectual, and social makeup of the individual – the human being – and his or her relations to other human beings and to the world, to metaphysical reflections on origins and destiny of the soul and its relation to divinity and time. In many systems of thought, these ruminations respond to and address core ethical questions of agency and moral responsibility. “Longue durée” systems in which the self reincarnates over long and even astronomical periods have to explain how this happens. In Buddhist moral philosophy, it is axiomatic that human deeds or acts entail reactions: they are like seeds that in the course of time will bear fruit for the doer. The rich narrative literature of the Avadānas (tales of the past-life exploits of people connected with Buddhism) that developed to illustrate the workings of karma frequently calls up a verse that makes this point: Human deeds do not just disappear Even after hundreds of millions of eons. When their time comes and conditions ripen Deeds bear fruit for breathing beings. Most of the well-known religious systems of India (Hinduism, Jainism, Buddhism) are such systems. Buddhism is unique in that the landmark teaching of Gautama Buddha is “non-self.” Writing perhaps in the second century CE, a poet praised the Buddha: When the mind is filled with egotism, the succession of rebirths does not cease. Egotism does not leave the heart when the view of self is present. No other teacher in the world proclaims non-self: Therefore, outside of your system of thought there is no other pathway to peace. (Wogihara 1932–1936, 697) Gautama Buddha did not posit the existence of an enduring self, but – on the contrary – deconstructed the individual into components and constituent parts. These he formulated in several 188

DOI: 10.4324/9781351030908-19

The Vātsīputrīya/Sām.mitīya

ways. An early summary was the five aggregates, the twelve sense doors, and the eighteen elements (skandha-āyatana-dhātu), a framework that developed into what modern thinkers have described as a “dharma theory” – an analysis of experience in terms of complex interactions between clusters of material and mental conditions (dharma). For example, Buddhaghosa’s commentary on the Enumeration of Phenomena (Dhammasaṅgaṇi) of the Pali Abhidhamma explains how certain types of mentality arise: When the first of the great skillful mentalities (cittas) of the sense-sphere arises, at that time the fifty and more constituents (dhammas) that have arisen by way of factors of mentality (citta) are indeed ultimate constituents in the sense of self-existents (sabhāva). There is nothing else, whether a being, or an entity, or a man or a person. (Ronkin 2005, 117) The question that Buddhist philosophy had to face is: how does moral responsibility work when there is no soul? In the Sūtra on the Ultimate Meaning of Emptiness (Paramārthaśunyatāsūtra), the Buddha said: It is like this, O monks: moral action (karma) exists and its fruition ( phala, vipāka) exists, but there is no doer or agent (kāraka) who casts aside the aggregates of this present life and takes up the aggregates of the next – except as a conventional designation (dharmasaṃketa). There was, however, one group of schools within the traditionally enumerated eighteen early monastic orders that asserted the existence of a person, an individual ( pudgala) who was neither the same as nor different from the constituent elements of being. To maintain the existence of the pudgala seems, from the contemporary understanding of Buddhism, to be unprecedented and unique. But this notion had a long evolution, starting from the time of the Buddha himself. It is not possible, however, to trace the evolution in any chronological detail. The group of schools that asserted the existence of the person was the Vātsīputrīya/Sāṃmitīya complex of schools. In this chapter, I attempt to delineate the contours of the Vātsīputrīya/Sāṃmitīya presence in India. The topic raises questions about the place of the Vātsīputrīya/Sāṃmitīyas in Buddhist history and in Indian and world philosophical thought – how this has been assessed and how to assess it. This entails examining the history of ideas in the aggregation of Buddhist sanghas, or monastic orders, both synchronically and diachronically – a task that is far beyond the scope of this contribution. Put simply, we need to avoid taking one or the other Buddhist school as original, orthodox, or mainstream, and instead regard the development of Buddhist thought as an organic process of exchange between the sanghas within the broader dynamics of Indian philosophy. The “eighteen Buddhist schools” did not develop unilaterally as discrete and self-contained units. Classical texts preserve several competing narratives of how the schools emerged. The problems raised in this contribution concern the overall development of Buddhism, both within and outside of the framework of the traditional monastic orders, schools of thought, and vehicles.

Buddhist Debates About Self and Person Many canonical passages discuss aspects of non-self and state that the view of a self is a wrong or mistaken view.2 Centuries of discussion and debate on the nature and status of the “person” are intricately intertextual: the different sides cite the Buddha’s words to support their theses. It is a debate among different interpretations of the import of the Buddha’s speech. 189

Peter Skilling

In the Sūtra on the nun Śailā’s encounter with Māra, the latter approaches the arhat nun Śailā and quizzes her about the origin of the “being” (satva). She retorts: Why do you think there’s a sentient being? Māra, this is only your mistaken idea. Empty is this mass of formations: No being is there here. Just as when the parts are assembled We recognize what we call a “chariot”: Like this, dependent on the aggregates, One speaks by convention of a “being.” (La Vallée Poussin 1912, 257–58) Śailā’s riposte was often cited in polemics about the “being”; for instance, in the Topics of Debate (Kathāvatthu, Taylor 1894, 66) and Questions of King Milinda (Milindapañha, Trenckner 1880, 27), and in the works of Kumārajīva (Lamotte 1944, 28), Vasubandhu, Candrakīrti, and other philosophers. We see here a shift from the ātmā to the being, which represents a major reformulation of the philosophical stakes from the critique of the self to a broader spectrum of modes of individuality. It appears that the early discourses accepted and used the notions of the conventional individual, the conscious being or person (satva, pudgala), without much reservation. The Vātsīputrīyas took what might seem to be a natural step and recognized the “person” as a valid constituent of human existence and adopted it into their dharma system. Other schools, including the Theravaṃsa (the lineage of schools claiming to descend from the early theras, the direct followers of Śākyamuni), saw this as a concession to the idea of an enduring ātmā – soul, self, or substantial ego – of mainstream Indian thought and emphasized in their writings that the being or person does not exist and cannot be epistemologically proven. Belief in an enduring soul was considered a serious obstacle to liberation. The shift is evident in the Topics of Debate. The Sarvāstivāda school went further and introduced a long list of synonyms of the being and the person into their Sūtra and Abhidharma literature, denying that any of these iterations of individuality existed. This line of thought strongly influenced the Perfection of Wisdom and Great Vehicle literatures, where similar lists occur regularly. The debate seems in part to have been fueled by the ambiguities of language. The word “person” occurs throughout the Buddhist scriptures, and one Sūtra in particular is known as “discourse on the bearer of the burden” (Bhārāhārasūtra). The Buddha teaches the monks that the burden is the five aggregates and the bearer of the burden is the person. But, if the person who bears the burden is different from the burden, then isn’t the person something more than just a name for the five aggregates? There were many attempts to explain the anomaly. For some, the Sūtra proved that the pudgala is an autonomous reality, but this gave rise to further questions. Their answer was that the person was neither the same as the aggregates nor different from the aggregates: it did not fit the usual abhidharmic categories. It was inexpressible (avācya), an “indeterminate self.”

Vātsīputrīya, Sām · mitīya, Pudgalavāda: Quandaries of Nomenclature

Modern academic writing most commonly calls the school pudgalavāda – “those who assert the factual existence of the person or individual” – in a word: “personalists.” But the Vātsīputrīya/

190

The Vātsīputrīya/Sām.mitīya

Sāṃmitīya family of schools itself did not use this term. Pudgalavāda is not a self-descriptor but rather a label coined by other schools to indicate what they saw as the school’s distinguishing, and most pernicious, tenet: the pudgala. Pudgalavāda has been the term of choice in Buddhist studies for over a century (Lusthaus 2009, 275),3 but I generally prefer to avoid the label – not only because it is polemical and pejorative, but even more because it obfuscates rather than clarifies the historical record. The label exaggerates, from the perspective of philosophical adversaries, the role of the pudgala. I use instead the somewhat awkward term “Vātsīputrīya/ Sāṃmitīya” or, for short and for the later period, simply “Sāṃmitīya.” I use the term to include the affiliated Bhadrayānīya, Dharmottarīya, Ṣaṇṇagarika, and Kaurukulla-Sāṃmitīya schools. Several modern scholars stand out in the study of the pudgalavāda in the English language: Theodor Stcherbatsky (1866–1942), Louis de La Vallée Poussin (1869–1938), André Bareau (1921–1993), Thích Thiện Châu (1931–1998), and Leonard Priestley.4 We might say that specific attention to the notion of the pudgala starts with Stcherbatsky’s provocatively titled “The Soul Theory of the Buddhists,” written in English and published in 1919 (Stcherbatsky 1919a, 823–54; 1919b, 937–58): that is, modern pudgalavāda studies are now a hundred years old. In 1920, La Vallée Poussin published an entry on the “Sammitiyas” in the Encyclopaedia of Religion and Ethics (La Vallée Poussin 1920, 168–69), but he also published many articles and books relevant to the theme before and after this (La Vallée Poussin 1925a, 34–38). Bareau wrote extensively on Buddhist sects (Bareau 1954, 1955a, 1956). By the end of the last century, Vietnamese monk Thích Thiện Châu ([1996] 1999) and Leonard Priestley (1999) produced balanced studies of the concept of the pudgala, demonstrating that the pudgalavādins had something to say and that this was worth listening to. New sources for the study of the Vātsīputrīya/Sāṃmitīyas were identified, and in some cases were edited and translated.

Vātsīputrīya/Sām · mitīya and the Doctrine of the Person

What is the relation between the “person” or “individual” and the “soul” or the “self ” – between pudgala and the ātman? Are they the same or different? Both are everyday terms of speech and reference and both have metaphysical usages. In Jainism, pudgala, material and non-sentient, is one of six fundamental substances. Ātman, in grammar a noun as well as an emphatic and reflexive pronoun, carries significant metaphysical baggage in Hindu schools of thought. The Vātsīputrīya’s term of choice is pudgala, not ātman, but the criticisms of other schools react (or perhaps overreact) to it as a kind of ātman. This is why some authors singled them out as “Buddhist tīrthikas.”5 When we treat the “person” as the signature tenet of a belief system or vāda, we need to consider the relationship between received doctrine and quotidian monastic life. The belief in or assertion of the pudgala – the person or the individual – looms large in the doxographical literature compiled by Buddhist authors from other, non-Vātsīputrīya/Sāṃmitīya schools or traditions, but the word pudgala is noticeably absent in most of the available texts of the school itself. Only two metaphysical treatises of the Sāṃmitīya school survive in their original Indian language – The Great Discourse on the Evolution and the Devolution of the Universe (Mahāsaṃvartanīkathā, Okano 1998) and the Stanzas of the Compendium of Abhidharma (Abhidharmasamuccayakārikā). The two works survive in more or less complete form, but neither of them discusses or even mentions the doctrine of the pudgala. This leads to the question: how did the Buddhist orders manage their identities? The cut and style of monastic costume maintained a physical, visual identity for the different orders, as it does today for the subdivisions within the Theravādin, Tibetan, and East Asian sanghas. Robes

191

Peter Skilling

announce hierarchical and gender identities. But barba non facit philosophum, l’habit ne fait pas le moine: we are not concerned here with matters of hairstyle and dress, but of ideas and philosophy. Was every Vātsīputrīya/Sāṃmitīya monastic expected to study and adhere to the idea of the pudgala? Did the pudgala mean anything to rank and file monastics? Did the pudgala insinuate itself into the monastic codes of the Vinaya, the teaching of the Buddha in the Sūtras, or the daily liturgies and spiritual exercises? What, if anything, did the pudgala mean to lay followers and supporters of the school? In the absence of any Vātsīputrīya/Sāṃmitīya canons, these questions remain open, but on the evidence of currently available texts, I suspect that the answer is that no, the pudgala does not seem to feature in the canonical texts of the school.

Methodological Considerations in the Study of the Vātsīputrīya/Sām  mitīya The existing paradigms for the study of school development are conditioned by the resources available today and the choices that we make as researchers (further dependent on our language skills or lack thereof). Since records of most other schools are no longer extant, the resource base is largely limited to the textual traditions of two schools, the Theravāda and the Sarvāstivāda/Vaibhāṣika. Most modern research has sided with these two schools and presumed that pudgalavāda is an aberration – that it is unorthodox and even non-Buddhist. The South Asian pudgalavāda group of monastic orders died out nearly five centuries ago and today has no voice whatsoever in the forums of philosophy (despite the fact that the debate on the pudgala continues to be part of Tibetan Buddhist curricula). Today, whether in Tibetan monasteries or in modern Buddhism and Buddhist studies, there is no one to speak on behalf of the Vātsīputrīya/Sāṃmitīyas or to argue their case. Nonetheless, we ought not to ignore them. Buddhism is not a monolith, and the Vātsīputrīya/Sāṃmitīyas are one of the several cogent streams of thought that constitute the complex of ideas that we recognize as “Buddhism.” Like other sanghas, the Vātsīputrīya/Sāṃmitīyas trace their lineage directly back to Śākyamuni and the first two recitation/redactions of his teachings. Like other sanghas, the Vātsīputrīya/ Sāṃmitīyas were fully fledged corporate bodies which interacted with state and society, and responded to the continual flux of philosophical thought around them. Like other sanghas, the Vātsīputrīya/Sāṃmitīyas were historical and cultural currents of Indian Buddhism with distinct identities. The Sāṃmitīya self-consciousness is nurtured by a distinctive account of the how their scriptures were redacted (see in what follows). They claim a distinct pedigree, a distinct history, through five recitation convocations. Balanced studies of the historical and philosophical evolution of the Vātsīputrīya/Sāṃmitīyas have been impeded by methodological issues. One is the persistence of Victorian frameworks that were developed at the inception of Buddhist studies. Early European scholars interpreted Indian religions, including Buddhism, on their own terms, importing inappropriate models from contemporary historiographies of the Christian churches and the fledgling discipline of “world religion.” They saw the evolution of Buddhist schools in terms of “orthodoxy” and “heresy,” as a history of schism, dissension, and departure from the original pure and untrammeled teaching of Gautama Buddha. They found this untrammeled teaching in the Theravāda that they encountered in Sri Lanka – or at least in its Pali scriptures, since they did not take long to decide that the latter-day understanding and practice of the religion was degraded and corrupt. The Theravāda was among the Buddhist schools that rejected the Sāṃmitīya theory, and modern Buddhist historiography generally echoes their polemics. The very idea of the person was seen as wrong – and there were no pudgalavādins around to argue for it. 192

The Vātsīputrīya/Sām.mitīya

Geography and Localization of the Vātsīputrīya/Sām · mitīya

From the very start, Buddhist monastics were mobile. Perennially peripatetic, they carried ideas, practices, ceremonies, and rituals with them wherever they went. Theories and doctrines were nurtured in shifting landscapes of exchange and debate. There is no systematic history or geography of the Vātsīputrīya/Sāṃmitīyas – or, indeed, of any other school. We can only attempt to sketch the geography and demography of the Buddhist Vinaya schools by consulting Chinese sources like the fabled accounts of Xuanzang and Yijing, which enable us to imagine the geographical distribution of the schools at the time of their visits: that is, in the seventh century. Later sources in Tibetan are piecemeal but often helpful. Indic literary references are few and far between, but script and inscriptions help to locate the early Buddhist schools. Xuanzang’s reports show that the Sāṃmitīyas were well represented at the major sites associated with the Buddha’s life – Kapilavastu, Varanasi, Sarnath, Sravasti, and Vaisali. At the great monasteries of Nālandā and Vikramaśila, the Sāṃmitīyas were represented side by side with the four main schools of the age. They were an active presence throughout the GangaYamuna basin from Mathura down to Bengal, and Xuanzang reports that the Vātsīputrīya/ Sāṃmitīyas were widespread in the middle and lower Indus valley (present-day Pakistan) and in the coastal regions of Pakistan and western India. Yijiing writes that “in Lāṭa and Sindhu . . . the Sāṃmitīya has a large number of adherents” (Li Rongxi 2000, 11). This idea is perpetuated by the rubric “Saindhava Śrāvaka” encountered in Tibetan sources like Tāranātha’s history describing adherents of the school in north India as “Sindhi,” “residents of Sindh.” It appears that the Vātsīputrīya/Sāṃmitīyas were the predominant school in Sindh up to the Arab invasions of the eighth century. Little remains of early Buddhist or pre-Muslim antiquities of the region, but what does remain suggests that there were numerous stūpas and elegant decorative styles in brick (Cousens [1929] 1975; van Lohuizen 1981, 43–57; Pal 2008). In Gujarat, there are a number of rock-cut monasteries. They are much abraded and lack any inscriptions. Which sanghas inhabited them? We do not know, but from the blanket statements about the Buddhism of the region in the Chinese travel accounts, they could well have been Sāṃmitīyas. At an early period, the Vātsīputrīya/Sāṃmitīyas were present along the western coast in Konkan and the adjacent western Ghats, as well as inland in the Deccan plateau, where at Nasik they enjoyed royal patronage of prestigious projects like monumental caitya halls. Xuanzang associates many of the Sāṃmitīya centers with the visits of Śākyamuni or previous Buddhas and with the celebrated emperor, Aśoka (note, however, that during his journeys across the subcontinent, the redoubtable Chinese pilgrim-scholar reported on many sites that claimed such illustrious associations). Buddhist monuments once dotted the north Indian landscape. Across north India, from Sindh in the west to Bengal in the east, they were interactive nodes in an extensive and extraordinary network of knowledge. By the time the British arrived and began – haphazardly, at least at first – their colonial research in archaeology and history, nothing was left of thousands of monasteries and stūpa sites beyond their foundations, and as economic activities expanded, even these were looted for bricks to build railway lines and new buildings – the nineteenth century was an age of both exciting archaeological discoveries and of thoughtless despoliation. Some, if not many, of these ruins would have once been Vātsīputrīya/Sāṃmitīya institutions, given that Xuanzang’s records mention at least 1,351 monasteries of the school with more than 66,500 monks (he rarely distinguishes between monasteries and nunneries) (Lamotte 1958, tables, 596–601). His figures are approximations  – but even so, in terms of monastic land holdings and population, the Sāṃmitīya were by far the largest school: more than half of the monks of India belonged to the Sāṃmitīya order. The monasteries would have been centers 193

Peter Skilling

of training and education where monks and nuns practiced the Vinaya and studied the scriptures and the Vātsīputrīya/Sāṃmitīya dharma system. The only structures associated with the school to survive are early rock-cut monuments in western India.

Vātsīputrīya/Sām  mitīya Scriptures and Canons The historical geography of the Vātsīputrīya/Sāṃmitīyas – as far as we can reconstruct it – shows a widespread and dynamic monastic order that was well established in the lower Indus and Ganga-Yamuna valleys. Like other schools, the Vātsīputrīya/Sāṃmitīyas must have had rich textual collections centered on Tripiṭakas containing their redactions of the Vinaya, Sūtras, and Abhidharma. The school’s scriptural language was a form of Middle Indo-Aryan, specified by Tibetan sources as Apabhraṃśa (Thích Thiện Châu [1996] 1999, 31–32). It was written in at least the Nagari and the Bhaikṣukī scripts; the latter, apparently unique to the school, is also called Saindhavī (Dimitrov 2020). But the textual products of Vātsīputrīya/Sāṃmitīya thought do not survive. The seventh-century Chinese scholar Yijing, who had first-hand experience in India, describes the Sāṃmitīya Tripiṭaka as 300,000 stanzas in length. Xuanzang carried fifteen texts and treatises of the school back to China, but the manuscripts do not survive (Li Rongxi 1996, 394). Many of the short passages cited in extant Sāṃmitīya works like the Treatise of the Sāṃmitiya School (*Sāṃmitīya-nikāyaśāstra, Sānmídibù lùn, T. 1649) have parallels in the canons of other schools like the Pali canon. While a few fragments of their canonical literature survive, nothing is left of their ritual, liturgical, and mythological literature. The result is that the Vātsīputrīya/Sāṃmitīyas are known to us through the writings of outsiders to the tradition and its textual sources. This includes great masters like Vasubandhu and Bhāviveka, who at least had some direct access to the textual tradition. We modern scholars do not even have access to the school’s texts and write using the reports of opponent schools. We may claim the benefits of historical consciousness, but we have access only to fragmentary records. Reasoned debate or criticism presupposes an understanding of the intellectual milieux of the historical audiences of the Vātsīputrīya/Sāṃmitīyas, but, in the absence of texts or social or literary records, such an understanding is lacking – even impossible – at present. This is hardly satisfactory. Some details of the Vinaya rules for monks and nuns are given in Buddhatrāta’s Exposition of Twenty-two Aspects of the Monastic Rules (*Vinaya-dvāviṃśativyakti, Lü èrshiêr míngliăo lùn, translation by Paramārtha, T. 1461). We do not even know what the canonical Abhidharma of the Vātsīputrīya/Sāṃmitīyas looked like – how many books it had, their titles, whether any of them had introductions to tell whether they were spoken by the Buddha, by disciples, or by others. The Vātsīputrīya/Sāṃmitīyas certainly had a fully developed dharma theory of their own. This is further evidence that their pudgala did not contradict or replace the Buddhist deconstruction of the self; whatever the “indeterminate person” was, it depended on a complex and sophisticated theory of interactive dharmas. We see this in the Stanzas of the Compendium of Abhidharma (Abhidharmasamuccayakārikās) and in Daśabalaśrīmitra’s long excerpts from scholastic texts; if not from their Abhidharma, then from a manual or manuals. The Abhidharma Treatise, Description of the Universe (*Lokaprajñapty-abhidharmaśāstra, Li shi a pi tan lun 立世阿毘曇論, T. 1644), a cosmological work translated into Chinese in 559 by Paramārtha, has been identified as belonging to the Sāṃmitīyas (Mahāsaṃvar tanīkathā, Okano 1998, 22 foll.) This text is one of the sources for the Description of the Universe (Lokapaññatti, Denis 1977), a cosmological compendium transmitted in Pali in Southeast Asia. The Sāṃmitīya affiliation is contested by others. 194

The Vātsīputrīya/Sām.mitīya

Recitation Convocations (Saṃgīti) of the Vātsīputrīya/Sām · mitīyas

Accounts of the Buddhist convocations or recitation/redactions consolidate the lineage and identity of the different schools. They establish a sense of continuity – a direct link to Śākyamuni and the great arhats. Not many accounts survive, and modern scholarship and piety have largely taken the Theravādin account as foundational for all schools.6 Remarkably, two roughly contemporary accounts of the recitation convocations of the Vātsīputrīya/ Sāṃmitīyas are preserved. One is “from the Āgama of the Ārya-Sāṃmitīya-school” (Okano 1998, 397, § 1) in Tibetan translation in Daśabalaśrīmitra’s Examination of the Compounded and Uncompounded; the other is in Sanskrit verse in Sarvarakṣita’s Great Discourse on Evolution and Devolution (Okano 1998, § 4.2.15). The account runs as follows: Long after the final nirvāṇa of the great sage Kāśyapa, in the age of strife, the last age, our teacher, the Tathāgata Śākyamuni arose in the world . . . the Tathāgata, the lamp, after establishing the continuity of the lamp of the holy Dharma, himself passed away. In the second month after the Tathāgata’s final nirvāṇa, . . . five hundred monks free of desire recited and compiled the teaching of the Buddha in the Saptaparṇa cave. And again, one hundred years after the Tathāgata’s final nirvāṇa, seven hundred monks free of desire recited and compiled the Dharma. And again, four hundred years after the Tathāgata’s final nirvāṇa, when the Ascetic’s community had become divided into different groups, each adhering to its own school, Vātsīputra recited and compiled the Dharma of one of these schools. From that time on, that school has been known as the Vātsīputrīyas, the expounders of the Dharma. And again, seven hundred years after the Tathāgata’s final nirvāṇa, a senior monk, the sage *Sammita, recited and compiled the traditions of that school. From that time on, that school has been known as the Sāṃmitīya school. And again, eight hundred years after the Tathāgata’s final nirvāṇa, the senior monks Bhūtika and Buddhamitra recited and compiled the traditions of that school. These are known as “the five recitation convocations of the Dharma of the Sāṃmitīya school.” The Sanskrit Great Discourse and Daśabala’s Tibetan excerpt name four prominent teachers (we might call them the patriarchs) of the school: Vātsīputra, *Saṃmita, and, as a pair, Buddhi and Buddhamitra. We do not know anything about their dates or their lives and intellectual careers, only their names and their association with the third, fourth, and fifth Tripiṭaka recitation/redactions. Their textual legacies do not survive – not a single extant work. The result is that we cannot trace their roles in the philosophical development of pudgalavāda.

Sām  mitīya Authors Not many authored works of the Sāṃmitīyas are preserved. Vasubhadra composed the Sutra of the Three Categories of Phenomena (*Tridharmakhaṇḍaka-sūtra), which is lost in the original Indic version but preserved in two different Chinese translations: the Sì ēhánmù chāojiĕ (T. 1505) and the Sānfădù lùn (T. 1506). Both translations include a commentary by Saṅghasena, the Book on the Three Categories of Phenomena (*Tridharmakhaṇḍaprakaraṇa). The author and translator of the Treatise of the Sāmmitīya School, or, from its colophon, Treatise on the Description of the Basis (*Āśrayaprajñaptiśāstra), are not known. It deals 195

Peter Skilling

with controversial philosophical topics, including the pudgala and the intermediate state, and quotes many Sūtras (Thích Thiện Châu [1996] 1999, 99–117). A short, independent commentary on the ye dharmā stanza entitled Commentary on the stanza of dependent arising (Pratītyasamutpādagāthā-vyākhyāna) is preserved in Tibetan translation. According to the colophon, it was written by Ācārya Vinayavarman of the Sāṃmitīya school. It awaits study and translation to see whether it throws light on any specific doctrines of the school (La Vallée Poussin 1962, 50, § 127.2; Yamaguchi et al. [1977] 1986, 51–52).7 Indic versions of the works of only two authors, Saṃghatrāta and Sarvarakṣita, are preserved. We know nothing about Saṃghatrāta, author of the Stanzas of the Compendium of Abhidharma. The author who stands out is the poet Sarvarakṣita; the colophon of the Mahāsaṃvartanī-kathā describes him as “the master, the excellent venerable one Sarvarakṣita, the great poet.” He is also the author of a Sanskrit grammar and the Prakrit Birth Story of Prince Maṇicūdā (Maṇicūḍā-jātaka), which is composed in verse (Hanisch 2006, 109–61; 2008, 195–342) – a glance at only a few verses immediately confirms that he is a “great poet.” Another Indian author is known only from his works and translations in Chinese. This is Paramārtha (499–569 CE), who is said to have been ordained as a Sāṃmitīya and to have studied at Valabhī, in present-day Gujarat, which in Xuanzang’s time was a flourishing Sāṃmitīya center (Okano 1998, 58–59). Paramārtha brought 240 bundles (about 20,000 fascicles) of Sanskrit books by sea to China and was a prolific translator who rendered important treatises by Asaṅga and Vasubandhu into Chinese. He was a leading proponent of the Vijñānavāda school in China, and through his exegesis, he was an influential philosopher in his own right. This does not preclude a Sāṃmitīya backgound: regardless of their monastic affiliation, monastics were free to follow their philosophical consciences. The association of teachers and philosophers to the Vinaya schools is not straightforward. Monks ordained within the Sāṃmitīya tradition could master and write on non-Sāṃmitīya philosophy. An outstanding example is Vimuktisena, who composed a commentary on the Ornament of Realization (Abhisamayālaṃkāra), a verse summary of the Perfection of Wisdom that tradition ascribes to Maitreya bodhisatva,8 who transmitted it to Asaṅga in the Tuṣita heaven. The colophon of the Sanskrit manuscript and the Tibetan translation describe Vimuktisena as a monk of the Kaurukulla tradtion of the Ārya Sāṃmitīyas. His Sāṃmitīya background did not inhibit him from mastering an abstruse and decidedly non–pudgala-oriented work like the Ornament of Realization.

Toward a Chronology of Vātsīputrīya/Sām  mitīya Ideas The Vātsīputrīya/Sāṃmitīyas did not invent the word pudgala, which is frequent in the early Sūtras. It is embedded in the dialogues and discourses that record the Buddha’s teachings over his forty-five–year career; the literature on the topic is immense and I am unable even to attempt to summarize it here. It is hard to see how terms for “individual” or “person” can be avoided in any language. In standard Buddhist usage, the Buddha is a great man (mahā-puruṣa) and the eight types of realized person are called the “eight individual persons” (Pali, aṭṭha-purisa-puggala), a term that is part of the homage to the three jewels in the hymnology of the recollection of the Buddhist order.9 In some passages, the word pudgala is prominent; the Vātsīputrīya/Sāṃmitīya draw attention to these in the formalized debates in the treatises. The Designation of Persons (Puggalapaññatti), one of the seven books of the Theravādin Abhidhamma, even uses the word puggala in its title. It deals with types or categories of individuals. Whenever the Designation of Persons was compiled, the term puggala was still quite respectable. Similar lists of pudgala 196

The Vātsīputrīya/Sām.mitīya

were drawn up by other schools, although none of them devote a whole work to the categories. The lists and definitions define the grades of attainment and map the paths and goals. The difference seems to be that the Vātsīputrīya/Sāṃmitīyas chose to integrate the pudgala into their dharma system and relate it directly to the aggregates, sense-bases, and elements, and to the processes of rebirth and the realization of nirvāṇa. This was novel, if not radical.

Critiques and Doxographies Some of the great Buddhist metaphysicians incorporated critiques of the concept of the pudgala into their works: for example, Harivarman (c. 250–350), Asaṅga (fourth century), Bhāviveka (c. 500–570), Candrakīrti (c. 600–650), and Śāntarakṣita (eighth century, commented upon by his disciple Kamalaśīla) (Priestley 1999, 48–52; Duerlinger 2009, 80–85). Inventories of Sāṃmitīya tenets are given in Vasumitra’s Treatise of the Wheel of the Different Divisions of the Tenets (Samayabhedacakra) and related doxographies preserved in Tibetan and Chinese, and in Bhāviveka’s Blaze of Reason (Tarkajvālā). These doxographies are prominent in Tibetan scholarship up to the present day, and major Tibetan doxographical works deal with the subject in some detail. That the schools gave so much space to the Vātsīputrīya/ Sāṃmitīyas reflects the historical weight of their ideas. Not much information is available after the travel accounts of the itinerant seventh-century Chinese scholar-monks Xuanzang and Yijing. By the eleventh century, we get miscellaneous mentions in written sources from India and Tibet. A Sanskrit history of the Vajrayoginī cult from Nepal notes that the (future siddha) Maitrīgupta ordained as a Sāṃmitīya at Vikramapura in the first half of the eleventh century and “listened to the Sūtra, Abhidharma, and Vinaya” (Lévi 1931, 423).10 The Tibetan historian Tāranātha refers to a Sāṃmitīya master named Jñānamitra, a śūdra (a man of lowest caste) of Tripura in eastern India, who was ordained at Jagaddala monastery in northern Bengal and was learned in the Vinaya and Abhidharma Piṭakas according to the Sāṃmitīya system (Tāranātha in Templeman 1983, 53). These references suggest that Sāṃmitīya ordination rites could be performed at Vikramapura and Jagaddala monasteries, where instruction in their Tripiṭaka was also available – if not at these monasteries themselves, then elsewhere in northern India. This was the time when Daśabalaśrīmitra, a twelfth-century paṇḍita, compiled his Examination of the Compounded and Uncompounded, a copious compendium of philosophy according to the Sarvāstivāda, Sthāvira, and Sāṃmitīya systems, as well as the Yogācāra and Madhyamaka. The presentation of the Sāṃmitīya doctrines, integral to his work, shows that Daśabalaśrīmitra had direct access to their texts. Sarvarakṣita’s Great Discourse on Evolution and Devolution continued to be copied in the Kathmandu valley. The Patna Dhammapada and the Stanzas of the Compendium of Abhidharma were recovered in Tibet in the first half of the twentieth century by Sankrityayan and Tucci, respectively. In addition to this, an unknown number of palm-leaf manuscripts preserved in Tibet are written in bhaikṣukī script, in what we might call a “Sāṃmitīya Prakrit,” showing that the school’s manuscripts continued to circulate. At a certain point – we cannot say when – ordinands of the school began to study, practice, and contribute to Mahāyāna thought. This may be equally true for members of other schools. Modern historians, influenced by the model of Christian history, have tended to overemphasize exclusivity, division, and sectarianism, schismatic thought that might be alien to the ecology of Buddhist and Indian ideas. An outstanding example is Vimuktisena, previously mentioned. This raises the question of the significance and boundaries of school affiliation. Monks like Vimuktisena seem to have been Sāṃmitīya by ordination only – not by adherence to its philosophical system. 197

Peter Skilling

The last report of Sāṃmitīya activity comes from the 1460s from central Tibet, where a monastic document records the visit of an itinerant monk named Lokottara who hailed from eastern India and was traveling from Kashmir to central Tibet. He stated that he belonged to the Sāṃmitīya school and was a student of Madhyamaka, Pramāṇa, and Vajrayāna (Van der Kuijp 1995).

Recovery: Sām  mitīya Texts in Tibetan and in Indian Languages Daśabalaśrīmitra A significant advance in Vātsīputrīya/Sāṃmitīya studies has been the study of Daśa­ba­ laśrīmitra’s Analysis of the Compounded and the Uncompounded, the only Indian work preserved in Tibetan translation that gives direct access to the thought of the Saṃmitīya school. The Paṇḍita, who most probably wrote in the twelfth century, devotes six complete chapters to their dharma system and gives other doctrines elsewhere in his work. This encyclopedic work is almost entirely lost in the original Sanskrit (but for three folios preserved in Cambridge University Library), but it is preserved in an excellent Tibetan translation (Szántó 2020, 129–45). Each chapter is a direct citation of (a) Sāṃmitīya text(s) chosen by Daśabalaśrīmitra. In this regard, the Sāṃmitīya chapters parallel Daśabalaśrīmitra’s citations of Sthāvira material which are taken directly from a single source, the *Vimuttimagga.

Sāṃmatīya Buddhology Daśabalaśrīmitra provides invaluable information on the concept of the Buddha (Skilling 2006). The Great Discourse on Evolution and Devolution gives a short verse summary of the events of Śākyamuni’s life.

Sāṃmatīya Cosmology Cosmology and cosmogony deal with the shape, structure, and formation and dissolution of the world system in which we humans live, including systems of time and eons. Two texts deal with this subject. One is Sarvarakṣita’s Sanskrit verse, Great Discourse on Evolution and Devolution, and the other is a prose section of the Tibetan Analysis of the Compounded and the Uncompounded. These show that the school had a fully articulated cosmology that – in broad strokes – resembles the cosmology of other schools.

Sāṃmatīya Scholasticism The Vātsīputrīya/Sāṃmitīyas developed a unique interpretation of the stages of the path (mārga) to release from suffering. The different schools disagreed on the number of karmic destinies or realms (gati). Some maintained that there are five, others that there are six. A major point of disagreement was whether or not there is an intermediate state (antarābhava) between death and rebirth. Daśabalaśrīmitra confirms other available texts that the Sāṃmitīyas accepted six destinies and the intermediate state. Most of the material cited by Daśabalaśrīmitra is abhidharmic in subject matter and style. Since the school’s Abhidharmapiṭaka does not survive, we cannot trace their source. The Stanzas of the Compendium of Abhidharma is an Abhidharma treatise composed in concise and tightly packed stanzas (kārikās); it resembles and at times even parallels Vasubandhu’s Treasury of Abhidharma (Abhidharmakośa). Because the verses are compact and laconic, 198

The Vātsīputrīya/Sām.mitīya

they are difficult to understand without a commentary. Some of the categories outlined in Daśabalaśrīmitra’s citations and the verses of the Stanzas of the Compendium of Abhidharma are the same as those of the Sarvāstivādin Abhidharma. This suggests that the dharma theories of the two schools developed in tandem and that, despite developing into distinct schools, they share core categories and doctrines. A text recently recognized as a Sāṃmatīya work is a “preacher’s manual,” dating perhaps to the fifth century. Composed in Sanskrit verse and prose, grounded in canonical citations in “Sāṃmitīya Prakrit,” this unique text deals with many of the practical undertakings of Buddhist practitioners in early South Asia (Szántó 2019).

Were the Sām  mitīyas and Vātsīputrīyas Buddhists? One question that has rankled Buddhist intellectuals is this: were the Vātsīputrīya/Sāṃmitīyas Buddhists? (La Vallée Poussin 1925b, 227–28). The doubts began in India, from, most probably, the very beginnings of the great pudgala debate. The question continued to be discussed in Tibet, and it is still debated in modern Buddhist studies. The question depends on context and the definition of “Buddhist”; there is no easy answer and no single answer. Buddhism had several canons, not a single canon. It had several Tripiṭakas – perhaps eighteen or more – not a single Tripiṭaka. Tripiṭakas were dynamic repositories of ideas, not static and passive collections; even when they were fixed and the texts did not change, the uses and interpretations did. Buddhism did not have a single track or path, but developed several  – even multiple – systems of practice and thought. One of these systems was the Vātsīputrīya/ Sāṃmitīya, who transmitted their own canon. A canon is a collection and accumulation of the ideas and practices of a school, transmitted orally for centuries before being written down; each school needed such a corpus for the study and practice of Śākyamuni’s Dharma. Debate on the canonicity of texts continued among the reciters and was never fully resolved. For example, the Miscellaneous Collection (Khuddakanikāya) of the Theravādin canon contains – in addition to early works shared with the other schools – a number of oddities, commentaries, Abhidhamma, and advanced buddhologies that were and became canonical. But they were canonical for the Theravādins alone: other schools would not consider them canonical or even know them. The members of all eighteen Buddhist schools regarded Śākyamuni as their teacher and they were all deeply and sincerely devoted to his person and teaching. All of the eighteen schools traced their scriptural collections, their Piṭakas and Tripiṭakas, back to the Buddha himself through the recitation-redaction of the scriptures held at Rājagṛha after his decease. All of them preserved and nurtured Sūtra and Vinaya collections, and developed Abhidharma systems. Yaśomitra comments on whether or not the Sāṃmitīyas’ notions of the pudgala disqualify them as Buddhists, saying, “It is not that there is no liberation for the Vātsīputrīyas: there is, because they are Buddhists” (Wogihara 1932–1936, 699.4). In his commentary on the Introduction to the Bodhisatva Path (Bodhisatvacaryāvatāra), however, Prajñākaramati refers to the pudgalavādins as “tīrthikas within the Buddhist fold” (La Vallée Poussin 1898, 297.8). In his General Presentation of the Classes of Tantra, Mkhas grub rje (1385–1438), a close disciple of Tsongkhapa, presents another angle: “pudgalavādins are not Buddhist because they believe in the pudgala, but they are Buddhist because they take refuge and practice ethics, etc.” Some say that the question rests with whether or not the schools upholds the “four seals”: 1. 2. 3. 4.

The saṃskāras are all impermanent. Everything with flux (sāsrava) is suffering. All natures are devoid of self. Nirvāṇa is tranquil and solitary. 199

Peter Skilling

Some say they do, and some say they do not. Once again, there are no Sāṃmitīyas to speak up and clarify their position.

New Perspectives: Fragments From a Sām  mitīya World Inscriptions on Sāṃmitīya statues or their pedestals bear a canonical citation that appears to be a Prakrit version of the summary of the Buddha’s teaching given by the monk Aśvajit to Śāriputra – a celebrated event that is recounted in the scriptures of all known schools, in Sanskrit, Pali, hybrid Sanskrit, and in a variety of translations. The Fortunate One, dear sir, teaches the nature of dependent arising and the cessation of dependently arisen dhammas: The Tathāgata teaches the cause Of the dhammas that arise from causes As well as their extinction. Such is the teaching of the Great Ascetic. The ye dharmā verse or stanza of dependent arising was, and is, widely disseminated in liturgies, inscriptions, and manuscripts across the Buddhist world. That the Sāṃmitīyas chose to inscribe this particular text shows that like the other schools they regarded, or revered, conditionality as the very heart or core of the Buddha’s teaching. The school’s public face is dependent arising and the stanza of causation – not the pudgala. The Vātsīputrīya/Sāṃmitīya was not a marginal group: it was a fully fledged Buddhist school. Priestley notes that “this was not a small sect on the fringes of Buddhism, but rather an integral part of the Indian Buddhist tradition” (Priestley 1999, vii). It was one among equals. We now have access to some of their works in Indian languages and translation, and further studies will allow us to appreciate more clearly their role in Buddhist thought and culture. These fragments from a Sāṃmitīya world give a new perspective on the school, suggesting that it was a vital and creative presence in north India up till the final age, the decline and disappearance of Buddhism. There seems to be a particular nexus of activity in the eleventh–twelfth centuries in north India: Sarvarakṣita’s literary productions, the copying of Bhaikṣukī manuscripts, the inscribing of images, and the ordinations as Sāṃmitīya monks, for example. This activity, and the accessibility of Sāṃmitīya materials, is reflected in Daśabalaśrīmitra’s decision to give significant space to Sāṃmitīya doctrines. Marginality and centrality are subjective categories, dependent on the writer’s choice of a center. I am ready to say that throughout the history of Indian Buddhism, the Vātsīputrīya/Sāṃmitīyas were anything but marginal, either as a monastic order or as an intellectual current. They were a significant presence in the material and metaphysical landscape of northern and western India from the beginnings of post-Nirvāṇa Buddhist philosophy, over a period of 1,500 years. It is time to recognize this in Buddhist studies. In a work published in 1999, Leonard Priestley wrote: It is hardly necessary to point out the importance, both philosophically and historically, of a form of Buddhism which differs strikingly in its interpretation of the Buddha’s teaching from what we have come to regard as orthodox, and yet was for some time, at least, the dominant form of Śrāvakayāna Buddhism in India. But the difficulties facing us in investigating the Pudgalavāda are considerable.

200

The Vātsīputrīya/Sām.mitīya

Orthodoxy and unorthodoxy are troublesome notions, and what role, if any, they have in the study of a fluid and dynamic Indian Buddhism – or, if we like, Indian Buddhisms – that evolved for more than 2,000 years is open to question. In this short chapter, I  have attempted to point out the significance of the Vātsīputrīya/ Sāṃmitīya tradition. Many difficulties still face us in the study of this school of practice and thought, but the new materials outlined in this chapter should advance our understanding considerably. The title of an earlier work referred to “rehabilitating the pudgalavādins.” By this, I meant bringing them in from the cold by showing that they were one of the Buddhist schools that significantly contributed to the development of Buddhism. I hope to encourage the non-sectarian study of Buddhist monastic culture in the broadest sense, embracing literature, ritual, art, and archaeology, as far as resources allow. By non-sectarian, I mean without taking any tradition as original or authentic, but rather examining Buddhist intellectual culture with open minds.

Notes 1 I am grateful to Eng Jin Ooi for help with the final stages of this chapter. 2 I use the word “canonical” to mean belonging to the speech of the Buddha as preserved in the scriptural collections of the different Buddhist schools without prejudice for any particular school. 3 See the remarks on “the term pudgalavāda, which scholars continue to use,” in Lusthaus (2009, 275), and his n. 1, “I leave pudgalavāda in the lower case . . . in order to indicate that it is not the proper name of a school or sect but a label attached to the Vātsīputrīya, Sāṃmitīya, etc. schools by their opponents.” With this politics of capitalization I concur. 4 I do not refer here to an extensive literature in Japanese, which includes translations of some of the key works from Chinese and Tibetan. 5 A tīrthika, “one who leads or guides to and across the ford to liberation,” is a general Buddhist term for other spiritual traditions. It is often inappropriately translated as “heretic.” 6 Exceptions are the work of early generations of European scholars such as Schiefner, Wassilief, Przyluski (1927), and La Vallée Poussin, and those of Bareau mentioned previously, as well as his Les premiers conciles bouddhiques (Bareau 1955b). 7 It is found only in the London Dunhuang manuscript collection; present-day Tanjurs – Tibetan canons of Buddhist scriptures – do not include the work, or any others by any author of this name. 8 In this essay, bodhisatva is spelled with a single “t” as it is spelled in the overwhelming majority of manuscripts and inscriptions, as also in loan words in other languages like Tibetan or Thai. Bodhisattva with a double “tt” is a modern normalization. 9 Arthaviniścayadharmaparyāya: text Samtani (1971, 47–48; trans. Samtani 2002, 33). The Commentary (Nibandhana) gives an etymological explanation (Samtani 1971, text 253.2; 2002 trans. 182). 10 For the date, see Tatz (1987, 695–711). I suppose that Vikramapura = Vikramaśilā.

Bibliography Bareau, André. 1954. “Trois traités sur les sectes bouddhiques attribuès à Vasumitra, Bhavya et Vinītadeva 1.” Journal Asiatique 242: 228–66. ———. 1955a. Les sectes bouddhiques du Petit Véhicule. Saigon: Ecole française d’Extrême-Orient. ———. 1955b. Les premiers conciles bouddhiques. Paris: Presses Universitaires de France. ———. 1956. “Trois traités sur les sectes bouddhiques attribuès à Vasumitra, Bhavya et Vinītadeva 2.” Journal Asiatique 244: 167–200. Cousens, Henry. (1929) 1975. The Antiquities of Sind with Historical Outline. Archaeological Survey of India, vol. XLVI, Imperial Series. Calcutta, reprt. Varanasi: Bhartiya Publishing House. Denis, Eugène. 1977. La Lokapaññatti et les idées cosmologiques du bouddhisme ancien, 3 vols. Lille: Université de Lille. Dimitrov, Dragomir. 2020. The Buddhist Indus Script and Scriptures: On the So-called Bhaikṣukī or Saindhavī script of the Sāṃmmitīyas and their Canon. Wiesbaden: Harrassowitz Verlag (Veröffentlichungen der Fächergruppenkommission für Außereuropäische Sprachen und Kulturen: Studien zur Indologie).

201

Peter Skilling Duerlinger, James. 2009. “The Theory of Persons of the Pudgalavādins.” The Indian International Journal of Buddhist Studies 10: 75–90. Hanisch, Albrecht. 2006. “Progress in Deciphering the So-called ‘Arrow-head’ Script Allowing Access to Sarvarakṣita’s Maṇicūḍajātaka, a Text of the Buddhist Sāṃmitīya School.” Journal of the Centre for Buddhist Studies Sri Lanka 4: 109–61. ———. 2008. “Sarvarakṣita’s Maṇicūḍajātaka. Reproduction of the Codex Unicus with Diplomatic Transcript and Palaeographic Introduction to the Bhikṣukī Script.” In Sanskrit Texts from Giuseppe Tucci’s Collection, edited by Francesco Sferra. Manuscripta Buddhica, 1; Serie orientale Roma. Rome: Istituto italiano per l’Africa e l’Oriente. La Vallée Poussin, Louis de. 1898. Bouddhisme, études et matériaux: Ādikarmapradīpa, Bodhicaryāvatāraṭīkā. London: Luzac & Co. ———, ed. 1912. Madhyamakāvatāra par Candrakīrti – Trad. tibétaine. St.-Pétersbourg: Imprimerie de l’Académie Impériale des sciences. ———. 1920. “Sammitiyas.” In Encyclopaedia of Religion and Ethics XI, edited by James Hastings, 168–69. Edinburgh: T. & T. Clark; New York: Charles Scribner’s Sons. ———. 1925a. Nirvâṇa (Études sur l’Histoire des Religions 5). Paris: Gabriel Beauchesne. ———. 1925b [1971]. L’Abhidharmakośa de Vasubandhu V. Repr. Brussels: Institut Belge des Hautes Études Chinoises. ———. 1962. Catalogue of the Tibetan Manuscripts from Tun-huang in the India Office Library. London: Oxford University Press for the Commonwealth Relations Office. Lamotte, Étienne. 1944. Le traité de la grande vertu de sagesse de Nāgārjuna (Mahāprajñāpāramitāśāstra), vol. I. Louvain: Bureaux du Muséon. ———. 1958. Histoire du bouddhisme indien: Des origines à I’ère Śaka. Louvain: Institut Orientaliste. Lévi, Sylvain. 1931. “Un nouveau document sur le bouddhisme de basse époque dans l’Inde.” Bulletin of the School of Oriental Studies, University of London 6 (2): 417–29. Li Rongxi, trans. 1996. The Great Tang Dynasty Record of the Western Regions. Berkeley: Numata Center for Buddhist Translation and Research. ———, trans. 2000. Buddhist Monastic Traditions of Southern Asia: A Record of the Inner Law Sent Home from South Seas. Berkeley: Numata Center for Buddhist Translation and Research. Lusthaus, Dan. 2009. “Pudgalavāda Doctrines of the Person.” In Buddhist Philosophy: Essential Readings, edited by William Edelglass and Jay L. Garfield, 275–85. New York: Oxford University Press. Okano, Kiyoshi. 1998. Sarvarakṣitas Mahāsaṃvartanīkathā: ein Sanskrit-Kāvya der Sāṃmitīya-Schule des Hīnayāna-Buddhismus. Sendai: Seminar of Indology, Tohoku University. Pal, Pratapaditya, ed. 2008. Sindh: Past Glory, Present Nostalgia. Mumbai: The Marg Foundation. Priestley, Leonard. 1999. Pudgalavāda Buddhisṃ: The Reality of the Indeterminate Self. South Asian Studies Papers, no. 12, Monograph no. 1. Toronto: University of Toronto. ———. n.d. “Pudgalavada Buddhist Philosophy.” Internet Encylopedia of Philosophy. Accessed June 25, 2020. www.iep.utm.edu/pudgalav/. Przyluski, Jean. 1927. Le concile de Rājagṛiha: introduction à l’histoire des canons et des sectes bouddhiques. Paris: Paul Geuthner. Ronkin, Noa. 2005. Early Buddhist Metaphysics: The Making of a Philosophical Tradition. London: Routledge. Samtani, Narayan Hemandas, ed. 1971. The Arthaviniścaya-Sūtra  & Its Commentary (Nibandhana) (Written by Bhikṣu Vīryaśridatta of Śri-Nālandāvihāra), Critically Edited and Annotated for the First Time with Introduction and Several Indices. Patna: K. P. Jayaswal Research Institute. ———, trans. 2002. Gathering the Meanings: The Compendium of Categories: The Arthaviniścaya Sūtra and Its Commentary Nibandhana. Berkeley, CA: Dharma Publishing. Skilling, Peter. 2006. “Daśabalaśrīmitra on the Buddhology of the Sāṃmitīyas.” In Nagoya Studies in Indian Culture and Buddhism – Sambhasa 25, 98–123. Nagoya: Department of Indian Studies, Graduate School of Letters, Nagoya University. Stcherbatsky, Theodar. 1919a. “The Soul Theory of the Buddhists.” Bulletin de l’Académie des Sciences de Russie VI série 13 (12–15): 823–54. ———. 1919b. “The Soul Theory of the Buddhists.” Bulletin de l’Académie des Sciences de Russie VI série 13 (16–18): 937–58. Szántó, Péter-Dániel. 2019. “The *Saddharmaparikathā, a Buddhist Preachers’ Manual from Late Antiquity.” Presented at the University of Vienna, 22 March; Unpublished Draft Courtesy the Author.

202

The Vātsīputrīya/Sām.mitīya ———. 2020. “A Sanskrit Fragment of Daśabalaśrīmitra’s Saṃskṛtāsaṃskṛtaviniścaya (Ch. 29 & 30).” Annual Report of the International Research Institute for Advanced Buddhology 23: 129–45. Tatz, Mark. 1987. “The Life of the Siddha-Philosopher Maitrīgupta.” Journal of the American Oriental Society 107 (4): 695–711. Taylor, C. Arnold, ed. 1894. Kathāvatthu. London: The Pali Text Society. Templeman, David, ed. and trans. 1983. The Seven Instruction Lineages (Jo Nang Tāranātha). Dharmasala: Library of Tibetan Works & Achieves. Thích Thiện Châu, Bhikshu. (1996) 1999. The Literature of the Personalists of Early Buddhism. Translated by Sara Boin-Webb. New Delhi: Motilal Banarsidass Publishers. Trenckner, V., ed. 1880. The Milindapañho: Being Dialogues Between King Milinda and the Buddhist Sage Nāgasena. London: Williams and Norgate. Van der Kuijp, Leonard W. J. 1995. “Some Indian and Sri Lankan Peregrinators in Central Tibet and Glo Bo Smon Thang During the Fifteenth Century.” Presented at the 7th Seminar of the International Association for Tibetan Studies, Graz. van Lohuizen, J. E. 1981. “The Pre-Muslim Antiquities of Sind.” In Sind Through the Centuries: Proceedings of an International Seminar Held in Karachi in Spring 1975 by the Department of Culture, Government of Sind, edited by Hamida Khuhro, 43–57. Karachi: Oxford University Press. Wogihara, Unrai, ed. 1932–1936. Sphuṭārthā Abhidharmakośavyākhyā by Yaśomitra, Part II. Tokyo: The Publishing Association of Abhidharmakośavyākhyā. Yamaguchi, Zuihō et  al. (1977) 1986. スタイン蒐集チベット語文献解題目錄 (A Catalogue of the Tibetan Manuscripts Collected by Sir Aurel Stein), Part Two. Tokyo: Toyo Bunko.

203

PART 4

Philosophical Founders

PHILOSOPHICAL FOUNDERS Introduction to Part 4

The great flowering of Buddhist philosophy that takes place in India has no founder. Even the Buddha himself is said to be just the most recent of those who have turned the wheel of the Dharma for the sake of the world; according to Buddhist traditions, other buddhas preceded him. But conventionally, it is nevertheless possible to identify developments that can, in retrospect, be recognized as new modalities in Indian Buddhist philosophical life. Of course, such beginnings are happening constantly, and this volume aims to show how a great number of individual philosophers leave their marks on how Buddhist philosophy gets done. Still, as in the West, the philosophical work of some figures affects subsequent thinkers in the tradition; in India, such figures have traditionally been identified as constituting new lineages of thought and practice. We identify the thinkers in this part as “philosophical founders,” taking our inspiration from the popular Tibetan schema of the “six ornaments and two supreme ones” (rgyan drug mchog gnyis). This group of eight luminaries, a popular subject for monastic scroll paintings, includes the six philosophers treated in this part – Nāgārjuna, Āryadeva, Asaṅga, Vasubandhu, Dignāga, and Dharmakīrti – along with two masters of monastic discipline (vinaya), Guṇaprabha and Śākyaprabha. The six founders broadly also represent three significant streams in Mahāyāna philosophical thought, that of the Madhyamaka, Yogācāra, and Pramāṇa traditions, with two masters associated with each stream. We begin with Nāgārjuna, the founder of Madhy­amaka, along with his chief immediate disciple, Āryadeva. Following this, we have Asaṅga, the founder (with Maitreya) of the Yogācāra, along with his chief disciple and brother, Vasubandhu. And we finish with Dignāga, the founder of the Buddhist Pramāṇa stream, along with his most influential disciple, Dharmakīrti. Each of these six figures is treated in this part by authors who take distinct approaches. As they are all iconic for so many of the Buddhist thinkers who follow them, readers will find their names referenced repeatedly throughout the rest of the book. Nāgārjuna (second–third century CE) is widely regarded as one of the most influential philosophers in any Buddhist tradition. His Root Verses of the Middle Way (Mūlamadhyamakakārikā), along with the other analytic treatises and devotional hymns that articulate his account of dependent origination and emptiness, became the basis of the Madhyamaka tradition of Mahāyāna Buddhism. Nāgārjuna’s notoriously terse verses led to a great diversity of interpretations by Buddhist intellectuals in India and in Central and East Asia, as well as among contemporary scholars. In her chapter, Amber D. Carpenter engages the most prominent interpretations of DOI: 10.4324/9781351030908-21

207

Philosophical Founders

the Root Verses – as metaphysical, semantic, epistemological, and anti-rationalist – and offers a reading that emphasizes Nāgārjuna’s return to the Buddha’s teachings on dependent arising ( pratītyasamutpāda). Thus, according to Carpenter, Nāgārjuna is a fundamentalist reformer, responding to the intellectual world of second-century Indian Buddhism by returning to what he believes is the true teaching of dependent arising and its rightful centrality in Buddhist thought. In contrast to the view of dependent arising as the way in which things arise or are in relation with each other, Nāgārjuna’s understanding of dependent arising functions to deconstruct metaphysical conceptions of individuation as such; according to Carpenter, for Nāgārjuna, dependent arising means there are no things arising. By linking the original teaching of dependent arising with the Mahāyāna understanding of emptiness and the distinction between ultimate and conventional truth, Nāgārjuna defends the Mahāyāna as the authentic embodiment of the Buddha’s teachings. Carpenter sees in Nāgārjuna’s Precious Garland (Ratnāvalī) the implications for ethics of Nāgārjuna’s philosophy of dependent arising, emptiness, and the two truths. Reading Nāgārjuna as embedded in communities of practice, she suggests that his ontology, epistemology, and philosophy of language do indeed inform the Buddhist practical-soteriological project. Āryadeva (second–third century) is traditionally said to have been Nāgārjuna’s student and is credited as a cofounder of the Madhyamaka. While many Indian philosophers argue that possessing an essence or intrinsic nature is a condition for something to be real, Āryadeva believes that there is no entity that meets this condition. Like Nāgārjuna, he employs a dialectical method to deconstruct any claim of intrinsic nature; he argues that there is no essence or identity that is somehow independent of causes and conditions. Phenomena are then precisely empty of the intrinsic natures that are attributed to them. Thus, we must abandon any philosophical views that presuppose real entities. Focusing on Āryadeva’s most significant work, the Four Hundred Verses (Catuḥśataka), Tom J. F. Tillemans gives an account of both Āryadeva’s anti-realist metaphysics and ontology and his ethics, as well as how they may – or may not – relate to each other. Āryadeva’s ethics are, as Tillemans argues, radically revisionist. That is, Āryadeva proposes a supererogatory morality that calls for a kind of moral sainthood. Tillemans suggests that there is a tension between Āryadeva’s descriptive ontology and revisionary ethics, as the former undermines any possible philosophical foundation for the later. Tillemans also explores Dharmapāla’s Yogācāra interpretation of the Four Hundred Verses and what it might tell us about early Yogācāra-Madhyamaka syntheses. The writings attributed to Asaṅga (fourth–fifth century) are foundational to the development of the Yogācāra tradition of Mahāyāna Buddhism. Indeed, Asaṅga’s texts are some of the earliest systematic presentations of core Yogācāra ideas. Jowita Kramer, in her chapter, focuses on Asaṅga’s two most important texts: the Compendium of the Abhidharma (Abhi­ dharmasamuccaya) and the Summary of the Mahāyāna (Mahāyānasaṃgraha). The Compendium of the Abhidharma is a systematic work of what some scholars call Mainstream (i.e., not Mahāyāna) Abhidharma that integrates characteristically Yogācāra ideas. In the Summary of the Mahāyāna, Asaṅga articulates the core ideas that distinguish Yogācāra from other Buddhist schools. These include the idealist ontology of the three natures, and that there is nothing external to the mind – that there is nothing other than presentation to the mind (vijñaptimātra). He also presents the doctrine of the eight forms of mind, including the storehouse mind (ālayavijñāna), a form of mind beneath the other forms that is the store of karma. Kramer suggests that while it may not be the case that Asaṅga himself first articulated these ideas, he was likely the first author to develop them into one coherent philosophical system. Perhaps no Indian Buddhist philosopher justifies a wariness about categorizing authors according to schools as much as Vasubandhu (fourth–fifth century), traditionally said to be Asaṅga’s 208

Introduction to Part 4

younger brother. Vasubandhu is widely regarded as one of the founding fathers of the Yogācāra school. Yet he wrote a comprehensive account of the Sarvāstivāda Abhidharma, the Treasury of Abhidharma Verses (Abhidharmakośakārikā). At the same time, in his own detailed and very influential Commentary on the Treasury of Abhidharma (Abhidharmakośabhāṣya) – the touchstone reference for Abhidharma in Tibetan scholasticism – he subjects the root verses to critique from multiple perspectives, especially from a view associated with the Sautrāntika school. Some of Vasubandhu’s other texts defending an idealist ontology and associated Yogācāra ideas were also highly influential. Still, as Jonathan C. Gold notes, while Yogācāra and Madhyamaka came to be regarded as rival schools, Vasubandhu never actually uses the term “Yogācāra,” and he does employ characteristically Nāgārjunian arguments. According to Gold, then, Vasubandhu should be read not as sequentially representing one school after another, but rather as a philosopher who works within multiple inherited traditions. Focusing on questions of causality and language, Gold traces a common theme across the diversity of Vasubandhu’s texts. In particular, Gold unpacks Vasubandhu’s view that while concepts and words cannot actually capture reality, even false concepts, when used skillfully, can point to an ineffable reality beyond language. Dignāga (late fifth–early sixth century) is credited with founding the influential Buddhist tradition of logic and epistemology known as Pramāṇa. Of course, Dignāga inherited and was informed by a rich tradition of philosophical inquiry into perception, cognition, and reasoning. He was especially influenced by Vasubandhu’s writings on argumentation. Still, Dignāga’s work, culminating in the Compendium of the Means of Valid Cognition (Pramāṇasamuccaya) together with his autocommentary, constitutes a significant development in Indian logic and epistemology. In contrast to some other Indian philosophers who accept verbal testimony (śabda) as a distinct valid means of cognition ( pramāṇa), Dignāga understands it as a form of inference (anumāna), one of only two valid means of cognition, along with perception ( pratyakṣa). He subdivides inference into inference-for-oneself and inference-for-others, which makes it possible to treat argumentation within the broader epistemological category of inference. As Kei Kataoka notes, this was part of Dignāga’s project of bringing epistemology (theories of the means of valid cognition), dialectics (theories of argumentation), and semantics (theories of word-referents) into one unified field of study. Relevant to all three domains, and one of Dignāga’s most original and important contributions, is the idea of pervasion (vyāpti), a form of invariable concomitance that can guarantee inferences. In addition to giving an overview of Dignāga’s project in the Compendium, Kataoka focuses on some of Dignāga’s most crucial concepts, including the exclusion of what is other (anyāpoha), the three conditions of a proper reason (trairūpyaliṅga), the self-awareness of cognitions (svasaṃvitti), and nonperception (anupalabdhi). Kataoka carefully traces what Dignāga drew from his predecessors, what constitutes his original contributions, and where his successors – the Mīmāṃsaka Kumārila and the Buddhist Dharmakīrti – recognized the limitations of his system. Developing the work of his predecessors, especially Vasubandhu and Dignāga, Dharmakīrti (ca. sixth–seventh century) is regarded as a cofounder of the Buddhist school of logic and epistemology, and was for centuries seen as the main representative of Buddhist philosophy for non-Buddhist philosophers in India. His most important work, the Commentary on the Means of Knowledge (Pramāṇavārttika), explicates and refines the Compendium of the Means of Knowledge (Pramāṇasamuccaya), synthesizing Dignāga’s epistemology, dialectics, and semantics with traditional Buddhist doctrines to give them a whole new significance. While Dharmakīrti is generally associated with the fields of logic and epistemology, as John Taber points out, he is primarily interested in defending the rationality of core Buddhist doctrines as he understands them. Thus, in addition to addressing his original contributions to theories of perception, inference, universals, and exclusions, Taber devotes much of his chapter to 209

Philosophical Founders

Dharmakīrti’s apologetic concerns. These include Dharmakīrti’s defense of the Buddha as tantamount to a means of knowledge ( pramāṇa) and the authority of Buddhist scripture, his critique of physicalism and defense of antirealism and momentariness, and his disagreements with Brahmanical philosophers. Throughout, Taber makes clear, Dharmakīrti is primarily interested in logic and epistemology not for their own sake, but insofar as they can provide a systematic and rational defense of the Buddhist soteriological project.

210

13 NĀGĀRJUNA Dependent Arising Without Any Thing Arising1 Amber D. Carpenter

With Nāgārjuna (c. second–third century CE), we come to the first named philosopher in the Buddhist tradition, and certainly one of the most significant. As with Plato, half the fun of Nāgārjuna (when it does not lead to blood on the carpet) is that there are nearly as many Nāgārjunas as there are serious students of his work – with the added layer of complication that it is even more radically contested what counts as Nāgārjuna’s work. Yet, so little is securely known about Nāgārjuna the person that the best way to anchor a referent for the name is via his works, traditionally including ‘analytical’ treatises, several practical works (for instance, the Letter to a Good Friend [Suhṛllekha]), and a number of hymns. While scholars have disputed whether the same person is indeed responsible for all these works, for philosophical purposes, it is not necessary to adjudicate the matter. For within philosophical discourse, the analytic text Root Verses of the Middle Way (Mūlamadhyamakakārikā; abbreviated as MMK) may be taken as criterial: Nāgārjuna is, by definition, “the author of the Root Verses of the Middle Way,” the founding text of what became over time a distinctive understanding of Mahāyāna Buddhist thought known as Madhyamaka.2 About Nāgārjuna-the-author-of-the-Root Verses of the Middle Way as a person, we know very little with certainty. It is said that he was born into a Brahmin family and converted to Buddhism, and possibly even worked as advisor at the Sātavāhana court on the Deccan Plateau.3 He wrote in Sanskrit, not in one of the various vernaculars as was common practice among Buddhists at the time, thus situating his thought within the wider pan-Indic intellectual discourse, both Buddhist and non-Buddhist. It is broadly accepted that Nāgārjuna so defined also wrote the Dispeller of Disputes (Vigrahavyāvartanī)4 and other treatises grouped together in Tibetan doxography as Nāgārjuna’s ‘analytic’ works, namely: the Seventy Verses on Emptiness (Śūnyatāsaptati), with accompanying commentary by Nāgārjuna and later commentaries by Candrakīrti and Parahitabhadra;5 the Sixty Verses on Reasoning (Yuktiṣāṣṭika), widely-cited in Madhyamaka debates (quotes appear in, for instance, Bhāviveka, Śāntarakṣita, and Candrakīrti, who also wrote a commentary on the whole);6 Crushing the Categories (Vaidalyaprakaraṇa), which critiques non-Buddhist metaphysical categories;7 and, possibly in this group, and at any rate mostly agreed among scholars to be by the author of the Root Verses, the Precious Garland (Ratnāvalī; abbreviated as RĀ), a letter of advice to a prince containing within it a healthy measure of articulation of Nāgārjuna’s Madhyamaka.8 DOI: 10.4324/9781351030908-22

211

Amber D. Carpenter

Naturally, disagreements continue.9 But in some cases, at least, settling these disagreements may not be vital for philosophical purposes: if the ideas in the Letter to a Good Friend, for instance, are so generic that any Mahāyāna Buddhist might have composed them, then we may allow the attribution to stand,10 so long as any interpretation of it is guided by an understanding of the Root Verses and not the other way round. The question acquires philosophical interest only if taken as evidence for some broader claim – for instance, that the only ethical implications of Madhyamaka are generically Mahāyāna. Similarly, regarding the hymns attributed to Nāgārjuna but possibly retrospectively ascribed to him: the Buddhist tradition of associating these with Nāgārjuna is long and deep, and from the point of view of what the figure of Nāgārjuna stands for in the development of Mahāyāna Buddhism, this association is significant. For philosophical purposes, however, the Root Verses will be used to interpret the hymns, and not the other way round.11 There is, of course, a great deal this leaves open. In particular, one might wonder whether the largely metaphysical12 – or semantic,13 or epistemological,14 or anti-rationalist15 – Root Verses have any bearing at all on ethical conduct. Here the beloved but disputed text, the Precious Garland, comes to the fore. If this text is by Nāgārjuna, in our sense of that name, then it could give, first of all, some confirmation that Nāgārjuna took the views articulated in the Root Verses to be consistent with continued engagement with conventional reality and with hortatory reform within that discourse; and second, it would give some indication of what such a Madhyamaka engagement might look like – or it might alternatively confirm that Madhyamaka differs from other Buddhist views only in its conception of the ultimate goal, without that bearing on everyday conduct. The Buddhism Nāgārjuna adopted upon entering the Buddhist fold was of the Mahāyāna sort – although precisely what Mahāyāna meant, apart from the growing clarity that it was different from the other Buddhist views and practices dominant at the time, was still under construction. Indeed, Nāgārjuna’s Root Verses constitutes an important move in the formation and definition of the Mahāyāna, as it gradually acquired that distinctiveness from other Buddhist views which retrospectively appears so categorical. In the early part of the first millennium of the Common Era, a subset of views within the eclectic Abhidharma tradition became increasingly correlated and distinguished from all other Abhidharma strands of thought. Usually associated with the Mahāsāṃghikas, these ideas included the supermundane nature of the Buddhas, and bodhisattvas who make vows to remain in saṃsāra for the sake of helping suffering beings. This subset of associated views began to come together with the new injection of ideas – such as Buddha nature (tathāgatagarbha) and an emphasis on emptiness – from the Perfection of Wisdom Sutras (Prajñāpāramitā Sūtras), texts ‘discovered’ from perhaps the turn of the millennium and recognised by some – but not all – Buddhists as authentic buddha-vacana, words of the Buddha.16 While Mahāyāna was, by Nāgārjuna’s time, a distinct label for those working out an interpretation of the Buddhadharma centred on these notions, such an affiliation was not yet so distinct as to warrant distinct communities of practice, and Mahāyānists of Nāgārjuna’s time and place would have been practicing in mixed monasteries alongside other assorted Abhidharma Buddhists.17 It is within this dialectical context that Nāgārjuna offers his interpretation of the Buddhadharma. He is a Mahāyānist, and in that, he can be expected to be critical of Abhidharma views where they differ from the Mahāyāna. At the same time, his aim is to legitimate the Mahāyāna, and his strategy for doing so will be two-fold: first, to offer the most defensible interpretation of key notions any Buddhist must recognise (and this will include the critical enterprise of demonstrating alternatives to be incoherent); second, to demonstrate that this reading more accurately captures the original thought of the Buddha, from which his 212

Nāgārjuna

Abhidharma competitors offer unwarranted departures. Nāgārjuna is thus typical of a fundamentalist reformer, legitimating what is innovative by presenting it as a reclamation of the original thought from which more established interpretations have in fact departed. Nāgārjuna seems not to have had much immediate impact on Buddhist thought, whether because his view was simply unexceptionably Mahāyānist18 or because his argument was to Sanskrit speakers so obviously fallacious.19 Although there were apparently others, only one early commentary on the Root Verses, entitled Fearless (Akutobhayā), survives at all. And there was an immediate successor, Āryadeva, who took up the Madhyamaka banner energetically enough to be considered nearly a co-founder of this distinctive interpretation of the Buddha’s teachings. A flurry of interest some centuries later, in the commentaries and debates of Buddhapālita and Bhāviveka in the late fifth century, and Candrakīrti in the sixth, revived Madhyamaka as a (to varying degrees distinct) Mahāyāna alternative to Yogācāra.20 Since the Root Verses is an enormously opaque text, most attempts to come to grips with Nāgārjuna’s thought therefore do so by reaching back through these layers of interpretation, particularly through the Tibetan reception of Candrakīrti’s interpretation of Madhyamaka.21 In what follows, however, I will use a reading of the Root Verses to anchor an articulation of Nāgārjuna’s thought from the other direction – viz., as it arose in the intellectual context of second-century CE Buddhist thought. Against the background of his own times, Nāgārjuna’s fundamentalist reformation hinges specifically on a recovery of what he takes to be the original and correct teaching of dependent arising ( pratītyasamutpāda), and the centrality this ought to hold in relation to all other teachings. Associating the distinctively Mahāyāna notion of emptiness with dependent arising should legitimate the Mahāyāna as more authentically capturing the Buddha’s original thinking than subsequent Abhidharma thought had managed. On Nāgārjuna’s view, dependent arising should replace – and not supplement or qualify – conceptions of being, thus undermining all metaphysical individuation tout court. Only this, he argues, will free us from the mesmeric power of individuating misconceived as apprehension instead of creation. The epistemic implications of abandoning individuation (without individuation, there is no assertion) explain both the style of the Root Verses and the elusive, apparently sceptical manoeuvres of the Dispeller of Disputes. And finally, so understanding the nature and centrality of dependent arising indeed revises the shape of our practical-soteriological task, from one of seeing ultimate reality as it is (instead of as it is constructed) to one of seeing experienced reality as dependent, or even seeing the ultimate reality of dependence in the fabrication of ordinary thought and life. The implications of this can be seen in the Precious Garland, which bridges everyday practical and ultimate concerns.

Fundamentalist Reformer: Centering Dependent Arising To see which fundamentals of the Buddha’s teaching Nāgārjuna wants to reclaim, there is nowhere better to look than the verse of praise of the Buddha that frames the Root Verses. These apparently formulaic verses which open Buddhist Sanskrit philosophical texts indicate the author’s priorities and the lens through which we are to understand the text which follows.22 The Root Verses are framed with: I salute the Fully Enlightened One, the best of orators, who taught the doctrine of dependent arising, according to which there is neither cessation nor arising, neither annihilation nor the eternal, neither singularity nor plurality, no coming nor going, for the auspicious cessation of prapañca.23 213

Amber D. Carpenter

Exactly what this prapañca is which ceases will have to emerge from an interpretation of the Root Verses. According to Siderits and Katsura (2013), “pra + √pañc literally means to be prolix or excessively wordy” (NMW, 126); it is often glossed in Buddhist-English discourse as ‘conceptual proliferation,’ and that will do as a working approximation. The relevant point for present purposes is that it is the teaching of dependent arising for which the Buddha is praiseworthy; and this teaching is the key to realising the supremely desirable end, characterised here as the stilling of prapañca. Nāgārjuna does not pick out the Buddha as a knower nor as supremely compassionate, and he entirely neglects familiar themes distinctive of the Buddha’s teaching – no self, non-clinging, impermanence, suffering and the roots of suffering.24 The opening verse does not so much as hint at themes more closely associated with the Mahāyāna, such as emptiness (śūnyatā).25 Instead, Nāgārjuna selects dependent arising as the sole basis for the Buddha’s praiseworthiness. Half the verse,26 in fact, is given over entirely not to praise of the Buddha but to characterisation of dependent arising – and the manner of its characterisation is striking. First, dependent arising is characterised entirely through matched pairs of negations, which is even more evident in the Sanskrit where each word of the first two lines begins with an alpha-privative. This is very far indeed from earlier sutta formulations of dependent arising as “from this, that” – as, for instance, “with ignorance as condition, formations [come to be]; with formations as condition, consciousness; with consciousness as condition, materiality-mentality.”27 Second, one of those negations is actually a negation of arising itself – unarising (anutpādam) seeming to negate the very thing that is asserted in dependent arising ( pratītysamutpādam). This is an acute form of the tension pervasive in the whole verse, which holds dependent arising responsible for an auspicious stillness (upaśamaṃ, translated ‘cessation’ above). This tension in honouring the Buddha as the teacher of unarising dependent arising signals that not only will Nāgārjuna hang his exposition of the Buddha-dharma on pratītyasamutpāda but his understanding of this central term will be challenging to the received orthodoxy.28 Nāgārjuna therefore opens the Root Verses proper by squarely situating his project with respect to the incontestable received teachings, the Discourses of the Buddha: Not from itself, not from another, not from both, nor without cause; Never in any way is there any existing thing [bhāvāḥ] that has arisen. (MMK 1.1) In its deployment of the tetralemma  – presenting and rejecting four ways beings (bhāvāḥ) might arise – this verse, and the chapter it announces, epitomises what the Naked Ascetic Kassapa Sutta (Acelakassapasutta, SN 12.17/ii.18–21) expresses more expansively, also deploying the tetralemma.29 In the Naked Ascetic Kassapa Sutta, the insistent ascetic Kassapa asks the Buddha whether suffering is created by oneself, by another, or both, or whether it arises spontaneously, and in each case, the Buddha replies negatively. He also rejects the suggestion that perhaps there is no suffering or he does not know that there is suffering. The baffled Kassapa asks the Buddha to teach him about suffering. “Kassapa,” the Buddha replies, [if one thinks,] “The one who acts is the same as the one who experiences [the result],”  .  .  . this amounts to eternalism. But, Kassapa, [if one thinks,] “The one who acts is one, the one who experiences [the result] is another,” . . . this amounts to annihilationism. Without veering towards either of these extremes, the Tathagata teaches the Dhamma by the middle: “With ignorance as condition, volitional formations [come to be]; with volitional formations as condition, consciousness. . . . Such is the origin of this whole mass of suffering.”30 214

Nāgārjuna

While this discussion clearly has a pragmatic import – it is the origin of suffering that is under discussion – the Buddha’s answer addresses the point from the most general angle, namely, how one should think of the origins of anything that arises. The false view of eternalism arises if one thinks, quite generally, “the one who acts is the same as the one who experiences” the result, while nihilism arises from supposing quite generally “the one who acts is different from the one who experiences.” Nāgārjuna’s first verse, seen in this connection, only makes explicit the generality implicit in the sutta’s tetralemma and, picking up on the ‘created by’ of the sutta, places the discussion squarely in the context of causes or conditions of arising. With such a sutta passage in the background, Nāgārjuna’s decision to present a discussion of conditions in the form of a tetralemma reinforces the message of the praise verse, that dependent arising is the focus of attention, since that is what is positively asserted as the Buddha’s teaching at the end of the sutta passage. Nāgārjuna’s audience will more specifically be primed already to anticipate the untenable implications associated with each limb of the tetralemma: the first option implying eternalism, the second nihilism, the third presumably inheriting the faults of both – since the “Dhamma by the middle” avoids both extremes, rather than mixing them – and the last is simply rejected. The explicit reiteration in Root Verses 1 of so much of the Naked Ascetic Kassapa Sutta draws attention to what it leaves tacit, namely “the Tathāgata teaches the Dhamma by the Middle.” If there is any work for Nāgārjuna to do, it will be in interpreting this, the proposed alternative to the four rejected options – “with ignorance as condition, volitional formations [come to be]; with volitional formations as condition, consciousness” and so on, up to “this whole mass of suffering.”31 Here is where Nāgārjuna will challenge prevailing interpretations and offer his own version of dependent arising as an alternative. What are these prevailing interpretations that Nāgārjuna challenges? While the centuries of interpretation since the Buddha’s parinirvāṇa took his teachings in various directions, and while it is difficult to determine with precision how exactly the various interpretations were developed at the time of the composition of the Root Verses, the evidence of a canonical Abhidharma text such as the Conditional Relations (Paṭṭhana) suggests that “the Dhamma by the middle: ‘With ignorance as condition, volitional formations [come to be]’ ” was predominantly interpreted as a claim about conditions and causes. Conditions for arising were many and various – and identifying, distinguishing, and describing the conditions by which things arise was tantamount to grasping dependent arising.32 For anything arising, there are multiple conditions for its arising and multiple further events or items for which it is a condition. This is dependent arising as a sort of hyper-distributed, massively ramified causation.33 This view is most naturally put in an ontological mode, but it can also characterise the phenomenological practices ascribed to the Theravādins of identifying and describing mental phenomena as conditions under which further mental phenomena arise.34 Whether ontological or phenomenological, appeal to dependent arising so conceived works as a part of Buddhist doctrine and practice by deflecting away from the tendency to pick out and privilege one thing as the cause, without asserting the sheer non-existence of causes (thus avoiding being stricken by nihilism). Instead of one cause, there are always many causes, none of them autonomous (self-causing, or uncaused). Meditative practices of distinguishing the many contributing factors from one another enable us to recognise this massively interconnected network of dependencies. Any apparent object of desire or aversion, fear, or pride thus dissolves into countless strands of endlessly ramifying causal conditions, undermining the coherence of these afflictive emotions. This general metaphysical point translates immediately to the personal context: there are no autonomous acts, and therefore no agents as we might feel ourselves to be or aspire to be. The self – both as agent and subject – dissolves into countless dependencies on both personal 215

Amber D. Carpenter

and impersonal factors. This eliminates the coherence of the notion of a real self that might be successfully distinguished from the rest. The result is not the discovery of the true self, but the acceptance that there is no self to seek, and no empty hole where a self could be.

Nāgārjuna’s Critique: Its Target and Motivation If this aptly, if impressionistically, characterises Abhidharma thinking about dependent arising, what, we might ask, could Nāgārjuna possibly find objectionable? We might begin at the broad exegetical level. In the tetralemma of SN 12.17, all four limbs are rejected, and Root Verses I.1 explicitly picks up on this. The “Dhamma by the Middle” – that is, dependent arising – is offered instead of any of those four options. But any analysis of dependent arising that figures it as causes and conditions, however multiple, must necessarily present these multiple conditions and their results as either the same as or different from each other – that is, as either the first or second limbs of the tetralemma. But these options (as well as the option of combining or rejecting both), were rejected by the Buddha, as leading to eternalism and nihilism, respectively. So, any apt understanding of the Doctrine by the Middle must construe it as an alternative to causes and conditions, not as a special version of it. This would be a textual basis for Nāgārjuna’s objection, and for supposing this is what he is objecting to. But his objection, seen in this light, is not to any particular theory of causation, and it would hold even when all parties acknowledge that causation, as a relation, is merely nominal and not ontologically robust. Indeed, Nāgārjuna’s objection is only incidentally about causation at all: in the first place, because distributed causation is how Ābhidharmikas mistakenly construe dependent arising (the topic it is essential to get right); and in the second place, because causal relations share problematic features of any relations and model the relevant features of any asymmetrical or asynchronous relation. The Buddha offered the Dhamma of the Middle as an alternative to either identifying or distinguishing cause and effect. What his Dhamma of the Middle rejects, then, is the identity or difference of any cause(s) and its/their effect(s) – and by extension, says Nāgārjuna, of any relata. The objection, it seems, is to the individuation of relata and, in Root Verses 1, of cause and effect in particular. Thus, Root Verses 1 aims to show that if the relata ‘cause’ and ‘effect’ can be individuated from each other, then they cannot arise dependently. If the relata exist prior to and independently of their relation,35 then each must have a distinct identity: “Nowhere does there exist any such thing as an existent [bhāvaḥ] without defining characteristic”;36 and “There is no existing thing [bhāvaḥ] that is without svabhāva.”37 But, Nāgārjuna reasons, if each relatum has its distinct identity independently of any relation to any other, then those distinct individuating natures themselves cannot be dependently arising.38 They were either there all along and so did not arise at all (the first limb of Nāgārjuna’s tetralemma) or else they pop into existence apropos of nothing (the fourth limb).39 To make that which distinguishes an existing thing from everything else  – that which makes it this individual  – dependent on another, as the second limb of the tetralemma suggests (and as perhaps the Ābhidharmikas thought they were doing) is to acknowledge that there is no individual there to be related prior to the relation, and thus the whole notion of a web of causal relations between discrete individuals falls apart. Massively distributed causation is not merely not dependent arising; it is antithetical to it, because it presumes the individuation of the relata, and such individuation requires independence. Thus, it is significant that the primary assertion of Root Verses I.1 is that “Never in any way is there any existing thing [bhāvāḥ] that has arisen.”40 No bhāvaḥ arises in any of the four ways because a ‘being,’ were there any, would be something distinct from other beings, something independently identifiable as ‘this.’ But there are no such things. 216

Nāgārjuna

The practical objection to the Abhidharma conception of dependent arising as distributed causation is that it leaves individuation intact, not just as an apt characterisation of reality but also as a mode of thought, with its attendant implicit values.41 There may not be wellindividuated personal selves on the Abhidharma view; but this is explained – as in the muchdiscussed chariot analogy in The Questions of Milinda (Milindapañha), which itself is rooted in SN 5.10 – by appeal to the constituent elements on which such apparent wholes depend. These constituent dharmas enjoy the independent identity which persons and chariots lack, and enjoy greater claim to reality in virtue of that fact. The Abhidharma conception of dependent arising thus dissolves the personal self into countless “selves,” retaining the notion that to be real is to have well-individuated identity independently of others – so that instead of pointing away from the self and the search for the self, the Abhidharma understanding of dependent arising as a way dharmas associate risks simply attaching us more to the notion that a clear delineation of independently specifiable individuals-in-relation gets at how things really are.

Nāgārjuna’s Alternative Nāgārjuna rejects distributed causation as the correct construal of dependent arising, and therewith the fundamental picture of reality as beings-in-relation. If well-individuated relata cannot be dependently arising, then conversely what dependently arises cannot be well-individuated. Thus, in the course of the Root Verses, the point from Chapter 1 is generalised across any relations and relata,42 with asymmetrical or asynchronous relations modelled on the examination of causation in Chapter 1 and symmetrical or synchronic relations modelled on the analysis of traversing-traversed in Chapter 2, which claims that it is both necessary and impossible to distinguish two conceptually dependent things (in this case, traversing and the space traversed, 2.1–5, or motion and mover43); in Chapter 6 on Desire-Desirer, where there is dependence, one cannot have co-occurrence and distinctness together (6.8); Chapter 10 on Fire-Fuel concludes, upon consideration of the necessary mutual dependency of fire and fuel (10.10–12, et passim), that “they are not considered by us to be wise instructors in the teachings of the Buddha who describe the subject and existents in terms of identity and difference” (MMK 10.16). In each case, it turns out to be both necessary and impossible for what is distinct to be in relation.44 This is a wholesale critique of beings as individuals, and one which can be extended to practices of individuation with minimal ontological commitment.45 Just as the Buddha’s anatta teaching was not the assertion of an absence where a ‘self’ could have been, so Nāgārjuna’s critique of individuation does not assert the contingent nonexistence of what might have existed. On the contrary, Nāgārjuna insists that it is rather the very dichotomy between ‘exists’/‘does not exist’ – set up by thinking of discretely identifiable individuals in terms of identity and difference – which is the problem: What entity is prior to arising and the rest, what entity is simultaneous, and what entity comes after — these do not exist; the concepts of existence and nonexistence no longer apply there. (MMK 9.12)46 So this is not the view that there is nothing instead of something; only “those of little intellect  .  .  . take there to be existence and non-existence with respect to things” (MMK 8.13). Rather, the conclusion is that there are no things – with the additional specification that it did not take much to be an Abhidharma ‘thing’ in the first place. Space, nirvāṇa, and the Four Noble Truths are, on one standard Abhidharma analysis, all distinguishable items – existing 217

Amber D. Carpenter

things in the sense of knowable dharmas – without being substances (dravya). And Nāgārjuna criticises these only tenuously thing-like individuals, too, on the same grounds that he criticises all purported beings (space, MMK 5; the Noble Truths, MMK 24; nirvāṇa, MMK 25). This is thus not merely the rejection of robust Aristotelian substances as bearers of properties or as primary existents, but of any kind of individuatable ‘this’ at all.47 In place of things-in-relation, there is dependent arising. “Whatever exists in dependence, that is free of svabhāva” (MMK 7.16).48 Thus, it is incoherent to demand an account of what arises dependently, for the very conception of dependent arising is the erasure of the model of individuals-in-relation.49 In a move to appropriate, and perhaps implicitly critique, Mahāyāna notions of emptiness, Nāgārjuna goes on in Root Verses Chapter 4 to characterise this as the appropriate sense of that key Mahāyāna term, emptiness (śūnyatā). Drawing on Chapter 1’s critique of causal relations, Nāgārjuna says that bodily form or rūpa (and likewise, the other “aggregates,” or skandhas, 4.7) is neither separate from, nor similar to nor dissimilar to its cause, nor is it without cause. And he concludes by characterising this as a refutation and explanation based on emptiness (MMK 4.8–9). An argument based on emptiness, that is to say, is an argument based on the unreality of discrete individuals, as evidenced in the neither-samenor-different character of any purported individuals.50 What arises dependently cannot be said to be identical to or different from that on which it depends (22.6–8); nor, for the same reason, can it be said to exist by means of svabhāva (22.9). “Thus,” according to Nāgārjuna, “both that on which he depends and the one who is dependent are altogether empty” (22.10). Emptiness is thus presented as non-individuatedness, or what follows from it51 – not just the lack of svabhāva in the sense of substantial (as opposed to insubstantial, merely apparent) reality, but lack of individuatedness, of the distinctness in virtue of which things are thought to exist as the identifiable things they are. Where this lack of individuation was the account of dependent arising proposed in place of any causal model in Chapter 1, it should be no surprise to find emptiness then used explicitly to characterise dependent arising in Chapter 24: “Dependent arising we declare to be emptiness” (MMK 24.18).52 Nor, given this assimilation, is it any wonder that prapañca is extinguished in emptiness (MMK 28.5). Where there is no individuation (the emptiness which is ubiquitous is emptiness of svabhāva), language and concepts  – which function, essentially, to discriminate discrete items as ‘same’ (as itself) and ‘different’ (from not-this) – necessarily lose their function.53 Recognising that dependent arising is absence of any individuation, there is quite simply nothing left for thought and language to do. The stillness of conceptualising promised in the framing verse is the natural result. Thus, the declared assimilation of emptiness and dependent arising is immediately followed by the observation that emptiness itself “is a dependent concept” (MMK 24.18). ‘Emptiness’ is not a thing or a substance or a quality that pervades reality – nothing enjoys that status; nor, moreover, is it independently specifiable in some more attenuated way. Emptiness is a way of conceptualising dependent arising, meaningful within this specific context as a way of indicating the ubiquitous absence of independent discriminability, or the pervasive nonindividuatedness due to dependency. In a precursor to Diṅnāga’s apoha theory, any apparently well-individuated concept, ‘emptiness’ included, derives whatever meaning it has through its association with other concepts, and not by accurately referring to real individuals.54 Perhaps ‘emptiness’ is a particularly useful way of characterising dependent arising; dissolving the apparent independent specifiability of anything whatsoever, it may well be “taught by the conquerors as the expedient to get rid of all views” (MMK 13.8).55 But it remains, for all that, a concept or a way of grasping something, dependent for both its sense and its usefulness on the very conceptions of individuation and existence which it seeks to undermine. Without a 218

Nāgārjuna

notion of discrete identities, there is nothing for emptiness to be empty of;56 and were there no habits of individuating and relating discrete items, there would be no good for understanding emptiness to do.57 As far as ‘the thing itself’ which the word ‘emptiness’ presumably refers to, this just is dependent arising; if there was ever a place for necessary self-predication, it is surely in the dependent arising of dependent arising. Bringing us back to the Discourse with Kassapa, the verse concludes, “just that is the middle path.”58 Most contemporary philosophical scholarship on Nāgārjuna recognises this as articulating a kind of anti-foundationalism, whether metaphysical, explanatory, or semantic.59 There is nothing that explains, grounds, gives being to, or validates anything else, which does not also depend on that very thing for its very identity. In this way, anything that could be picked out or individuated in any way is equally non-basic.60 This equalising of everything as dependent seems reinforced by the infamous claim at Root Verses 25.19, that “there is no distinction whatsoever between saṃsāra and nirvāṇa.” Even in the most basic practical distinction between that from which we move and that towards which we move, their individuation as such is mutually dependent, and they are no different from each other in this respect.61 Such anti-foundationalism, however, may be compatible with considering dependent arising as, in a certain sense, ultimate: dependent arising explains or grounds the possibility that anything whatsoever should arise, without ever explaining that this rather than that should arise here.62 As Nāgārjuna understands it, dependency necessarily impugns the well-individuated nature of any purported existent, rendering impossible the individuation necessarily prior to any relating. This is the true heart of the Buddha’s teaching of dependent arising – not the fragmentation of being into countless impersonal beings and a frenetic mental hygiene of chasing down endlessly unfolding relations between them,63 but rather the utter absence of real individuals, together with a mental hygiene of ceasing to individuate what is not really individual.

Tetralemma, Scepticism, and Speechlessness There are, then, significant implications of Nāgārjuna’s construal of dependent arising for language, thought, and the sorts of philosophical argument for which these are the necessary medium. For all language and all thought is individuating, and this is so even if we think that meaning arises from use and not by reference to entities – if words do not delineate, they cannot mean or be used as words at all.64 Likewise, all argument relies on the explanatory priority of explanans to explanandum. If individuation itself is incoherent, there will be no way to state the position which does not rely on the very incoherence it diagnoses. Even to say that dependent arising is ultimate reality – as we must in English – says too much, for it invokes the existence and identifying-individuating power of ‘is’ which Nāgārjuna eschews. Similarly regarding “emptiness is lack of svabhāva,” a claim Nāgārjuna immediately follows up with “It is empty” is not to be said, nor “it is not-empty,” nor that it is both, nor that it is neither; [“empty”] is said only for the sake of instruction. (MMK 22.11) One might use words as a kind of paedagogical gesture, to facilitate the appreciation of reality as not carved in the way words necessarily do carve: “The ultimate truth is not taught independently of customary ways of talking and thinking” (MMK 24.10). But any such gesture must be immediately undercut by reminders that “one should assert neither real nor non-real”65 in order to dispel the false impression that the customary ways of speaking are accurate articulations of how things are, or indeed that things (distinct and related to one another) are. It is an 219

Amber D. Carpenter

unavoidable implication of this construal of dependent arising that the via negativa is the only way, and even here one must be careful that the negations do not imply an assertion of their opposites: not being must be supplemented by not non-being. This makes Nāgārjuna’s way of presenting his project an elegant – perhaps even necessary  – fit for the particular aims of his project. Since the aim is to present a construal of dependent arising which dislodges the individuals-in-relation picture, and its associated identify-and-relate mode of thinking, the manner in which this is presented matters. That manner is most conspicuously the destructive tetralemma itself, which structures the first chapter of the Root Verses, and looser but similar constructions in subsequent chapters (see MMK 3.2, 5; 4.1, 12.1, 18.8, 22.17–18). Much scholarly work has accordingly been done on the logic of the tetralemma;66 more attention still has been given to Candrakīrti’s codification of Nāgārjuna’s method as exclusively prasaṅga – exclusively making use of others’ views in order to reveal an unwanted (to the opponent) implication of those views.67 More immediately relevant than the logic of the tetralemma to Nāgārjuna’s philosophical goals as outlined is the fact that Nāgārjuna negates each limb  – just as the Buddha did in Saṃyutta Nikāya 12.17, considered previously. And more salient than potential codifications of method is the regularity with which (apparently) exclusive and opposite pairs are both negated.68 One might worry, however, that this leaves the Mādhyamika in an awkward position. If everything sayable is inadequate simply in virtue of being sayable, then it looks as if every claim is equally inadequate. But if every claim in conventional language is equally inadequate, then it would seem that, conversely, any claim is as valid as any other: the teachings of Gotama the Nyaiyāyika would be just as (in)valid as the teachings of Siddhartha Gautama, Sage of the Śakyas. And there is a deeper problem. The very practice of reason-giving is a practice of asserting relations of priority and dependency among things. (The ‘cause’ discussed in Root Verses I.1 is hetu, which means ‘reason’ as much as ‘cause’; and the critique counts as much against explanation as against causation.69) And fundamentally, any knowledge claim implicitly asserts the independence and priority of the means by which one knows. For any knowledge claim, my knowledge is only as secure as my basis for believing it. The proximate basis is my evidence – thus, the asymmetry implicit in any explanatory enterprise. But any evidence must be accessed in some way, whether through direct experience or through reasoning. And thus, any evidence can only be as strong as my mode of access to it is reliable. That is to say, any knowledge whatsoever depends upon reliable means of cognition, or pramāṇas (if my vision is faulty, no visual evidence in the world can deliver accurate knowledge of visibles to me). In this way, a reliable means of cognition underlies the very possibility of valid cognition – and so is prior to any particular valid cognition, and independent of it. Seeing must be determinate independently of any particular visual cognition, if any particular visual evidence is to count for anything. But this sort of independence and priority is just what Nāgārjuna has undermined with his new interpretation of dependent arising.70 He has, therefore, undermined the very possibility of having a valid cognition of emptiness or dependent arising – and with this, he seems to have lost the means for asserting dependent arising, or emptiness, itself. This seems to be the animating concern of Nāgārjuna’s Dispeller of Disputes. Valid means of cognition, if they are determinate and distinct from their objects and the knowledge they give rise to, must be either validated by some yet prior means of cognition (perhaps inference grounds confidence in perception) or must be self-validating. If Nāgārjuna rejects both of these – as he must, since the former leads to a vicious regress, while the latter reintroduces independent foundations the Root Verses argued against – then he is left unable to claim any validity for his own claims about emptiness, lack of individual being (bhāvāḥ) or svabhāva, and dependent arising. 220

Nāgārjuna

While Nāgārjuna presents this dilemma as an objection to his proposed interpretation of the Middle Way in terms of emptiness, this is in fact a pretext for an extension of that interpretation into a trenchant critique of the whole epistemic project. For, after quickly disposing of the trivial form of the objection based on the misconstrual of ‘emptiness’ as non-existence, Nāgārjuna rebuts the objection by arguing that any opposing view is as much skewered on this objection as his own  – and indeed, more so. No means of cognition is self-validating, nor is it possible to have infinitely stacked modes of cognition validating each other – this is so, Nāgārjuna argues, whatever one takes one’s metaphysics to be. The dependency which was incompatible with individual identities in the Root Verses likewise infects all the tools by which one might argue for or against it (or anything else). So, trying to launch an argument against Νāgārjuna’s Madhyamaka by relying on such tools is misguided  – the tools themselves would be revealed as either only conditionally or capriciously authoritative, thus demonstrating rather than defeating Nāgārjuna’s point. Knowledge as the objector conceives it, when he complains that Nāgārjuna cannot lay claim to it, is simply impossible in any case. Nāgārjuna’s strategy and conclusion here is quite similar to the classic Problem of the Criterion, which Sceptics of Graeco-Roman antiquity deployed to such effect.71 But Nāgārjuna’s aim is not the same as the sceptic’s.72 For the argument leads to sceptical conclusions only for one who is committed to a certain conception of knowledge – a conception on which causes are prior to, independent of, and explanatory of their effects; and a conception on which therefore causes of valid cognition, the apparatus for cognising itself, must be prior to and independent of the object of cognition. Dispense with this prejudice and the non-independence of means of cognition does not impugn cognitions at all. One recognises merely that their credibility is neither freestanding nor indefeasible. This makes honest humility and a standing openness to reconsideration the best epistemic condition to be in, for seeking certainty (even relative certainty) is a fool’s errand. But only Nāgārjuna’s Madhyamaka is willing to make this move, and so only the Mādhyamika is free of the very scepticism of which they stand accused. Of course, dispensing with the epistemic foundationalist prejudice is more easily said than done, for if one goes about it via vacua as Nāgārjuna demands, it seems to entail eschewing the claim to have asserted anything at all. Indeed, Nāgārjuna encourages this understanding of the implications: “If I had any thesis (pratijñā)” he (in)famously says (Dispeller 29), “that fault would apply to me. But I do not have any thesis, so there is indeed no fault for me”73 – leaving uncomfortably ample room for speculation about what it means for whatever we read in the Root Verses to not be tantamount to expressing a thesis, position, or claim (pratijñā can mean any of these), and likewise what it meant there to say that emptiness rids one of all views. An extremely capacious understanding would include any assertion or claim whatsoever about anything – indeed, any articulate speech, thought, or view at all, since we saw previously that thought and language necessarily individuate where there are no corresponding individuals. On such an understanding of ‘view,’ the Root Verses might be understood as a work of literary fiction, or at any rate a mere device to induce a certain mental state.74 A more restricted understanding of ‘view’ construes it specifically as ‘speculative’ views, theoretical or metaphysical claims, or  – as Siderits and Katsura have it  – even more specifically with metaphysical views about the ultimate nature of reality.75 Then the point would be not the cessation of conceptualising altogether, but rather to cease understanding concepts as even potentially well-defined terms for referring to a well-delineated reality. An approach to the question which would integrate it into the preceding discussion of dependent arising and emptiness might begin by associating ‘view’ with emptiness: something counts as a ‘view’ just in case it purports to be non-empty, which is to say lays claim to being independently well-defined or referring to some well-defined individual whose identity 221

Amber D. Carpenter

as that individual does not depend upon the very thing it explains, grounds, or refers to. So understood, there is an obvious sense in which Nāgārjuna has expressed no views, for he has already acknowledged the dependency of everything he has said – emptiness itself depends for its intelligibility and coherence on prior and mistaken notions of individual beings. His own claims are explicitly empty in just the sense that everything is: arising dependently and not identifiably distinct from the conditions on which they depend, including the tools for cognition which are themselves dependent on the cognising and object cognised. It is, then, not a thesis or a claim in the sense that the interlocutors mean it at all. On the one hand, Nāgārjuna has been careful not to even try to give a positive specification of his position as distinct from all other positions, preferring instead to give us reason to stop adopting views and mental practices that even implicitly presume independent identity. On the other hand, in as much as even the via negativa relies implicitly on distinguishing thoughts, recognising reasons, and relations between statements, Nāgārjuna can happily accept that any characterisation of or argument for his view is dependent – and not just incidentally dependent on prior causal factors, but dependent for whatever apparent identity and distinctness it has on precisely those mistaken notions of individuality and being they aim to undermine. But this preliminary understanding of the sense in which Nāgārjuna has no views – and how having no views might enable him to escape the self-refutation objection while redeeming the meaningfulness of the Root Verses – only goes so far. If Nāgārjuna gets the better of his opponents, it is only in as much as he recognises the inevitable emptiness of his own claims, their non-absoluteness or lack of ultimate grounding in some self-sufficient mode of cognition, while his opponents lack awareness that their own views are in the same position. But dialectically, this should only put all contending views on all fours: all views arise dependently, lack coherent and distinct content, and rely for their intelligibility on similarly dependently arising factors. If this is where we are left, we might think that Madhyamaka avoids logical self-contradiction only at the drastic cost of a kind of practical self-abnegation: if the Madhyamaka view is correct, that very fact eliminates the point of saying so and the possibility of showing it. Two distinct worries may be distinguished here, the first to do with everyday practice and the second with insight and the ultimate Buddhist goal. If thought and language can never get it right about how things really are, and if everyday speech is not even trying to make claims about how things really are, then everyday language can never be criticised for having got it wrong. Everyday language is an effective instrument for achieving everyday ends, and so long as it succeeds in that objective, there is nothing going wrong and nothing in need of correcting. This, however, suggests that there is no space for a reforming Buddhist outlook on practical everyday affairs – we must simply leave everything in its place.76 This would leave even the Buddha’s ambiguous talk of the self, so at odds with everyday notions, as itself a bit of idle chatter without practical relevance. The second concern is that if the means of substantiating arguments is itself undermined, then there is no adjudicating between views – not even between the view that we should abandon all views, and the view that we ought to aim, as the Eightfold Path suggests, for Right View. In short, it seems the Mādhyamika is saved from speechlessness only by destroying the inherently teleological character of the Buddhist path. For if ordinary speech generally is fine just as it is, so long as it is intended and taken pragmatically – as indeed it ordinarily is  – and without implying some conviction about the really real basis underwriting it, then what needs reforming? If “there is no distinction whatsoever between saṃsāra and nirvāṇa” (MMK 25.19), where is there to go? And if all language that does try to say anything about ‘how things really are’ is equally falsifying, this seems to render Buddhist insight utterly mute.77 222

Nāgārjuna

Practical Implications, Proximate and Ultimate We might see our way out of this by turning to the text in which Nāgārjuna seems most explicitly concerned with just this issue.78 The Precious Garland is presented as ‘advice for a prince’ – so it is explicitly concerned with everyday life and how that might be lived better. What there is to be said on this score, however, does not come neatly disentangled from ultimate concerns of a specifically Madhyamaka kind, nor is it disconnected from the ultimate aim of escaping worldly life altogether. The text sets up two distinct ends – happiness and the highest good — with virtuous practices and wisdom as their respective means (RĀ 3–4).79 This may seem at first to reinforce the idea that the Root Verses’ view of dependent arising as lack of individuation leaves us with two disconnected alternatives: either we fully embrace its implications for language and thought, and thus remain mute and disengaged (literal quietism); or we accept the categories of the world as they stand and aim for worldly goods as these are understood (practical quietism).80 Even if one could alternate between the two, one could never use the former to reform the latter, nor the latter to approach and articulate the former. This disjunctive quietism would seem to leave the practical commitments of the committed bodhisattva impotent. So it is essential to see how the text gives us more than that and, in doing so, addresses how it could be that “the ultimate truth is not taught independently of customary ways of talking and thinking” (MMK 24.10) – indeed, how it could be that Nāgārjuna follows his extended critique and re-conception of dependent arising in the Root Verses with a presentation of the twelve-fold chain of dependent arising (MMK 26) that is, as Katsura and Siderits (307) put it, “the perfect model of orthodoxy.”81 The Precious Garland articulates two crucial ways in which the two ends (worldly wellbeing and ultimate attainment) and their respective paths converge. First, on a practical level, we are assured that aiming for happiness will indeed eventually issue in attaining the highest good.82 The text then bears this out in its extended discussion of sense-pleasures, which takes the main constituent of happiness as ordinarily conceived and uses the prince’s interest in it to motivate a concern not only for virtue but for wisdom and liberation. Note, for instance, how what starts as a familiar Abhidharma exercise on the foulness of the body as a means to inducing disenchantment with sensual pleasures (RĀ 148–70) – essential for virtuous practice  – leads to an orientation towards “unparalleled enlightenment” rooted in “wisdom not relying on duality” (RĀ 174); and later how the analysis of pleasure at RĀ 348–63 relies on specifically Madhyamaka positions and techniques concerning emptiness and insubstantiality. Indeed, it turns out that one of the four main virtuous practices for a happy life is none other than wisdom (RĀ 139), so that to aim at happiness turns out, as promised, to lead one to enlightenment. The higher goal is actually built into the path to the lesser goal. This convergence of higher and lower goals grounds a second convergence, which operates in the opposite direction. For now, leading to wisdom operates as a constraint on what could count as virtuous practice. The fact that a certain ‘ordinary outlook’ or way of conceptualising leads – or does not lead – to wisdom as the Mādhyamika understands it is a reason for considering such an outlook as either apt (134–35) or in need of reform. Thus, we have a basis for critique of ordinary practice, after all, and indeed a basis in ultimate truth and not just in the several prescriptions found in scripture – though not in the way the Abhdidharma might have suggested. Common ways of speaking and thinking are indeed riddled with error, not because they get at the wrong objects, but because thinking in that way causes suffering. Given dependent arising/non-individuation/emptiness, no distinctions and ways of individuating are correct characterisations of how things really are, and wisdom will consist 223

Amber D. Carpenter

in recognising this. Ordinary modes of thought can be criticised if and when they block this recognition. Thus, those conventional uses of language are apt which lead one towards recognition of ubiquitous dependency and its concomitant insubstantiality – and this is indeed what the Precious Garland does with the prince, a paradigm worldly figure, in verses 125–75, and elsewhere. A perfectly orthodox exposition of the twelve links of dependent arising may do the same. Uncritical participation in the language of individual persons (‘I,’ ‘mine,’ and associated desires: ‘I want this to be mine’) probably will not. A thorough examination of the relationship between individuation, causation, existence, and reality – such as the Root Verses itself – would be a use of language which facilitates this, as would thinking of emptiness as the lack of substantial individuation necessarily implied by dependency. Going along with whatever the world happens to say about individuation and causation may or may not be fruitful, depending on the circumstances. Perhaps there is no immediate harm in using the language of persons for everyday transactions, but it can be difficult to do so while simultaneously maintaining the perspective that recognises the purposes of these transactions themselves are most often riddled through with confusions about self and individuation that cause suffering.83 And so it may well be appropriate for even ordinary folk with worldly commitments – the prince, for instance – to engage in active critique of ordinary language and concepts (RĀ 37–73). The Mādhyamika is not restricted to using all ordinary language with only its ordinary meanings without critique, for critique may arise from the conduciveness of practice to happiness and final liberation. Nor then, conversely, is the Mādhyamika restricted to using ordinary language for only its ordinary non-philosophical purposes; reasons can be given and reasoning efficaciously engaged in, without indulging in the self-refuting presumption that one’s reasons and reasoning enjoy an implausible epistemic or ontological independence. None of this requires that philosophical reasoning be a necessary and exclusive means to ultimate realisation of the Buddha’s teaching of dependent arising. There is no reason to restrict the effect of language on the mind in that way, and every reason to suppose that nonargumentative use of language may also have salutary effects on one’s phenomenology, state, or outlook – indeed, nothing speaks against reading the Root Verses itself in such a literary way, if that accomplishes the same end.84 It is noteworthy in this regard that the four hymns most centrally associated with Nāgārjuna engage explicitly in reasoning to achieve their aim,85 as well as in rehearsing terms of approbation in a more poetic way.86 We need not suppose some strict distinction between philosophical engagement on classic philosophical topics such as cause, relation, individuation, essence, and existence, on the one hand, and ‘merely’ skillful teaching on the other.87 On the contrary, these might often be one and the same, as is demonstrated not just in the Root Verses, but also in the Precious Garland. But if one does engage philosophically with the Root Verses (and other analytical texts, such as the Dispeller of Disputes), whether negatively or constructively,88 what one can no longer do is naïvely suppose that such engagement is a search for some lynchpin, whether that be one thing or many dharmas. Nor may one engage in practices of individuating, identifying, relating, and classifying – making Nāgārjuna’s work into a worldview – as if such a mental exercise were essentially different from the more modest individuating and relating we do in everyday life. The Root Verses, however, has this advantage over ordinary thought: it carries its own antidote within it. By enabling us to recognise the individuating we do while we are doing it – making, indeed, such recognition inescapable – the Root Verses ensures that we are no longer implicitly convinced of the superficial unimpeachability of our thinking. To engage in thinking through the Root Verses is to become aware of the dependency of that very activity – and it may possibly even suggest an alternative.89 224

Nāgārjuna

Conclusion To sum up: Nāgārjuna is a fundamentalist reformer, taking us back to what he asserts is the core of the Buddha’s teaching – viz., dependent arising – and the correct understanding of that. His critique of the Abhidharma aims to transform how we think of dependency and therewith the distinction between ultimate and conventional, and of the Buddhist goal and the path to it. As Nāgārjuna tells it, the Ābhidharmikas are working with a view according to which existing things dependently arise. Dependently arising characterises the behaviour or relations of things with respect to one another; but it can only do so on the presupposition of individuation; that is, of a prior distinction between two things (elements of experience, existing things) subsequently related. When seen in this light, it becomes clear that this Abhidharma understanding of dependent arising does not, after all, do away with a picture of reality that fundamentally sees individuals as acting on and being affected by each other; it only complicates that picture. It depersonalises the picture, to be sure, but the basic elements into which we analyse experience or existent things retain the characteristics of persons  – having a distinct identifying nature which makes them what they are by distinguishing them from everything else. Such a model, Nāgārjuna aims to show, positively precludes dependency and eo ipso dependent arising. For dependency is incompatible with having a discrete identity and therefore incompatible not only with causation, but with any relating that presumes (as it commonly does) the distinctness and identifiability of the relata. If causation is especially relevant, it is because (i) it is the terms in which his Buddhist opponents understand dependent arising, so it is the view of dependent arising he has particularly to displace in his project to recover the Buddha’s original teaching by the middle; and (ii) causation can be used as a model for any asymmetrical relation, and in particular for epistemological dependency relations inherent in explanation. Nāgārjuna’s larger project in the Root Verses is to defend a replacement conception of dependent arising ( pratītyasamutpāda instead of existence/non-existence) against the Ābhidharmika supplementary conception of dependent arising ( pratītyasamutpāda as a way in which things arise). Instead of conditioned relations between discrete individuals, instead of the being/nonbeing, same/different dichotomy this implies: dependent arising. Eschewing individuation itself (and not just an ontological commitment, since it is as much the mental habit of individuating that generates suffering as any additional existence claim on top of that); recognising the imperfect individuation of things, their imperfect sameness and difference from others and themselves; and recognising this incomplete differentiation as due to dependency – just this is the Middle Way. This leaves much still up for debate. For instance: does relating – even the conventional, epistemological, and non-ontological activity of relating – indeed presume the prior individuation of its relata? Are existence and essence so mutually bound up that rejecting the latter implies losing also the former? Is Nāgārjuna’s alternative proposal even coherent – or has it not, perhaps even wittingly, removed its own conditions for the possibility of being coherent? And in spite of the suggestions offered in this chapter, there is still ample room for debating the practical implications of this: is the implication a kind of recommendation to extreme quietism, a re-conception of the ultimate prajñā to be obtained as simply not thinking anything at all, and a rejection of critical engagement with everyday notions? Is Nāgārjuna a sceptic in the classical mould, using critical thinking to defeat any ambitions of critical thought, for the purposes of a disengaged equanimity? Or if we should stay engaged, how does the Madhyamaka view support or revise that project? How will correct understanding of pratītyasamutpāda – a distinctively Madhyamaka, non-Abhidharma understanding – make one a better leader in the world? It is indeed fitting that such questions should remain open, for the power and inescapability of Nāgārjuna in Buddhist philosophy consist in how much new terrain his thought opened up 225

Amber D. Carpenter

and the standard he set for the degree of penetration, subtlety, and systematicity to be brought to the task.

Notes 1 I am very grateful to Paul Livingston for insightful and thoughtful comments on an earlier draft, and to Jay L. Garfield and William Edelglass for their comments on a later draft. Graham Priest’s comments on the penultimate version helped me to clarify important points. Enormous gratitude goes also to Nicholas Lua for exceptional research assistance. And finally, I thank the editors of the volume – particularly William Edelglass, by whom to be edited is a joy – for their invitation, for their patience, and for constructive support in bringing this to fruition. 2 Ruegg (1981) is still unsurpassed for a scholarly philosophical overview of the history and literature of Madhyamaka thought in India. 3 Joseph Walser (2005, 61–88), argues for the dates and place, for which there is cautious scholarly support; Walser argues more specifically that Nāgārjuna “wrote the Ratnāvalī within a thirty-year period at the end of the second century in the Andhra region around Dhānyakaṭaka (modern-day Amaravati)” (Walser 2005, 61). 4 Westerhoff (2010) offers a lucid translation of, cogent commentary on, and introduction to the Dispeller of Disputes. 5 Komito (1987) translates just the verses from Tibetan into English, as does Tola and Dragonetti (1987), which includes the Tibetan, as well. Lindtner (1982, 34–65) translates verses with auto-commentary from Tibetan into English. Erb (1997) offers a partial translation of Candrakīrti’s commentary into German. 6 Lindtner (1982, 102–19) offers a translation of the Sixty Verses, as does Tola and Dragonetti (1995); Loizzo (2007) includes a translation of Candrakīrti’s commentary along with the verses. 7 Westerhoff (2018a) offers a translation of Crushing the Categories, tackling in the introduction the question of attribution. Tola and Dragonetti (1995) also offers a translation into English. 8 Translation of the Precious Garland from the Sanskrit can be found in somewhat old-fashioned English in Tucci (1934; 1936). Hopkins (1998) provides a more modern English translation, from the Tibetan. 9 Pind (2001) and Tola and Dragonetti (1995) have challenged the attribution of Crushing the Categories, for instance, while in the opposite direction, Lindtner (1982) has a more expansive (and therefore more contentious) list of authentically Nāgārjuna texts, including in addition to the previous: the Letter to a Good Friend, Proof of Convention (Vyavahārasiddhi), Four Hymns (Catuḥstava), and Verses on the Heart of Dependent Arising (Pratītyasamutpādahṝdayakārika), among others. 10 Nāgārjuna need not think that his Middle Way is distinct from other Mahāyāna interpretations of the Buddha-dharma in all respects in order to think there are important enough differences. 11 We are, incidentally, similarly situated regarding Plato’s dialogues and the letters traditionally attributed to him. See Burnyeat and Frede (2015) for recent discussion. 12 See Arnold (2005, 6, 117; 2010; 2012); also see Westerhoff (2009a) – an excellent resource for getting to grips with Nāgārjuna’s Madhyamaka philosophically – and Shulman (2011). Thomas Wood (1994) proposes a metaphysical nihilist interpretation; see also Westerhoff (2016). Oetke (2007) proposes metaphysical illusionism as a viable alternative (though Oetke [2011] rejects metaphysical readings). Ferraro (2013; 2014) proposes ‘realist antimetaphysics,’ which is as much metaphysical as Kant’s critical philosophy may be. Priest and Garfield (2002) offer a logical reading of the Root Verses, more metaphysical than epistemological in its claim that Nāgārjuna endorses a paraconsistent logic that sanctions true contradictions, though it may also be given a semantic construal, as Garfield endorses in Siderits and Garfield (2013). 13 Ruegg (1977; 1983); Garfield (1996); Siderits (2003b; 2007) (semantic non-dualist). 14 Arnold (2010) focuses on explanation, while Ferraro (2017) is epistemological in a rather different, antimetaphysical way. Matilal (1986, Ch. 2) and Mills (2018) present Nāgārjuna as a skeptic, as does Garfield (1994) (explored further in relation to later Madhyamaka in Dreyfus and Garfield [2010; 2021]). 15 Huntington (2018, 156) argues that to insist that Nāgārjuna’s sole or overriding concern is to make rational sense is to forget that looking for meaning in these terms is – at least in principle – only one possible methodological approach to interpretation of his writing.

226

Nāgārjuna Huntington characterizes the Root Verses as a literary work of art aimed at inducing “a recognition of the limits of reason and a willingness to surrender to the groundless space below or beyond those limits, which are as well the limits of memory and imagination, of desire and fear” (Huntington 2007, 125). See also de la Vallée Poussin (1933, 59), and more recently Stepien (2019) for an anti-argument reading of Nāgārjuna and of the Root Verses. Such an approach has less difficulty explaining the practical implications of the Root Verses, but while it may arise from and lead to provocative philosophical reflections, it does entail removing the content of the text from the sort of philosophical discourse in which Bhāviveka and Candrakīrti, for instance, placed it  – a discourse in which, as Tillemans (2017, 111) puts it, “the particularity of the Madhyamaka is that it emphasizes philosophical analysis as a method of leading to the quietening of thought.” 16 The Perfection of Wisdom in Eight-Thousand Lines (Aṣṭasāhasrikā Prajñāpāramitā Sūtra), which by some accounts was composed by – or at least prevalent among – the Mahāsaṃghikas of the Andhra region (where Nāgārjuna may also have hailed from), may be especially relevant; according to P. L. Vaidya (1960, xvii) it was the exclusive prajñapāramitā text to which Indian Buddhist philosophers attended. 17 This is a cause for some of the difficulty in determining just how distinctive and exclusive Mahāyāna and non-Mahāyāna Buddhist ideas were. As Walser (2005) observes, under these conditions of communal living, Mahayanists had to present themselves as not so very different from their brethren if they hoped to have resources devoted to the copying and preservation of their texts. 18 As Westerhoff claims, the Root Verses is “not primarily conceived as a treatise with a specific sectarian orientation, but as a fundamental Mahāyāna text” (Westerhoff 2018b, 121). 19 Hayes (1994) argues that transparent equivocation explains the delayed effect of Nāgārjuna on the development of Buddhist philosophical thought. 20 Garfield and Westerhoff (2015) largely explores the rivalry between these two Mahāyāna views, but Westerhoff’s and Shulman’s contributions suggest that the hostility may not be there in Nāgārjuna. There were also important Mahāyāna syncretists. 21 This is so for at least the bulk of the Anglophone scholarship. A notable exception to this is Kalupahana (1986); Shulman (2008) is a more recent instance, while Shulman (2010) identifies the tendency to assimilate Nāgārjuna and Candrakīrti in the contemporary Anglophone scholarship on Madhyamaka, and gives specific reason why we should be cautious about taking Candrakīrti as our authority in understanding Nāgārjuna. 22 Diṅnāga’s Compendium of Pramāṇas (Pramāṇasamuccaya), for instance, opens by praising the Buddha as “the personification of the pramāṇas.” Vasubandhu’s Treasury of Abhidharma, with Commentary (Abhidharmakośabhāṣya), largely concerned with prajñā as the “discernment of dharmas” (AKBh. I.2a), begins by honouring the Buddha as the “teacher of the truth” who “has in an absolute manner destroyed all blindness,” by which, the commentary notes, “the Blessed One is sufficiently designated” (AKBh. I.1). In each case, the Buddha is praised for having those features which are most valued by and essential to the ideas that follow. 23 The translation is that of Siderits and Katsura, leaving out material they supply in square brackets, using ‘(dependent) arising’ for their ‘(dependent) origination,’ and leaving untranslated ‘prapañca,’ which they render as ‘hypostatisation.’ 24 Notice that both Vasubandhu and Diṅnāga, by contrast, connect their selected qualities of the Buddha to his beneficence: “he has drawn the world out of the mire of transmigration” (AKBh. I.1); he “seeks the benefit of [all] living beings” (PS Ι.1.1). They praise other qualities of the Buddha, as well, in particular his beneficence, while Nāgārjuna selects exclusively the Buddha’s teaching of dependent arising for praise. 25 Nor is there reference to svabhāva (translated, variously, as ‘own-nature,’ or ‘self-nature,’ ‘intrinsic nature,’ ‘essence,’ and even ‘own-being’), as that which is lacked, despite the centrality of this to discussions of Madhyamaka; indeed, svabhāva is somewhat less central to the Root Verses generally than its prevalence in Madhyamaka discourse would suggest. Even Chapter 15, which Candrakīrti presents as an analysis of svabhāva, may rather be, as Bhāviveka’s title suggests, “An Analysis of the Existent and Nonexistent.” See Shulman (2007, 148–499) for a sustained argument that this term of art indeed holds less central significance for Nāgārjuna than most contemporary interpretations suggest. 26 In the Sanskrit, this is the first half and not, as in the English translation, the middle portion. 27 MN 38; cf. SN 12.1 (PTS ii.1) 28 See Arnold (2005, 169) for a similar understanding that in the Root Verses, “pratītyasamutpāda must be associated with a novel, not commonly accepted import.” Oetke (2007) concludes that “the

227

Amber D. Carpenter linguistic meaning” of pratītyasamutpāda “and its notional significance is preserved whereas its conceptual import or its intension is completely novel” (Oetke 2007, 27). 29 Alex Wayman argues that “Nāgārjuna, in the matter of the catuṣkoti, is heir to and continuator of teachings in the early Buddhist canon,” the catuṣkoti (tetralemma) in its distinct uses being “found in early Buddhism and later in the Mādhyamika school,” and “well represented in passages of early Buddhism, as preserved in the Pāli canon” (Wayman 1977, 10–11). 30 Translated by Bhikkhu Bodhi in Connected Discourses of the Buddha (Wisdom Publications). 31 That the focus is on interpreting dependent arising is underscored by the MMK’s only explicit reference to a named discourse: “The Instructing of Katyāyana” (MMK 15.7a) refers to SN 12.15 (the Kaccāyanagotta Sutta), the core of which again has the Buddha not “veering towards either of these extremes,” here between existence and non-existence, and again teaching “the Dhamma by the middle: ‘With ignorance as condition, volitional formations [come to be]; with volitional formations as condition, consciousness’ ” (trans. Bodhi 2000). Kalupahana (1992, 161) calls the Root Verses as a whole “a grand commentary on the Buddha’s discourse to Kaccāyana” (viz., SN 12.15); but the Naked Ascetic Kassapa Sutta (SN 12.17), which includes the “Doctrine of the Middle” passage of SN 12.15, prefigures also the tetralemmic approach to expressing it. 32 Ronkin (2005) does some of the careful work necessary for articulating the interpretive moves made from Nikāya to Abhidharma Buddhism. Chapter 5 in particular details the transition from the Nikāya view of dependent arising to the early Abhidharma interpretation of it in terms of causation. 33 This need not imply a well-worked-out theory of causation as such; contrast Kalupahana (1975) with Ronkin (2005) on whether there is or need be a theory of causation implicit in the analysis of conditions. Shulman (2008) argues for the priority of mental conditioning in the early Buddhist notion of dependent arising (rejecting pratītyasamutpāda as a “general principle of causality,” 315), while Meyers (2018) argues that ‘dependent arising’ and ‘causation’ are false friends with only superficial and misleading similarity, but is concerned specifically with what modern notions of causation infelicitously import. 34 Gethin 1992 and 2004 both focus on the practical aspect of Abhidharma classifications. Heim (this volume) offers an anti-metaphysical phenomenological reading of texts from the Pāli Abhidhamma. The advantage of the understanding of Nāgārjuna’s critique which I unfold below is that its target might include not only such an ontologically reticent version of the Abhidharma, but also Abhi­ dharma views at any point along the trajectory from dhamma as classificatory category to dhamma as ultimately existing entity (see Cox 2004; Ronkin 2005, Ch. 2.). Whether it would so readily tell against the sort of ‘philosophical perception’ and meditative phenomenon that Shulman 2014, Ch. 1, attributes to early, sutta Buddhism is less evident. Indeed, one reading of Nāgārjuna may be as arguing that just this is the original Buddha-dharma that the Mahāyāna returns us to, after Abhidharma deviations and codifications. 35 Bastow (1994, 496) calls this “a principle which in general seems to be noncontroversial: that the existence of a relation demands the existence of its terms” and cites Saṃghabhadra as explicitly recognising it. 36 MMK 5.2 uses lakṣaṇa for defining characteristic, accentuating that it is distinct identity which is at stake. 37 MMK 13.3. See note 25 on translating svabhāva. This is, as Siderits and Katsura render it, said in the opponent’s voice, but it thereby confirms that this is the conception at issue (see Shulman 2007, 150n29 for discussion of this point). This translation drops bracketed texts added by Siderits and Katsura, and retains “existing thing” for bhāvaḥ, to underline the connection between this and MMK I.1: The very same thing which “does not arise in any way” is something which necessarily has a proper, individuating nature. 38 Ronkin (2005, Ch. 4) presents Abhidharma dharma theory as concerned centrally with individuation; and this includes reading the associated notion of svabhāva as this arises from the early Abhidharma (Ronkin 2005, 112–22) as that which identifies a discrete individual as what it is, and thus individuates it. 39 All of this can be read in causal terms, or in more broadly explanatory terms. 40 Compare MMK 12.1, “Some say that suffering is self-made, some that it is made by another, some that it is made by both, and some that it is without cause; but it is not correct to think of suffering as an effect” (emphasis added). 41 See Carpenter (2014).

228

Nāgārjuna 42 The focus on relata and relations is even clearer in the titles given these chapters by Buddhapālita and Bhāviveka than those more commonly used in English translations, given by Candrakīrti. For instance, where Candrakīrti has, “What is Prior” (Chapter 10), Bhāviveka offers “Appropriator and Appropriated”; Chapter 18 on “Self ” is, for Bhāviveka, “On Self and Things”; for Chapter 20, instead of Candrakīrti’s “Assemblage,” Bhāviveka offers “Assemblage and Causal Factors.” 43 MMK 2.18–20 reads: It is not right to say that the goer is identical with the act of going; nor again is it right to say that goer and act of going are distinct. If act of going and the goer were identical; then it would also follow that agent and action are one. If, on the other hand, the goer were thought to be distinct from the act of going, then there would be the act of going without a goer, and a goer without an act of going. The analysis of sense-fields (āyatanas, Ch. 3), for instance, is modelled on this. See also the “Hymn to Him Who Has Gone Beyond,” 11; “Hymn to the Unthinkable One,” 11. Shulman (2010) argues that even the first half of this chapter concerns the relation between agent and activity, rather than between locus and activity as Candrakīrti’s interpretation, and the Siderits and Katsura translation, suggest. 44 If Bastow (1994, 494) is correct and the Sarvāstivādins already defined their dharmas by their relations to each other (“the very being of a dharma is interpenetrated by those other dharmas which are causally responsible for its arising”), then Nāgārjuna is here either bringing to the surface an incoherence with their explicit commitments, or else pushing those commitments to their logical conclusions. 45 This is important if Heim is right in interpreting Pāli Abhidhamma, of which Nāgārjuna should certainly have been aware, as primarily doing phenomenology, without making claims about ontology or the nature of existing things (see Heim, Chapter 9 in this volume). 46 Chapter 8, on object and agent, particularly stresses the inadequacy of the real–unreal dichotomy; “intrinsic nature and extrinsic nature, existent and nonexistent – who see these do not see the truth of the Buddha’s teachings” (MMK 15.6). See also 4.4, 5.7 (and 5.2 quoted in what follows). MMK 15.6 is immediately followed by a reference to SN 2.17. 47 Hence, the sense to be found in calling Nāgārjuna a nihilist, in a specific metaphysical sense. It is not that he asserts there is nothing, but he does conclude there are no things. See Westerhoff (2016) for a careful and incisive discussion of the various senses of nihilism, those that Nāgārjuna would be keen to avoid, those with which his opponents (then and now) might reasonably charge him, and the specific sort of nihilism which avoids those objections, and is distinct from the nihilism decried ubiquitously in Buddhism as an ‘extreme view.’ This latter sense, which Westerhoff attributes to Nāgārjuna, is rather similar to the view set out here, but Westerhoff underscores the ‘substantiality’ supposedly implied by svabhāva, while this account focuses on its individuating character. 48 Compare RĀ 38–62, which associates correct understanding of dependent arising with rejecting the existent/non-existent dichotomy (v. 38, 42cd, 48–49, 56). 49 Similarly, to suppose that after the critique of Root Verses I Nāgārjuna owes some alternative account of causation, some explanation of how things do in fact arise, is to miss the main point of the critique. Causation is critiqued to clear space for an alternative understanding of dependent arising. If the result of the critique is an understanding that dependency impugns individuation, then to even suppose there might be an account of causation is to be in the grip of an illusion. 50 Emptiness as an alternative to existence (and non-existence), understood as having a distinct identity, is worked out in Chapter 13 (on suffering) and Chapter 15 (which Candrakīrti considers an analysis of essence (svabhāva), but which Bhāviveka considers “Analysis of the Existent and Nonexistent”); from here, as talk of svabhāva diminishes somewhat, references to emptiness (śūnyatā) – previously sparse – accelerate. Chapters 13 and 15 particularly look like they may be Nāgārjuna’s attempt to correct mistaken notions of emptiness arising among the Mahāyānists. 51 “Dependent arising is emptiness” ’ in 24.18 is immediately followed up in 24.19 with, “There being no dharma whatsoever that is not dependently originated, it follows that there is also no dharma whatsoever that is non-empty” (MMK 24.19). That is, because dependently arising, therefore emptiness. This might be understood as the reason why dharmas are empty (they are empty because dependently arising); and it might equally be articulating how we are to understand that pervasive emptiness (dharmas are empty in the sense that they depend and are not independent in any respect whatsoever, including in respect of their distinct, individuating identity). 52 Compare, from the “Hymn to Him Who Has Gone Beyond the World,” 22: “Dependent Origination has been considered by you to be just emptiness. ‘There is not an independent being’: (this is) your

229

Amber D. Carpenter incomparable lion’s roar” (in Tola and Dragonetti 1985). All quotations from the hymns are taken from this translation. 53 From the “Hymn to the Unthinkable One,” 40 (minor adaptations from Tola and Dragonetti’s translation): What is neither one nor not one, neither both nor not both, without base and not manifest, unthinkable; what cannot be pointed out, what neither arises nor disappears, is neither liable to destruction nor eternal – that, similar to space, is not within the range of words and knowledge. 54 If one asks about Emptiness, the thing itself, the answer is that there is no thing, Emptiness, just as there are generally no things, as discrete individuals with their own definitive identity. But one can use the circle of related concepts through which it gets its meaning to better grasp the alternative to things-in-relation, viz. dependent arising. 55 There is considerable controversy over how exactly to understand the dṛṣṭi (views) which emptiness should get rid of. See discussion following. 56 And likewise, wherever one wishes to indicate the specific emptiness of any particular thing, this necessarily depends on first individuating it: the emptiness of the chariot or of the rūpa dharma derives any specificity it has dependently upon the reference to the chariot or the rūpa dharma. 57 This is a rather deflationary take on the ‘emptiness of emptiness,’ which has been the subject of a great deal of careful and ingenious debate in the contemporary Anglophone literature. Garfield (1994) argues that MMK 24.18 “provide(s) the fulcrum for Candrakīrti’s more explicit characterization of the emptiness of emptiness as an interpretation of Nāgārjuna’s philosophical system – the interpretation that is definitive of the Prāsaṅgika-Mādhyamika school” and offers an interpretation of MMK I that agrees in making “the doctrine of the emptiness of emptiness” the heart of Nāgārjuna’s view (Garfield 1994, 221–22). Siderits (2003b)’s semantic reading also focuses on the emptiness of emptiness, arguing that the upshot is “that the ultimate truth is that there is no ultimate truth – there is only conventional truth” (Siderits 2003b, 11). While Siderits’ view avoids paradox through disambiguation, Priest and Garfield (2002) and Garfield and Priest (forthcoming) embrace the paradox. Such a reading, however, risks being lopsided – recoiling from the fact that words do not correspond to reality to conclude that “there is only conventional truth” (Siderits 2003b, 11) looks very much like the semantic equivalent of discovering nothing exists as an identifiable individual and concluding non-existence. As Nāgārjuna reminds us, the Buddha taught two truths/realities (MMK 24.8), worldly (lokasaṃvṛtisatyaṃ) and ultimate ( paramārthataḥ) truth/reality, the distinction between which must be understood (and failure to grasp the latter prevents one attaining nirvāṇa, 24.10). If worldly truth (lokasaṃvṛtisatyaṃ) is saṃsāra, then this reading of the emptiness of emptiness adopts the extreme position that the claim “there is no distinction whatsoever between saṃsāra and nirvāṇa” (MMK 25.19) is tantamount to the positive assertion that there is only saṃsāra. We should be wary of following Candrakīrti’s lead here and taking the emptiness of emptiness as the core of the Root Verses, particularly when alternatives exist. Nāgārjuna might rather be understood, for instance, as reining in the enthusiasm of his Mahāyāna brethren, by advocating that ubiquitous emptiness be understood merely as a way of designating the essenceless non-individuatedness of reality. See note 51 on 24.19. The opening thirteen verses of the Seventy Verses on Emptiness could be read as arguing in the same way, namely “Because dependently arising, therefore without distinct nature, and that just is to say ‘empty.’ ” 58 Oetke (2007) offers patient analysis of subtly distinguished interpretations of MMK 24.18, including the reading, not pursued here, which takes the verse to identify ‘conceptual designation,’ and not just dependent arising, with emptiness. 59 Presumably that is the force of the ‘non-dualism’ in Siderits 2007’s ‘semantic non-dualism.’ 60 It is not to be taken for granted here that this ‘equally’ is warranted; everything could be dependent, yet some things might be more basic or fundamental than others. Ubiquitous dependency does not itself entail equality of all dependent entities. 61 This difficult verse seems to claim further that nirvāṇa is not a distinct state or mode of being from saṃsāra, and this is hard to make good sense of. Nirvāṇa is frequently understood as cessation, and cessation looks very much like the ‘stillness’ that should arise upon the correct grasp of dependent arising, in the frame verse, but will not otherwise arise. If nirvāṇa is the stilling of prapañca and saṃsāra the proliferation of concepts (non-stilling of prapañca), then there does seem to be a great difference between the two, and the purpose of the Root Verses is to actualise the change from one to the other. If, however, nirvāṇa is not the stillness recommended in the frame verse, then it is also no longer the goal of the Buddhist path, but something quite different he is talking about. It is tempting

230

Nāgārjuna to consider here whether Nāgārjuna might be overstating his case for a certain literary-soteriological effect at 25.19 – and in general, this text deserves much closer consideration of how the literary elements function argumentatively, as seen for instance in Plato’s dialogues or the Introduction to the Bodhisattva Way (Bodhicaryāvatāra). 62 In this respect, it is rather like Plato’s chora in the Timaeus. In just the way Plato would take this to be an incomplete and inadequate explanation of reality and what exists, pratītyasamutpāda could be understood as non-foundational, for it can never adequately explain or ground any particular being. (My thanks to Alexis Pinchard, whose work on Nāgārjuna and Plato suggested the association.) 63 Gethin (1992) brings out the open-ended fecundity of the Abhidharma matrices as a salutary practice. See also Heim (2014, 86): “The Abhidhamma is engaged simultaneously in open-ended possibility and reductive analysis. The Abhidhamma’s work with the Dhamma takes us deeply into ultimate matters ( paramattha) and into the irreducible factors of our experience that cannot be analysed further, even while the relations between them can extend and vary almost infinitely.” 64 We cannot even use language – thinking of Wittgenstein’s example at the beginning of his Philosophical Investigations – if ‘block,’ ‘slab,’ and ‘bring’ are indistinguishable from each other. Diṅnāga recognises this when he argues that any naming or categorising involves conceptualising distinctions (ad PS I.1.3–4), and his comment on PS I.1.3d explicitly indicates that whether language refers to a real individual or not has no bearing on the matter. Even his apoha theory of meaning necessarily individuates – which is why, although the theory may be a useful tool for getting us to cease conceptualising or to navigate everyday reality more skilfully, apoha theory cannot connect concepts and percepts (nor does it pretend to: “a general term does not express particulars,” PS V.2). Even if a word means by excluding other words (PS V.1, 11), these exclusions are just an alternative way of individuating. Plato’s worry in the Theaetetus that the extreme Heraclitean flux-theorists fall into a similar predicament (Tht. 156a–157c) is operating in similar terrain. 65 For instance, “it is correct to call nirvāṇa neither existent nor an absence” (MMK 25.10cd). 66 Nāgārjuna’s deployment of the tetralemma has been the topic of significant scholarly attention, particularly as it is later associated by Candrakīrti with a distinctive interpretation of Nāgārjuna’s thought. Oetke (1996) discusses the use of the tetralemma in Indian philosophy generally; Ruegg (1977) is a classic discussion of the tetralemma in Mahāyāna. Tillemans (1999, 200), after considering a “logically trivial interpretation of the tetralemma” that could take it as “a uniquely ‘therapeutic’ use of language” (Tillemans 1999, 190), argues instead that it is an example of non-classical logic that does not deviate considerably from classical logic. See also Ganeri (2001, Ch. 2) and Priest and Garfield (2002). 67 While Candrakīrti’s dispute with Bhāviveka on this point is the origin of it, this became an especially important marker of a distinct interpretation of Mādhyamaka as Indian Buddhism travelled into Tibet; see Dreyfus and McClintock (2003). 68 Examples include: “An action does not possess conditions; nor is it devoid of conditions” (MMK I.4); “vision does not see; nor does nonvision see” (MMK 4.5); “an agent that is both real and unreal does not bring about an object that is real or one that is unreal” (MMK 8.11); “Dissolution does not at all exist either with or without arising; arising does not at all exist with or without dissolution” (MMK 21.1); and so on. Cf. “ ‘Hymn to the Unthinkable One,” 12–15. 69 Arnold 2010 and 2012 lean heavily on the explanatory priority aspect. 70 Nāgārjuna recognises this explicitly in Root Verses 9 and in the Dispeller. 71 Garfield (1996) notes the similarity to Pyrrhonists in strategy and aims. 72 For a contrary view, see Matilal (1986, Ch. 2), and more recently Mills (2018). 73 Translation, Westerhoff (2010); the choice of ‘thesis’ for pratijñā reflects a narrow rather than capacious interpretation of what is disavowed, and what emptiness eliminates. Westerhoff (2009b) offers detailed consideration of three ways in which one might understand, and the tradition in fact has understood, this verse. A similar claim is found in the Sixty Verses, v. 50. 74 Huntington (1995) and more recently Stepien (2019). 75 The capacious reading would seem to connect back to the ‘auspicious stilling of prapañca’ associated with emptiness in the opening verse, but this itself depends on whether prapañca is any conceptualisation at all, or merely (as Siderits and Katsura translate it) hypostasisation – that is, conceiving of things as ‘having intrinsic nature.’ The restricted view is closely related to understanding Nāgārjuna to be concerned specifically to reject realist semantics: e.g. “Nāgārjuna makes no claims that would be true or false in virtue of real entities with intrinsic natures” (Tillemans 2017, 117). See Ruegg (1983, 2003, Part I) for articulation of this semantic interpretation of Madhyamaka thesislessness. 76 This concern about quietism is raised by Tillemans (2011) and in this volume.

231

Amber D. Carpenter 77 Tillemans (2016, Ch. 5) characterises Candrakīrti (but not Bhāviveka) as such a quietist. Tillemans (2011; 2019) identifies this quietist reading of Candrakīrti as the “typical Prāsaṅgika” perspective (2011, 156; 2019, 640), with Tillemans (2011) exploring an alternative, more philosophically promising version of Prāsaṅgika. Tillemans (2017) presents ‘Nāgārjuna and Early Madhyamaka’ as quietist of a sort, though perhaps not a pernicious sort. It is worth noting that classical Mediterranean skepticism faced this same threat of two-fold quietism: a default conservatism regarding what is commonly accepted (since all grounds for critique have been undermined), and a simultaneous ideal mental calm (ataraxia) consisting in holding and asserting no convictions whatsoever (not even that everything is unknowable). 78 Many other and ingenious ways out have been tried – see for instance Priest, Siderits and Tillemans (2011). 79 These two run throughout the discussion, returning for instance at vv. 44–57, 212–13, 221–23, 230 and 381. Carpenter (2015) discusses in detail the correct understanding of these two goals, and the structure of the ethics in the Ratnāvalī which avoids presenting them as parallel alternative ends. Shulman (2011) offers insightful close comment on what he calls the ‘philosophical’ portions of the Precious Garland, although one might (and the reading set out here does) query the aptness of cutting off certain portions of the text from others as being ‘philosophical.’ 80 One way of reading verses 174–75 might support ‘happiness’ and ‘enlightenment’ as non-intersecting alternatives, though it relies on taking them out of the context of the whole. 81 What follows would be an alternative to the Katsura-Siderits suggestion that Chapter 26 is compatible with what came before by being pegged to conventional reality, while the foregoing chapters concern ultimate reality; the worry is that resolving tensions in this way so separates the two truths that there is no way that one could be used to lead to the other. 82 “In one who first practices high status, definite goodness arises later” (RĀ 3). Translations from the Precious Garland are those of Hopkins (1998), which translates from the Tibetan translation (‘high status’ is roughly worldly well-being, while ‘definite goodness’ is the ultimate enlightened state). Translation from the Sanskrit is available in Tucci (1934; 1936), though the English is quite outdated. 83 Maintaining an ironic engagement, as Siderits (2003a, Ch. 5) calls it, is not easy, but the stuff of bodhisattvas. 84 See Huntington (2018). However, determining whether the end achieved is indeed the same, and the correct one, would seem to require engaging in philosophical reflections on the explicit content of the Root Verses. Only direct philosophical engagement with the Root Verses could suggest that reading it as one reads fiction would be salutary – it does not come with such reading instructions on its surface, or present as fiction at first glance. 85 Analogical reasoning, for instance, at “Hymn to the Unthinkable One,” 4–6; a recap of MMK 1 at “Hymn to the Unthinkable One,” 9–12; “Hymn to Him Who Has Gone Beyond the World,” 4–7, 11, 14 17. 86 The “Hymn According to the Supreme Truth” especially, and much of the “Hymn to the Incomparable One.” 87 Contra Schroeder (2000). 88 Negatively, as in Candrakīrti’s way, doing no more than turning others’ commitments against them, or constructively, as in Bhāviveka’s way, offering suitably contextualized assertions and reasons which are then undercut or their own dependency (their emptiness) made explicit. 89 One of the great debates within the Madhyamaka originating from Nāgārjuna is whether there is ‘something else,’ or whether the recognition of the dependency of one’s thought, as well as any objects of thought, suffices to effect the blissful transformation to stillness.

Bibliography Arnold, Dan. 2005. Buddhists, Brahmins and Beliefs. New York: Columbia University Press. ———. 2010. “Nāgārjuna’s ‘Middle Way’: A  Non-Eliminative Way of Understanding Selflessness.” Revue Internationale de Philosophie 64 (253): 367–95. ———. 2012. “The Deceptive Simplicity of Nāgārjuna’s Arguments Against Motion.” Journal of Indian Philosophy 40: 553–91. Bastow, David. 1994. “The Mahā-Vibāṣā Arguments for the Sarvāstivāda.” Philosophy East and West 44 (3): 489–500.

232

Nāgārjuna Bodhi, Bhikkhu, trans. 2000. The Connected Discourses of the Buddha: A Translation of the Saṃyutta Nikāya. Somerville, MA: Wisdom Publications. Burnyeat, Myles, and Michael Frede. 2015. The Pseudo-Platonic Seventh Letter. Edited by Dominic Scott. Oxford: Oxford University Press. Carpenter, Amber D. 2014. “Ethics of Substance.” Aristotelian Society Supplementary Volume 88 (1): 145–68. ———. 2015. “Aiming at Happiness, Aiming at Ultimate Truth – In Practice.” In Moonpaths: Ethics and Emptiness, edited by the Cowherds, 21–42. New York: Oxford University Press. Cox, Collett. 2004. “From Category to Ontology: The Changing Role of ‘Dharma’ in Sarvāstivāda Abhidharma.” Journal of Indian Philosophy 32 (5–6): 543–97. Dreyfus, Georges B. J., and Jay L. Garfield. 2010. “Madhyamaka and Classical Greek Scepticism.” In Moonshadows: Conventional Truth in Buddhist Philosophy, edited by the Cowherds, 115–30. New York: Oxford University Press. ———. 2021. “Madhyamaka and Skepticism.” In Skepticism in the Indian Philosophical Tradition, edited by Matthew R. Dasti and E. Mills. Leiden: Brill. Dreyfus, Georges B. J., and Sara L. McClintock, eds. 2003. The Svātantrika-Prāsaṅgika Distinction: What Difference Does a Difference Make? Boston: Wisdom Publications. Erb, Felix. 1997. Śūnyatāsaptativṛtti: Candrakīrtis Kommentar zu den “Siebzig Versen über die Leerheit” des Nāgārjuna (Kārikas 1–14); Einleitung, Übersetzung, textkritische Ausgabe des Tibetischen und Indizes. Stuttgart: F. Steiner. Ferraro, Giuseppe. 2013. “Outlines of a Pedagogical Interpretation of Nāgārjuna’s Two Truths Doctrine.” Journal of Indian Philosophy 41 (5): 563–90. ———. 2014. “Grasping Snakes and Touching Elephants: A Rejoinder to Garfield and Siderits.” Journal of Indian Philosophy 42 (4): 451–62. ———. 2017. “Realistic-Antimetaphysical Reading Vs Any Nihilistic Interpretation of Madhyamaka.” Journal of Indian Philosophy 45: 73–98. Ganeri, Jonardon. 2001. Philosophy in Classical India: An Introduction and Analysis. London: Routledge. Garfield, Jay L. 1994. “Dependent Arising and the Emptiness of Emptiness: Why Did Nāgārjuna Start with Causation?” Philosophy East and West 44 (2): 219–50. ———. 1996. “Emptiness and Positionlessness: Do the Mādhyamika Relinquish All Views?” Journal of Indian Philosophy and Religion 1: 1–34. Garfield, Jay L., and Graham Priest. Forthcoming. “Madhyamaka, Ultimate Reality, and Ineffability.” In Reasons and Empty Persons: Mind, Metaphysics, and Morality: Essays in Honor of Mark Siderits, edited by Christian Coseru. New York: Springer. Garfield, Jay L., and Jan Westerhoff, eds. 2015. Madhyamaka and Yogācāra: Allies or Rivals? New York: Oxford University Press. Gethin, Rupert. 1992. “The Mātikās: Memorization, Mindfulness and the List.” In In the Mirror of Memory, edited by Janet Gyatso, 149–72. Albany, NY: SUNY Press. ———. 2004. “He Who Sees Dhamma Sees Dhammas.” Journal of Indian Philosophy 32 (5/6): 513–42. Hayes, Richard P. 1994. “Nāgārjuna’s Appeal.” Journal of Indian Philosophy 22 (4): 299–378. Heim, Maria. 2014. The Forerunner of All Things: Buddhaghosa on Mind, Intention, and Agency. New York: Oxford University Press. Hopkins, Jeffrey. 1998. Nāgārjuna’s Precious Garland: Buddhist Advice for Living and Liberation. Ithaca, NY: Snow Lion Publications. Huntington, C. W. Jr. 1995. “A Way of Reading.” Journal of the International Association of Buddhist Studies 18 (2): 279–308. ———. 2007. “The Nature of the Mādhyamika Trick.” Journal of Indian Philosophy 35 (2): 103–31. ———. 2018. “Nāgārjuna’s Fictional World.” Journal of Indian Philosophy 46 (1): 153–77. Kalupahana, David J. 1975. Causality: The Central Philosophy of Buddhism. Honolulu, HI: University of Hawai’i Press. ———. 1986. Nāgārjuna: The Philosophy of the Middle Way. Albany, NY: SUNY Press. ———. 1992. A History of Buddhist Philosophy: Continuities and Discontinuities. Honolulu, HI: University of Hawai’i Press. Komito, David Ross. 1987. Nāgārjuna’s “Seventy Stanzas”: A Buddhist Psychology of Emptiness. Ithaca, NY: Snow Lion Publications. Lindtner, Christian. 1982. Nagarjuniana: Studies in the Writings and Philosophy of Nāgārjuna. Copenhagen: Akademisk Forlag.

233

Amber D. Carpenter Loizzo, Joseph. 2007. Nāgārjuna’s Reason Sixty (Yuktiṣaṣṭikā) with Chandrakīrti’s Commentary (Yuktiṣaṣṭikāvṛtti). New York: American Institute of Buddhist Studies at Columbia University. Matilal, Bimal K. 1986. Perception: An Essay on Classical Indian Theories of Knowledge. Oxford: Clarendon Press. Meyers, Karin. 2018. “False Friends: Dependent Origination and the Perils of Analogy in Cross-Cultural Philosophy.” Journal of Buddhist Ethics 25: 785–818. Mills, Ethan. 2018. Three Pillars of Skepticism in Classical India: Nāgārjuna, Jayarāśi, Śrī Harṣa. Lanham, MD: Lexington Books. Ñāṇamoli, Bhikkhu, and Bhikkhu Bodhi, trans. 2009. The Middle Length Discourses of the Buddha: A Translation of the Majjhima Nikāya. 4th ed. Somerville, MA: Wisdom Publications. Oetke, Claus. 1996. “Ancient Indian Logic as a Theory of Non-Monotonic Reasoning.” Journal of Indian Philosophy 24 (5): 447–539. ———. 2007. “On MMK 24.18.” Journal of Indian Philosophy 35 (1): 1–32. ———. 2011. “Two Investigations on the Madhyamakakārikās and the Vigrahavyāvartanī.” Journal of Indian Philosophy 39 (3): 245–325. Pind, Ole Holten. 2001. “Why the Vaidalyaprakaraṇa Cannot Be an Authentic Work of Nāgārjuna.” Wiener Zeitschrift für die Kunde Südasiens/Vienna Journal of South Asian Studies 45: 149–72. Priest, Graham, and Jay L. Garfield. 2002. “Nāgārjuna and the Limits of Thought.” In Beyond the Limits of Thought, edited by Graham Priest, 249–70. Oxford: Oxford University Press. Priest, Graham, Mark Siderits, and Tom Tillemans. 2011. “The Two Truths About Truth.” In Moonshadows: Conventional Truth in Buddhist Philosophy, edited by The Cowherds, 131–50. New York: Oxford University Press. Ronkin, Noa. 2005. Early Buddhist Metaphysics: The Making of A Philosophical Tradition. Abingdon, Oxon: RoutledgeCurzon. Ruegg, David Seyfort. 1977. “The Uses of the Four Positions of the ‘Catuṣkoṭi’ and the Problem of the Description of Reality in Mahāyāna Buddhism.” Journal of Indian Philosophy 5 (½): 1–71. ———. 1981. The Literature of the Madhyamaka School in India. Wiesbaden: Otto Harrossowitz. ———. 1983. “On the Thesis and Assertion in the Madhyamaka/ dBu ma.” In Contributions on Tibetan and Buddhist Religion and Philosophy, edited by Ernst Steinkellner and Helmut Tauscher, eds. Wiener Studien zur Tibetologie und Buddhismuskunde 11. Vienna: Arbeitskreise für Tibetische und Buddhistische Studien. ———. 2003. Three Studies in the History of Indian and Tibetan Madhyamaka Philosophy. Wiener Studien zur Tibetologie und Buddhismuskunde. Vienna: Arbeitskreise für Tibetische und Buddhistische Studien. Schroeder, John. 2000. “Nāgārjuna and the Doctrine of ‘Skillful Means’.” Philosophy East and West 50 (4): 559–83. Shulman, Eviatar. 2007. “Creative Ignorance: Nāgārjuna on the Ontological Significance of Consciousness.” Journal of the International Association of Buddhist Studies 30 (1–2): 139–73. ———. 2008. “Early Meanings of Dependent Origination.” Journal of Indian Philosophy 36 (2): 297–317. ———. 2010. “The Commitments of a Madhyamaka Trickster: Innovation in Candrakīrti’s Prasannapadā.” Journal of Indian Philosophy 38 (4): 379–417. ———. 2011. “Ratnāvalī: A Precious Garland of Buddhist Philosophical Systems.” Indo-Iranian Journal 54: 301–29. ———. 2014. Rethinking the Buddha: Early Buddhist Philosophy as Meditative Perception. New York: Columbia University Press. Siderits, Mark. 2003a. Personal Identity and Buddhist Philosophy: Empty Persons. Aldershot, England: Ashgate. ———. 2003b. “On the Soteriological Significance of Emptiness.” Contemporary Buddhism 4 (1): 9–23. ———. 2007. Buddhism as Philosophy: An Introduction. Aldershot, England: Ashgate. Siderits, Mark, and Jay Garfield. 2013. “Defending the Semantic Interpretation: A  Reply to Ferraro.” Journal of Indian Philosophy 41 (6): 655–64. Siderits, Mark, and Shōryū Katsura. 2013. Nāgārjuna’s Middle Way: Mūlamadhyamakakārikā. Somerville, MA: Wisdom Publications. Stepien, Rafal K. 2019. “Abandoning All Views: A Buddhist Critique of Belief.” The Journal of Religion 99 (4): 529–66.

234

Nāgārjuna Tillemans, Tom J. F. 1999. “Is Buddhist Logic Non-classical or Deviant?” In Scripture, Logic, Language: Essays on Dharmakīrti and His Tibetan Successors, 187–205. Translated by John Dunne. Somerville, MA: Wisdom Publications. ———. 2011. “How Far Can a Mādhyamika Buddhist Reform Conventional Truth? Dismal Relativism, Fictionalism, Easy-Easy Truth, and the Alternatives.” In Moonshadows: Conventional Truth in Buddhist Philosophy, edited by The Cowherds, 151–65. Oxford: Oxford University Press. ———. 2016. How Do Mādhyamikas Think? And Other Essays on the Buddhist Philosophy of the Middle. Somerville, MA: Wisdom Publications. ———. 2017. “Philosophical Quietism in Nāgārjuna and Early Madhyamaka.” In The Oxford Handbook of Indian Philosophy, edited by Jonardon Ganeri, 110–32. New York: Oxford University Press. ———. 2019. “Mādhyamikas Playing Bad Hands: The Case of Customary Truth.” Journal of Indian Philosophy 47 (4): 635–44. Tola, Fernando, and Carmen Dragonetti. 1985. “Nāgārjuna’s Catustava.” Journal of Indian Philosophy 13 (1): 1–54. ———. 1987. “Śūnyatāsaptati: The Seventy Kārikās on Voidness (According To The Svavṛtti) Of Nāgārjuna.” Journal of Indian Philosophy 15 (1): 1–55. ———. 1995. On Voidness: A Study on Buddhist Nihilism. New Delhi: Motilal Banarsidass. Tucci, Giuseppe. 1934. “The Ratnāvalī of Nāgārjuna.” Journal of the Royal Asiatic Society of Great Britain and Ireland 66 (2): 307–25. ———. 1936. “The Ratnavali of Nagarjuna.” Journal of the Royal Asiatic Society of Great Britain and Ireland 68 (2–3): 237–52, 423–35. Vaidya, P. L., ed. 1960. Aṣṭasāharśrikā Prajñāpāramitā, with Haribhadra’s Commentary Called Āloka. Darbhanga: The Mithila Institute. de la Vallée Poussin, Louis. 1933. “Réflexions sur le Madhyamaka.” Mélanges chinois et bouddhiques 3: 4–59. Walser, Joseph. 2005. Nāgārjuna in Context: Mahāyāna Buddhism and Early Indian Culture. New York: Columbia University Press. Wayman, Alex. 1977. “Who Understands the Four Alternatives of the Buddhist Texts?” Philosophy East and West 27 (1): 3–21. Westerhoff, Jan. 2009a. Nāgārjuna’s Madhyamaka: A Philosophical Introduction. Oxford: Oxford University Press. ———. 2009b. “The No-Thesis View: Making Sense of Verse 29 of Nāgārjuna’s Vigrahavyāvartanī.” In Pointing at the Moon: Buddhism, Logic, Analytic Philosophy, edited by Mario D’Amato, Jay L. Garfield, and Tom J. F. Tillemans, 26–37. New York: Oxford University Press. ———. 2010. The Dispeller of Disputes: Nagarjuna’s Vigrahavyāvartanī. Oxford: Oxford University Press. ———. 2016. “On the Nihilist Interpretation of Madhyamaka.” Journal of Indian Philosophy 44 (2): 337–76. ———. 2018a. Crushing the Categories (Vaidalyaprakaraṇa). New York: American Institute of Buddhist Studies. ———. 2018b. The Golden Age of Indian Buddhist Philosophy. Oxford: Oxford University Press. Wood, Thomas E. 1994. Nāgārjunian Disputations: A Philosophical Journey Through an Indian Looking-Glass. Honolulu, HI: University of Hawai’i Press.

235

14 ĀRYADEVA Quietism and Buddhist Ethics Tom J. F  . Tillemans

Introduction Āryadeva was a second–third century CE Indian Buddhist author who – along with his teacher Nāgārjuna (c. 150–250 CE) and perhaps an obscure, lesser figure Rāhulabhadra  – is considered a co-founder of the “Philosophy of the Middle” (Madhyamaka). This Great Vehicle (Mahāyāna) philosophy is characterized by its use of dialectics to reject any attribution of intrinsic natures (svabhāva) to things, an intrinsic nature being a type of aseity – what a thing is in itself, completely independently of other factors. And while possessing an intrinsic nature is thought by most Indian thinkers to be a necessary condition for anything to be a fully real entity (bhāva), Āryadeva, Nāgārjuna, and their followers, the Mādhyamikas, held that all things are what they are only dependently. The conclusion: things are “empty” (śūnya) of intrinsic natures; it is impossible for there to be fully real entities and impossible to ascribe any fully real properties to them. Like Nāgārjuna, Āryadeva is thus not a metaphysical realist (Tib. dngos smra ba = Skt. *bhāvavādin); his thought is in sharp contrast with that of his Indian opponents and most of his coreligionists. I will take up his philosophical perspectives on metaphysics and ethics in more detail in what follows.1

Life Following Candrakīrti (sixth century CE), one of Āryadeva’s two Indian commentators, Āryadeva was born in Siṃhaladvīpa as a prince.2 He subsequently traveled from Siṃhaladvīpa – likely Śrī Laṅka  – to South India and became a disciple of the great Mādhyamika thinker Nāgārjuna, the author of “the sixfold corpus of [Madhyamaka] reasoning,” including the Verses on the Philosophy of the Middle (Madhyamakakārikā). After supposedly founding monasteries in South India, he moved to Northern India, to Nālandā.3 All accounts concord on Āryadeva’s skill in defending Buddhism from non-Buddhist thinkers. And indeed, his major work, The Four Hundred Verses (Catuḥśataka), does show a strong familiarity with non-Buddhist doctrines of the time, especially those of Brahmanical schools such as the Sāṃkhya and Vaiśeṣika, which it subjects to trenchant criticism. In a famous debate, supposedly with a Mātṛceta or Durdharṣakāla  – whom the Tibetan historian Tāranātha (1575–1634) implausibly claims to become the second century CE Buddhist poet Aśvaghoṣa (Chimpa and Chattopadhyaya 1970, 236

DOI: 10.4324/9781351030908-23

Āryadeva

131–32) – he is said to have defeated his opponent with dialectics, mantras, and supernatural feats, converting him finally to Buddhism (see Cabezón 2008, 75–78). The Chinese biography goes a step further: Āryadeva ended his days murdered by the disciples of a non-Buddhist philosopher whom he had defeated in debate (Taishō 2048, Robinson 1967, 27–28, Lang 1986, 7). As is frequent in hagiographical versions of Indian debates, we learn about the buildup and the denouement, the magical tricks and the perfidy, but relatively little of philosophical content (Cabezón 2008, 89–90).

Works The philosophies one attributes to a second- or third-century thinker will significantly differ depending upon the texts one regards as authentic. Such questions of textual authenticity are generally not simple matters. Various works, besides the Four Hundred Verses, have been attributed to Āryadeva, especially in the Tibetan and Chinese canons (see Tillemans 1990, 6–7). In fact, there can be little doubt that the name “Āryadeva” was widely applied to authors of works that could not have been written by the second- or third-century author of the Four Hundred Verses.4 We shall, in what follows, examine aspects of the thought of the author of the Four Hundred Verses.5 I consider this treatise – and, to a much lesser degree, the Hundred Verse Treatise – as the best representative of Āryadeva’s thought, although it may well be that at least some other texts that critique non-Buddhist systems, e.g., the Hundred Letter Treatise (akṣaraśataka; trans. Gokhale 1930), are by the same historical individual, too.6

Overview of the Four Hundred Verses Before giving a philosophical analysis of Āryadeva’s ideas, here is a brief overview of the content of the Four Hundred Verses. Chapters one–eight discuss ethical themes, many of which are common to most Buddhist thinkers whether of the Great Vehicle or other traditions. The first four chapters, for example, criticize the four basic illusions (viparyāsa) governing our psyches, viz., our mistakenly taking transitory life as permanent, what is painful as pleasurable, what is dirty as clean, and what is selfless as having a self.7 The next chapters deal, respectively, with the bodhisattva’s practices leading to enlightenment, the elimination of the passions (kleśa) that impede such practices, and the elimination of attachment to the objects of the senses; the eighth chapter deals with the proper practices of Buddhist disciples. It is in the latter half of the Four Hundred Verses that we find the detailed refutations of various ontologies, Buddhist and non-Buddhist, as well as instruction on methodological principles for Madhyamaka thought. Chapters nine–eleven refute, respectively, the reality of permanent entities (nitya), personal selves (ātman), and time (kāla); Chapter twelve refutes “heretical views” (dṛṣṭi) – i.e., inter alia, non-Buddhist views on liberation, scripture, and asceticism – and critiques the importance attributed to high-caste birth; Chapter thirteen refutes Buddhist and non-Buddhist positions on the reality of the sense faculties and their objects (indriyārtha); Chapter fourteen refutes our deep-seated hypostatization of dichotomies, or “extremes” (antagrāha); Chapter fifteen refutes the reality of conditioned things (saṃskṛtārtha); Chapter sixteen, entitled “Cultivating understanding in the master and disciple,” is on Madhyamaka method and logical issues that arise in executing that method. In what follows, I will first seek to better understand Āryadeva’s philosophical orientation and method in the latter half of the Four Hundred Verses, and then take up the first half on ethics and religion; I close with the question of the systemic coherence of the work. 237

Tom J. F. Tillemans

Quietism about ontology Āryadeva’s orientation in the last chapters of the Four Hundred Verses and the Hundred Verse Treatise can, in my opinion, best be described as quietistic. He neither advances nor defends any theses that are ontologically committed; that is to say, that presuppose acceptance of fully real entities (bhāva). This includes positive statements as well as negations and all conjunctions or disjunctions of the two. In having no theses and no debates about them he closely follows Nāgārjuna, who had famously said in verse 50 of his Sixty Verses on Reasoning (Yuktiṣaṣṭikā): Superior individuals have no theses (Tib. phyogs = Skt. pakṣa) and no debates; how could there be any opposing theses for those who have no theses [themselves]?8 Now, it might be objected that quietism is, in one way or another, a relatively widespread theme in Buddhism, Madhyamaka or not. There are, for example, a variety of meditative techniques to arrive at an irenic state beyond conceptual thought, at nirvāṇa, a state of peace, or the “silence of the noble ones” (āryatūṣṇīṃbhāva). One finds some form of quietism in the Pāli canon, Tantra, and Zen. True, but the particularity of the Mādhyamikas, including Āryadeva, is that they emphasize philosophical analysis as a method leading to that quietening of thought, diagnosing rationally and with dialectics where it goes wrong in making ontologically committed truth claims and debating about them. That is why it is of special interest to philosophers.9

Two senses of “no thesis,” “no debates” Understanding what Āryadeva says presupposes clarity on the thought he inherited. This is particularly so when it comes to Āryadeva’s “no thesis” methodology and metaphilosophy. Unfortunately, it seems inadequately recognized in much modern scholarship that Nāgārjuna and Āryadeva both promote two markedly different ways in which a Mādhyamika philosopher does not have theses, that is to say, one way that does not clearly concern the propriety or legitimacy of philosophical debate and another that certainly does. First, the Mādhyamika does not have a thesis because there is not anything fully real, including thesis-statements. And second, the Mādhyamika does not have a thesis because she does not make, or personally endorse, truth claims presupposing anything fully real, all such claims being refutable. It is the second sense of “no thesis” which most clearly leads to quietism, as it embodies the normative stance that one should not be committed to the content of thesis-statements or debate about that content. The second also lends itself to a semantic interpretation of the no-thesis stance, viz., that a Mādhyamika does not accept or debate about any theses whose terms range over fully real entities (bhāva). Although it seems to have been important in Mādhyamikas’ debates with their Brahmanical contemporaries, what I am calling the first sense does not turn on the general stance that all ontologically committing propositions are somehow refutable. Nor is it about the impossibility of making claims via statements with a realist semantics. It turns on the real existence or nonexistence of just one statement, viz., the Mādhyamika’s own fundamental principle that nothing has any intrinsic nature (svabhāva). Let us look at the textual data a bit closer. In Nāgārjuna’s treatise, the Dismissal of Disputes (Vigrahavyāvartanī), a non-Buddhist metaphysical realist – a Naiyāyika or Vaiśeṣika – argues in the first verse that Nāgārjuna’s own statement that nothing has any intrinsic nature is self-refuting, because if true, it would imply that this very thesis-statement itself does not really exist and therefore could not do anything;

238

Āryadeva

it could not mean anything, prove anything, or refute anything. Nāgārjuna replies in verse 29 of the Dismissal: If I had any thesis ( pratijñā) at all, then I would, for that precise reason, have faults. I don’t have any theses and thus I don’t have the fault [of self-refutation of which you accuse me].10 Here, Nāgārjuna recognizes that if he had a fully real thesis-statement to the effect that nothing is fully real, then the real existence of that very statement would falsify what it asserts; but, so his reply goes, he does not have any such real thesis-statement, precisely because nothing is real, and hence does not fall into self-refutation. Indeed, the Dismissal repeatedly argues that reasoning, statements, etc., can function all the while being unreal, just as two illusory or magical projections can do things (or, more accurately, seem to do things) to each other. In chapter sixteen of the Four Hundred Verses, and in a broadly similar passage in the Hundred Verse Treatise, chapter ten (Tucci 1929, 82), Āryadeva takes up that charge of self-refutation. He says in the Four Hundred Verses 16.2: [Objection:] Since the [Mādhyamika] speaker, what he states [and his own words] would also be [unreal], then his saying that [all] is empty would be incoherent. [We reply:] That [i.e., the speaker, assertion and words] which arises in dependence is not existent in any of the three. Instead of appealing to reasoning as simply an efficacious play of illusion, he invokes the basic idea that all things – including speakers, what is stated, and words – are dependent upon each other. However, for our purposes, the objection and reply are the same as in Nāgārjuna’s Dismissal. In both cases, the no-thesis stance invoked is not the generalized, quietistic refusal to make truth claims about the real; it is about the mere existence, or illusoriness, of one statement and its speaker. In short, this is a somewhat specialized debate that should not be mistakenly conflated with Mādhyamika’s generalized rejection of truth claims or the rejection of any semantics that presupposes fully real entities (see Oetke 2003 against Seyfort Ruegg’s interpretation of Dismissal, verse 29, in such a generalized fashion). The undeniable philosophical interest and importance of a rejection of realist semantics does not mean that the specific passages in the Dismissal, like verse 29, are to be read as themselves expressing that rejection. The same critique would apply to semantic readings of Āryadeva’s 16.2. Let us now turn to the second sense of “no thesis.” The stance that one should not make or endorse any truth claims that imply commitment to fully real things, or that have a realist semantics, is at its clearest in Āryadeva’s systematic refutation of each of the “four alternatives,” or the tetralemma (catuṣkoṭi). According to the tetralemma, an entity either is . . .; is not . . .; both is . . . and is not . . .; or neither is . . . nor is not. . . . These four supposedly exhaust the possible formulations of any metaphysical thesis (on the tetralemma, see Seyfort Ruegg 2010, chapter 3). Here is how Āryadeva describes the tetralemma in verse 21 of chapter fourteen in the Four Hundred Verses: Being, nonbeing, [both] being and nonbeing, neither being nor nonbeing: such is the successive method that the wise should always use with regard to identity and the like [i.e., all other theses].

239

Tom J. F. Tillemans

For Āryadeva, each of these four is typically refuted by reductio ad absurdum arguments, so-called prasaṅga; i.e., consequences that would be unacceptable to the proponent of the alternative in question. Āryadeva’s extensive use of absurd consequences is thus taken by many as evidence that he is a “Consequentialist Mādhyamika” (Tib. dbu ma thal ‘gyur pa; Skt. *prāsaṅgikamādhyamika) – roughly, a Mādhyamika who only refutes others and does not positively prove any ontological positions of his own. Consequentialist Madhyamaka is the way to read Āryadeva’s work via the commentary of the sixth century Mādhyamika Candrakīrti, the Catuḥśatakaṭīkā, and it does lead to a refusal of theses, for the would-be intrinsic natures of entities turn out to be so riddled with inconsistencies on analysis that no one can rationally make full-fledged truth claims about them. While Tibetan thinkers, and indeed most modern scholarship, have embraced some variant of this Candrakīrtian interpretation of Āryadeva, it is not the only way to understand his thought and refusal of theses. The other important way is in terms of Yogācāra idealism, by relying upon the Yogācāra’s trademark stance, viz., the three-nature (trisvabhāva) theory.

Āryadeva as a Yogācāra quietist While Consequentialist Madhyamaka rejects any and all intrinsic natures (svabhāva) and all truth claims that presuppose their existence, Āryadeva – if we take a Yogācāra interpretation of his thought – rejects only one type of nature among three: his quietism is thus correspondingly circumscribed. In particular, according to this interpretation, Āryadeva refutes the existence of subject-object dualities and the linguistically conditioned in favor of “mind alone” (cittamātra). More generally, his Madhyamaka dialectic would serve to reject all intrinsic natures conditioned by conceptual thought and language; they are shown to be “imagined natures” ( parikalpitasvabhāva) that are unreal and nonexistent. Āryadeva would leave the ineffable, dependent nature ( paratantrasvabhāva) and the perfect nature (  pariniṣpannasvabhāva) intact as existent.11 It is quite likely that some commentators on Nāgārjuna  – i.e., the fifth–sixth-century Yogācāra commentators, Guṇamati, Gunaśrī or Sthiramati – would have held a three-nature interpretation of Nāgārjuna, although the surviving textual fragments of the first two authors are very meager (see Tillemans 1990, 57–58). In the case of Sthiramati (An hui 安慧; c. 510– 570 CE), his commentary (Taishō 1567, Da cheng zhong guan shi lun 大乘中觀釋論) remains only in a difficult Chinese translation by Wei jing 惟淨. That said, the Yogācāra interpretation of early Madhyamaka was clearly a strong contender in India. Yogācāra-Madhyamaka synthesis continues all the way to the eleventh-century thinker Ratnākaraśānti and beyond to Tibet, its followers being designated under a huge variety of names – “those who profess emptiness of what is other (gzhan stong pa)” and other appellations. Pertinently, Ratnākaraśānti’s self-definition was as a three-nature Mādhyamika (rang bzhin gsum gyi dbu ma pa = *trisvabhāvamādhyamika) (see Luo 2018). The three-nature reading of Madhyamaka is indeed the interpretation promoted by Dharmapāla (Hu fa 護法; c. 530–561 CE), who commented upon the last half of the Four Hundred Verses and whose work does not survive in Sanskrit, was never translated into Tibetan, but only comes to us in Xuanzang’s Chinese translation (Taishō 1571, Guang bai lun shi lun 廣百論釋論). It was Dharmapāla’s commentary on the last chapter of Āryadeva’s Four Hundred Verses that provided his responses in a crucial debate with the Mādhyamika Bhāviveka (c. 500–570 CE), a debate that set Yogācāra and Madhyamaka on significantly different trajectories. (For a translation of key parts of this chapter of Dharmapāla’s commentary, see Hoornaert [2004]).12 See also Kajiyama (1968/1969) on the historical relationship between Bhāv(av)iveka, Sthiramati, and Dharmapāla.) Āryadeva was thus the focal point for a major 240

Āryadeva

parting of the ways in India. Interestingly enough, he was also probably a focus for the Tibetan synthetic movements – his commentator Dharmapāla sometimes figures in their lineages, too. What textual evidence might have prompted Dharmapāla to see Yogācāra in Āryadeva? As Hoornaert pointed out (2004, n. 11), it seems that Dharmapāla was working with a Sanskrit manuscript of the Four Hundred Verses that was significantly different from that on which Candrakīrti commented. Dharmapāla’s Sanskrit source must have contained some key Yogācāra terminology lacking in that used by Candrakīrti. In particular, Dharmapāla’s threenature stance on Āryadeva comes out most clearly in his commentary to Four Hundred Verses 16.23, a verse that can be translated as follows from the Chinese: [Objection:] If things are empty of intrinsic natures (ben xing 本性 = svabhāva), what benefit [do we gain] by seeing that emptiness (kong 空 = śūnyatā)? [Response:] It is the realization of emptiness that eliminates the fetters of false conceptualizations (xu wang fen bie 虛妄分別 = abhūtaparikalpa).13 The term abhūtaparikalpa is a notoriously important Yogācāra term, one that easily evokes the three-nature theory. False conceptualizations are themselves dependent natures ( paratantrasvabhāva), as they arise causally and produce effects, but their content consists of imagined natures ( parikalpitasvabhāva). Indeed, Dharmapāla does go into an exposition of the Yogācāra three-nature theory immediately after this verse. If, however, we translate the same verse as it figures in Tibetan in Candrakīrti’s commentary – the actual Sanskrit of the passage is unfortunately lost – it is apparent that abhūtaparikalpa would not have occurred.14 We simply have the basic and recurrent Buddhist idea that conceptualization obscures or blocks direct seeing of the truth. [Objection:] If entities exist by their intrinsic natures, what benefit do we gain in seeing emptiness? [Response:] Seeing is fettered by conceptualizations (rtog pa = kalpanā); they are what is to be refuted here [in the Four Hundred Verses]. It is also clear that the rest of verse 23 is very different in the manuscript used by Candrakīrti and thus not at all surprising that it led to a very different commentary from that of Dharmapāla. In Candrakīrti’s explanation, there is no clear reference to Yogācāra ideas. The crucial point of Dharmapāla’s interpretation of Āryadeva is to avoid what he sees as a nihilistic notion of ultimate truth ( paramārthasatya).15 While many Mādhyamikas construe the ultimate truth, emptiness (śūnyatā), as a simple negation ( prasajyapratiṣedha) or an absence of any real intrinsic nature whatsoever, Dharmapāla, in his commentary to the Four Hundred Verses 16.23, stresses that it is only the linguistically conditioned imagined natures that are customary truth, while the ineffable dependent and perfect natures are ultimate truths. Moreover, he makes it clear (f. 247c) that those linguistically conditioned natures are the same as universals, or common characteristics (sāmānyalakṣaṇa), while the dependent natures are particulars (svalakṣaṇa). There is thus a deliberate rapprochement with the nominalist philosophy of the Epistemological school of Dignāga (c. 480–540 CE). As Dharmapāla says: Words designate only common characteristics (共相, sāmānyalakṣaṇa), but the individual characteristics (自相, svalakṣaṇa) of things are beyond the scope of words. Individual characteristics are not nonexistent (非無), and common characteristics are not existent (非有). (247c, translation Hoornaert 2004) 241

Tom J. F. Tillemans

The upshot is that ultimate truth for him is certainly not the simple absence of all intrinsic nature, as it is for Mādhyamikas from Bhāviveka and Candrakīrti on; it is the combination of the ineffable dependent nature – i.e., particulars perceived without any linguistic conditioned natures – and the perfect nature. The two philosophical interpretations of Āryadeva thus differ markedly: the Yogācāra sees him as accepting a wordless, conceptless reality composed of particulars, whereas for a Mādhyamika, Āryadeva’s ultimate truth is the denial of any reality whatsoever.

Ethics What are we to make of Āryadeva’s ethics in the Four Hundred Verses? Importantly, if Āryadeva, like Nāgārjuna, rejected Buddhist and non-Buddhist claims that involve ontological commitment, he did not think that he also had to reject or remain quietistic about ethical positions and religious notions of enlightenment, bodhisattva paths, and the like. Indeed, as we saw in our outline of the chapters of the Four Hundred Verses, half of the work argues for basic Buddhist ethical views on impermanence, suffering, purity, selflessness, altruistic intentions, freedom from attachments, and the conduct of bodhisattvas. Āryadeva thus clearly thought that such argumentation did not fall victim to his metaphysical critiques or his quietism about having theses. The result is a complex position that reoccurs, in one way or another, in major Madhya­ maka writing, from Nāgārjuna’s Jewel Garland (Ratnāvalī) to Candrakīrti’s commentaries (vṛtti, ṭīkā) on Nāgārjuna and Āryadeva, Śāntideva’s Engaging in the Practices of a Bodhisattva (Bodhicaryāvatāra), and on to the Path and its Stages (Lam rim) of Tsong kha pa (1357–1419). In these works, we find a more or less nuanced rejection of ontologies coupled with a robust advocacy of the canonical Buddhist ethics and religious practices promoted in the Discourses (Sūtra) traditionally attributed to the Buddha, the Scholastic Teachings (Abhidharma), Code of Discipline (Vinaya), or the Great Vehicle literature laying out the bodhisattva path.

Descriptive versus revisionary The Oxford philosopher P.F. Strawson made a now well-known distinction between descriptive and revisionary metaphysics: “Descriptive metaphysics is content to describe the actual structure of our thought about the world, revisionary metaphysics is concerned to produce a better structure” (Strawson 1959, i). We could extend Strawson’s descriptive-revisionary contrast to ethics. An ethics that advocates supererogation, moral sainthood, extreme selfsacrifice, passionless non-attachment to anything, and in general very substantial betterment to our present views and behavior will be said to be revisionary, while one based more on how people do think and act will go in the direction of description. What would the Madhyamaka of Āryadeva, Nāgārjuna, Śāntideva et al. look like in this light? These philosophers will end up with a significant tension in their overall stance: they adopt a descriptive approach to all matters ontological but a strongly revisionary approach to ethics. Ethics and bodhisattva practices concern customary (saṃvṛti) truths, or customarily real matters about which one can and should argue, while ontology is about ultimately ( paramārtha) real things about which one should neither take a position nor argue; instead of bettering the world’s customary views on what there is, one supposedly limits oneself, faute de mieux, to description of what the world acknowledges (lokaprasiddha) in keeping with its epistemic procedures.16 The ethical discussions in the Four Hundred Verses, however, do not leave the world’s ethical views largely intact, but rather seek to show that the world should very significantly better its views and behavior. In effect, Āryadeva legitimizes the revisionism inherited from canonical non-Madhyamaka Buddhist sources. A  problem of systemic coherence then 242

Āryadeva

arises: if Madhyamaka tells us to do little more than describe what the world acknowledges on customary matters concerning what is and is not – essentially because there is no realist ontology that would provide leverage for anything better – then why the radical revisionism in ethics? That philosophical tension between description and revision is not just limited to Āryadeva: it is what we find in Nāgārjuna, Śāntideva, and many others. Let us look at how it plays out in more detail in the Four Hundred Verses. The picture, in my opinion, may well need some significant rethinking if Madhyamaka Buddhism is to be persuasive to a modern reader.

Misogyny It must be said from the outset that, for many contemporary readers, the ethical chapters in Āryadeva’s Four Hundred Verses will meet with stiff resistance, in part because arguments for the Buddhist goal of passionless detachment are often focused on strong deprecation of women and their sexuality. Misogyny was, no doubt, widespread in the ascetical monastic milieu in which much of the Great Vehicle (Mahāyāna) historically originated.17 It is also apparent in the Pāli Discourses (Sutta) of early Buddhism, like the Numerical Discourses (Aṅguttaranikāya) (see Bhikkhu Bodhi 2012, 683–84), although the most virulent forms seem more frequent in the Great Vehicle (Sponberg 1992, 21). In any case, Āryadeva manifestly inherited monastic asceticism and deprecation of women, bolstering it with questionable dialectics in chapter three of the Four Hundred Verses on the uncleanliness of the female body; Nāgārjuna had done likewise extensively in his Jewel Garland (Ratnāvalī). The origins of Āryadeva’s deprecation of women are thus partially explicable historically. Nonetheless, the disturbance it poses to those who think about his ideas philosophically will not simply go away with basic historical facts about the early Mahāyāna context. There is, in effect, a question as to what is core and what is peripheral in this ethical thought. Minimizers of the disturbance might well try to point out that no less an ethicist than Aristotle defended condemnable institutions such as slavery as just and natural in Politics I.iii–vii, and that his promotion of slavery does not vitiate the many important ideas in the Nicomachean Ethics, nor virtue ethics in general, as it is somehow unessential and peripheral to them. The Four Hundred Verses’ deprecation of women, however, may present a more intractable problem than Aristotle’s acceptance of slavery, as the deludedness of male feelings for women fits very closely with a larger Buddhist discourse on the suffering of saṃsāra and the need for allpervasive renunciation to be free of it. True, Āryadeva and Candrakīrti, in the first chapters of the Four Hundred Verses (see Lang 2003), do not argue only against the cleanliness of the female physique and the value of women generally: they also argue against the illusion of pleasure, against physical beauty, romantic love, love for family members, grief at their death (1.12–13), dedication to work and all other worldly exertions (2.18), and many other key features of human emotional life. Nonetheless, in keeping with much of Buddhist canonical literature, Āryadeva and both his Indian commentators took men’s love and desire for women as a commanding object lesson in deep-rooted, saṃsāric delusion. These are treated as paradigm cases of tainted ignorance and sources of suffering. The uncomfortable truth is that in most Indian Buddhist traditions, and in non-Buddhist texts (e.g., the Upaniṣads) alike, there was a large consensus that powerful contempt and repulsion directed toward women should be cultivated as salvific.18

Intuitions, consensus, moral disagreement Āryadeva’s method in the first eight chapters of the Four Hundred Verses often seems to be an attempt to plumb the depths of people’s ethical intuitions, showing that if we were not in 243

Tom J. F. Tillemans

the grip of self-grasping, were informed, and were consistent with our fundamental intuitions, we would accept canonical Buddhism’s ideas and change much of what we now think. Is that approach, based on an actual or ideal consensus, or perhaps just a convergence of views, likely to justify Āryadeva’s revisionism? Not easily, for justifying canonical Buddhism’s ethics purely or predominantly by means of a metaethics of intuitions and convergence is a very tall order (see Finnigan 2015). While consistency with intuitions can lead to changes of beliefs without appeals to metaphysics, scripture, omniscient buddhas, divine will, or some other source of argumentational leverage, the changes would usually leave the larger more fundamental framework of beliefs untouched. Unacceptable, radical conclusions will instead lead people to say that premises are false or that the reasoning is riddled with non sequiturs. At some point, revisionary ethicists typically need a set of facts – moral and non-moral – to solve disagreements about the truth of counterintuitive ethical statements. Āryadeva, as we shall see, acknowledges that he does, too. Some might well object, however, that Āryadeva could not have thought that Buddhist ethics were true and based on facts because Mādhyamikas, in their metaphysics, do not accept any statements of any customary truths (saṃvṛtisatya) to be true and about facts. Āryadeva, thus, only used argument and rhetorical techniques as expedients, or skill in means (upāyakauśalya), to induce changes of mind for religiously beneficial results but himself rejected truth and facts in all matters customary. This – or some variant upon it – is a quite widespread interpretation of Nāgārjuna’s and Āryadeva’s stance and is not without textual support. Dharmapāla, notably, seems to have read Āryadeva’s ethics largely in this fashion as only upāyakauśalya: For only emptiness is true, all else is not the truth, but is [just] the Tathāgata using skillful means (shan quan fang bian 善權方便 = upāyakauśalya) to preach and publicize [the Dharma] in accordance with the desires of sentient beings. (Tillemans 1990, 92) There was also a recurring, philosophically significant debate in Tibetan Madhyamaka about whether the customary is just, at most, wrongly thought to exist (Tib. blo ’khrul ba’i ngor yod pa) and whether there can ever be right knowledge (tshad ma = pramāṇa) of it (Tillemans 2016, 51–52). Indeed, given the Madhyamaka philosophy of emptiness and its idea that people always mistakenly hypostatize the customary, the question legitimately arises whether any customary truth is true.19

Scripture and the humanly unknowable It is thus a live issue as to what Indo-Tibetan Mādhyamika thinkers could have meant when they spoke of artha (fact, state of affairs) and satya (truth) – whether they only recognized, in the end, ersatz, semblant “facts” and “truths” or, perhaps, deflationary facts and truths, as I and others have argued elsewhere (see Tillemans 2019; Cowherds 2011, chapter 8). We shall henceforth proceed in a neutral fashion staying uncommitted. It would be wrong, however, to say that Āryadeva’s ethical arguments and claims are immune from rational disagreement because he simply never relied on any notion of facts and truth. Mādhyamikas, on the contrary, repeatedly argue that facts and states of affairs have important implications, and, notably, that facts about moral causality and retribution, or karma, ground what one should and should not do. Āryadeva, Candrakīrti, and – to a somewhat lesser degree – Dharmapāla, too, are no exceptions.20 For Āryadeva and successive Mādhyamikas, the issue in grounding ethics on karma is essentially epistemic and not to be conflated with the metaphysical issue of the full-fledged 244

Āryadeva

reality (or lack of it) of facts, moral or otherwise, or the propriety of truth claims about them. The problem is one of human knowledge: the relevant facts – or “facts” – that would eliminate disagreement cannot, in principle, be adequately understood by ordinary rational humans with their own reason alone. Many of those facts are, crucially, suprasensible (atīndriyārtha). Āryadeva, in effect, granted that justification of Buddhist revisionary ethical positions could not be based only on the religiously unschooled intuitions of the common person and the observable or inferable facts recognized by the world. Instead, ethics required belief in the special kinds of customary facts described in Buddhist scriptural statements, i.e., “radically inaccessible” (atyantaparokṣa) facts about the details of moral causality, or karma. For Āryadeva, as for subsequent Mahāyānists, much of the workings of that subtle, retributive causality will be unfathomable by any ordinary human beings in that such causality is humanly unobservable, not inferable from anything observable, and thus only understandable through trustworthy scriptures authored by extraordinary individuals with understanding of the suprasensible (Tillemans 1999, chapters 1–2). How could such subtle (but nonetheless customary) facts make rational individuals abandon their opposition and radically revise their beliefs? Mādhyamikas, especially those who follow Candrakīrti, stress that the customary is that which is, or should be, acknowledged by the world (lokaprasiddha) in keeping with the world’s own epistemic procedures. The obvious question, then, is whether a rational individual could believe that Buddhist scriptures are trustworthy on unfathomable matters if what they say is, in principle, not fathomable by the world itself. Four Hundred Verses 12.5 attempts a response. Āryadeva tells us that when, in an ethical deliberation, there is doubt about the veracity of the Buddha’s descriptions of imperceptible karmic consequences, we should nonetheless believe them to be true because the Buddha has a proven track record of truth in other significant areas; notably, in his teaching on emptiness. This response is crucially important in the history of Buddhism and is still current; it is invoked regularly by Tibetans, including the Fourteenth Dalai Lama. Dharmapāla and Candrakīrti develop an account of scriptural trustworthiness on the basis of 12.5, and, if we accept Tibetan accounts, that Āryadevan account was the same as what we find in the sixth and seventh centuries in Dharmakīrti’s discussions of scripture, i.e., the “triple analysis” (Tib. dpyad pa gsum) of the reliability of scriptures (see Tillemans 1999, chapters 1 and 2; 2000, 78–79). We can go further: Āryadeva would have played a major role in grounding Mahāyāna ethics on the suprasensible, because he was, as far as we can see, the first to have explicitly formulated the dominant Buddhist epistemic method later known as “scripturally based inference” (āgamāśritānumāna). Here is what Āryadeva said in 12.5: When someone entertains doubt concerning the imperceptible things ( parokṣa) taught by the Buddha, he should develop a [rationally founded] belief in these very things on account of emptiness (śūnyatā).21 His point, especially if we follow Candrakīrti’s interpretation, is not that the truth of statements about emptiness (śūnyatā) logically entails the truth of statements about karmic facts. It is rather an epistemic point about justified belief: humans supposedly can, with their own critical acumen, determine that the teachings on emptiness are an example (dṛṣṭānta) of a time when the Buddha got a very important matter perfectly right; therefore, so it is argued, because of his reliability on something essential like emptiness, it is also rational for humans to believe his other statements even when they cannot – on their own – determine the truth-values. There is, in effect, a transfer of credibility from one set of teachings to another. In his commentary to 12.5, Candrakīrti will go further and claim that transferring credibility is in keeping with 245

Tom J. F. Tillemans

worldly practices and intuitions – it is, as he puts it, following “your very own principles alone (svanayenaiva).” In other words, it follows intuitions and epistemic practices that are acknowledged by the world (lokaprasiddha). We regularly believe people who are proven experts. As I have argued in more detail elsewhere (Tillemans 2016, chapter 8), that argument can be seen to be badly flawed upon reflection. Having attested competence and credibility in one subject area does not transfer to another, unless the two are demonstrably closely related. Most of us are probably familiar with the trap of wrongly according extended credibility to people simply because they happen to be qualified in one important field. Those would-be universal experts become unreliable, not because they never correctly understood significant things, but because they overstep their qualifications: their expertise is made to extend to subjects not related to the area in which it has been rightly recognized. The flaw in the appeal to the Buddha’s track record is similar. It is not at all clear that trustworthiness concerning important general principles like emptiness does reasonably transfer to explanations concerning the details of karma in all their specificity and complexity, because the relationship is not clear. While emptiness of intrinsic nature, as a general principle, may be closely linked with the general feature that phenomena arise dependently due to causes and conditions – as Buddhists from Nāgārjuna on have stressed  – knowing that much would hardly suggest that one somehow knows the specific details of what is retribution for what. In the end, I do not think that Āryadeva’s verse 12.5, Candrakīrti’s construal of it, or the closely related Dharmakīrtian triple analysis will provide a publicly debatable proof of scriptural trustworthiness to convince people outside an already committed Buddhist community. In other words, it will not eliminate moral disagreement among rational people of very different views. Indeed, Dharmakīrti and his commentators seem to concede as much. Dharmakīrti was quite skeptical about the probative status of “scripturally based inferences” establishing such things as subtle karmic causality, and he said clearly in his Own Commentary (svavṛtti) to verse 1.217 that reasoning “in this way is not a bona fide inference at all (na khalv evam anumānam anapāyam) as there is no necessary connection between words and states of affairs (anāntarīyakatvād artheṣu śabdānām)” – the mere fact that something is said can never entail its truth.22 Śākyabuddhi (seventh–eighth centuries) and Karṇakagomin (c. 800) give us another angle: “it is not an inference because of factual matters” (vastutas tv ananumānam); it is only one “on account of the thought of people who want to engage [on the spiritual path] ( pravṛttikāmasya puṃso ‘bhiprāyavaśāt)” (see Tillemans 1999, 43 and n. 11). In other terms, it is a type of faith-based reasoning that will be used among religious aspirants, but not one that could be invoked in a neutral, public context of unbelievers.23 Dharmakīrti, to his credit, recognized that limitation. It is unlikely that Āryadeva and Candrakīrti did.

Epistemic humility? A larger problem then arises. Is basing ethics, in part or in whole, on a retributive, moral causality that is nonetheless outside the range of human beings’ own knowledge a coherent, rational standpoint at all? Arguably, for Buddhist metaphysical realists of the non-Madhyamaka schools, like Dharmakīrti, it is not an incoherent position to say that actions have humanly inaccessible dispositions to lead to certain results, dispositions that are as they are irrespective of all beliefs and dissent that rational people might have. Such realist philosophers could, in effect, accept a type of epistemic humility about karmic facts – unknowability does not count against real existence, but only calls for acceptance of our personal cognitive limitations.24 A Mādhyamika who disavows metaphysical realism and professes adherence to what is customarily acknowledged by the world (lokaprasiddha) does not, however, have the ready option 246

Āryadeva

of such humility. Candrakīrti’s argument appealing to “your own principles alone” does not work well, and it is hard to see that any other like it will. For a Mādhyamika, the conclusion, then, is not epistemic humility, but is one that is potentially more disturbing to traditional Buddhists: if subtle karmic cause and effect is not acknowledged as a fact by most of us in the world and not justifiable by the world’s epistemic procedures, much talk of karma-based distinctions in religious ethics risks being simply false. In the end, what the unresolved tension between description and revision in the two sections of the Four Hundred Verses shows us is that a Madhya­maka philosophy that emphasizes quietism about ontology and a strong reliance on worldly intuitions will have to move in a different direction from Buddhist traditions that have heavily relied on the suprasensible and the faith-based, discussable only within protected confines. The price to pay for the philosophy of the second eight chapters of the Four Hundred Verses would seem to be much of the ethics of the first eight. Instead of Āryadeva’s positions, Madhyamaka’s emphasis on the world’s ideas would seem to require a much more secularly oriented, publicly debatable ethics that values what people do think with their own autonomous reason.

Abbreviations CS Catuḥśataka of Āryadeva. CST Catuḥśatakaṭīkā of Candrakīrti. GBLSL Guang bai lun shi lun 廣百論釋論. Taishō 1571. Dharmapāla’s commentary on the last eight chapters of CS. Taishō Taishō Shinshū Daizōkyō. The Tripiṭaka in Chinese. Ed. by J. Takakusu and K. Watanabe. 100 volumes. Tokyo, 1925–1935.

Notes 1 I am indebted to Finnigan (2015; 2018). In Tillemans (2016, chapter 8), I had said that Mādhyamikas try to justify their ethical views to non-believers by appeals to intuitions and what is acknowledged by the world. No doubt they tried, but I doubted they would succeed without major changes to their ethics. Finnigan (2015) shows just how hard it would be to pull their rabbit out of such a hat. 2 The biographical details found in Candrakīrti are amplified in Tāranātha’s History of Buddhism in India (Chimpa and Chattopadhyaya 1970, 123 et seq.) 3 Āryadeva’s being a native of Śrī Laṅka may possibly be corroborated by references in the Ceylonese chronicles Dīpavaṃsa and Mahāvaṃsa to a “Deva” who lived in the second half of the third century at the time when the Indian Vetullavāda sect of Great Vehicle Buddhism was temporarily implanted in Śrī Laṅka (Seyfort Ruegg 1981, 50). His Ceylonese origins are to be taken with some caution, as Chinese sources have him being born into a South Indian brahmin family. It is far from clear who, if anyone, was a student of Āryadeva, but some Tibetan sources speak of Rāhulabhadra, who wrote an important Praise of the Perfection of Wisdom (Prajñāpāramitāstotra); others claim that Rāhulabhadra was the teacher of Nāgārjuna. See Seyfort Ruegg (1981, 54–56) on Rāhulabhadra. 4 A case in point is the Destruction of Errors about Madhyamaka (*Madhyamakabhramaghāta), assigned to an Āryadeva by Tibetans: this text copiously borrows from the Verses on the Heart of Madhyamaka (Madhyamakahṛdayakārikā) and Torch of Dialectics (Tarkajvālā) of Bhāviveka, a celebrated Mādhyamika who lived in the sixth century (500–570 CE). Another such case is the Compendium on the Essence of Knowledge (Jñānasārasamuccaya), a text which the Tibetan canon ascribes to Āryadeva, but which gives the fourfold presentation of Buddhist doctrine typical of the doxographical (siddhānta) literature, a genre which considerably post-dates the third century (trans. Mimaki 1976). Less clear are the Hand Treatise (Hastavālaprakaraṇa) and its Commentary (Hastavālaprakaraṇavṛtti), both of which the Tibetan canon attributes to Āryadeva, although the Chinese canon most likely has it right in attributing them to the fifth century author Dignāga (trans. Thomas and Ui 1918). Finally, the Chinese canon has five works attributed to Āryadeva, the longest two of which are known as “Hundred Verse Treatises”: Taishō 1570, the Extensive Hundred Verse Treatise (Guang bai lun 廣百論), is the last half of the Four

247

Tom J. F. Tillemans Hundred Verses (Catuḥśataka); Taishō 1569, the Hundred Verse Treatise (Bai lun 百論 = *Śata(ka) śāstra; trans. Tucci 1929), bears many thematic affinities with the Four Hundred Verses but is nonetheless a different work and not simply an extract of the Four Hundred. The Hundred Verse Treatise became a key text of the Chinese Madhyamaka school known as the “Three Treatise” (San lun 三論) tradition (May 1979; May and Mimaki 1979). If the identification of the author Qing mu 青目 with Āryadeva should be correct, then the author of the Four Hundred Verses may have also written a commentary on Nāgārjuna’s Verses on the Philosophy of the Middle (May 1979, 481; Tillemans 1990, 7, fn. 17). 5 The Four Hundred Verses has been translated into English in Lang (1986), who uses the Sanskrit fragments and the Tibetan; Sonam (2008) translates the Four Hundred Verses and rGyal tshab rje’s commentary from the Tibetan; chapters nine–sixteen are translated into Italian in Tucci (1925) based on the Sanskrit fragments, Tibetan, and Chinese (Taishō 1570). Lang (2003) introduces and translates the first four chapters along with Candrakīrti’s commentary; May (1980, 1981a, 1981b, 1982, 1984) provide a French translation of chapter nine of the Four Hundred Verses and Candrakīrti’s commentary; Tillemans (1990) contains a translation of chapters twelve and thirteen, along with Candrakīrti’s and Dharmapāla’s commentaries; Johnson (2012) translates chapter fifteen and Candrakīrti’s commentary. An edition of the Sanskrit fragments and their corresponding Tibetan translation is found in Suzuki (1994), following upon the works of Haraprasād Shāstrī (1914) and Bhattacharya (1931). 6 Tibetan tradition identifies major Mādhyamika authors – i.e., Nāgārjuna, Āryadeva, and Candrakīrti – with later tantrikas having the same name. In particular, the Mādhyamika Nāgārjuna is identified with tantric Nāgārjunapāda, also known as Ārya Nāgārjuna (c. seventh century CE); Āryadeva is said to be the disciple of Ārya Nāgārjuna and the author of the Lamp that Integrates the Practices (Caryāmelāpakapradīpa; trans. Wedemeyer 2007), which furthers the Ārya (noble) tradition of exegesis on the Tantra of the Esoteric Community (Guhyasamājatantra); he is also said to be the author of the Treatise on the Purity of Mind (Cittaviśuddhiprakaraṇa) connected with Ārya Nāgārjuna’s Five Stages (Pañcakrama). To be clear, however, one cannot follow the Tibetan tradition in attributing tantric texts to the same Mādhyamika Āryadeva; hence, one badly needs a differentiation between at least two individuals, Āryadeva I being the Mādhyamika and Āryadeva II a much later tantric author (Mimaki 1987; Tillemans 1990, 6–7). 7 For the canonical schema of four illusions (viparyāsa), see Abhidharmakośabhāṣya ad V.8, translation La Vallée Poussin (1971, tome IV, p. 21). 8 Tibetan text in Lindtner (1990, 114). 9 There is a tension here in that while Āryadeva practices analysis and uses argumentation adroitly, he insists that the Buddhist’s invoking emptiness (śūnyatā) is somehow not simply a move in debate: “The tathāgatas [= buddhas] did not state this [idea that things do not have real intrinsic natures] for the sake of debate (vāda). Nonetheless, it burns up rival assertions, just as fire [burns up] fuel” (Four Hundred Verses 12, verse 15; see Tillemans 1990, vol. 1, 128, vol. 2, 46–47). 10 Vigrahavyāvartanī 29, cited in Candrakīrti’s Prasannapadā (ed. La Vallée Poussin 1970, 16.7–10). 11 For more on the three natures, see Chapter 16 and Chapter 22 in this volume. 12 There is also a rather significantly flawed translation in Keenan (1997). 13 Taishō 1571, 246a: 若法本性空 見空有何德 虛妄分別縛 證空見能除. Our translation. 14 The Tibetan is: gal te rang bzhin gyis dngos yod//stong mthong yon tan ci zhig yod//rtog pas mthong ba ‘ching ba ste//de ni ‘di ru dgag par bya//. If the Sanskrit text translated into Tibetan had used abhūtaparikalpa, one would have expected something more along the lines of the usual equivalent, yang dag pa ma yin pa’i kun tu rtog pa, instead of simply rtog pa. See Gadjin Nagao, Index to the Mahāyāna-sūtrālaṃkāra. Part One. Tokyo, 1958, s.v. abhūtaparikalpa. 15 Interestingly, the broad lines of the polemic against a purely negative ultimate truth are also found in Tibetan Mādhyamika writers like Go rams pa bSod nams seng ge (1429–1489), but without the allegiance to Yogācāra’s three-nature theory. 16 Such is the position of Candrakīrti in his commentary to Four Hundred Verses, especially in his arguments in chapter thirteen against the logicians’ misplaced attempts to revise worldly epistemic standards (see Tillemans 1990, vol. 1, 176–79). 17 Contrary to how Mahāyāna has been typically depicted, it was not a predominantly egalitarian lay movement (see Boucher 2008, 50–52; Nattier 2003, 96–100). 18 See Cabezón (2017, chapter 4). While there certainly are tantras, as well as a few sūtras (e.g., the Mahāyānopadeśa or Vimalakīrtinirdeśa), that challenge misogyny, that challenge was not the norm in Indian Buddhism. It should be noted, in this vein, that Theravāda and some Mahāyāna Buddhists practiced an extreme contemplation on disfigured female bodies in morgues and cemeteries as a way of cultivating renunciation (see Cabezón 2017, 227 et seq.; Wilson 1996). The practice is known as

248

Āryadeva aśubhabhāvanā, “meditation on the hideous,” and is for the edification of men to impress upon them the actual repulsive nature of the female physique. It continues to this day. Buddhist canonical literature has no remotely similar place for an aśubhabhāvanā on the male physique. 19 Cowherds (2011, vi, 15, emphasis in original) sets out a book-length treatment of four questions, one of which is formulated as “What is true about customary truth?” or “What kind of truth is it, and what sort of truth theory would best fit it?” 20 For example, Candrakīrti uses the phrase “concerning suprasensible states of affairs” (atīndriyeṣv artheṣu) in connection with Āryadeva’s 12.5. Dharmapāla ad 12.5, in Chinese, seems to gloss only emptiness (śūnyatā) as yi 義 (real fact, state of affairs, artha), preferring shi 事 (phenomenal facts, things) for all that is customary, including imperceptible things, i.e., shen shi 深事, parokṣa. 21 Translation and text of 12.5 and Candrakīrti’s commentary in Tillemans (1990, vol. 1, 120 and vol. 2, 17–19). For Dharmapāla, Tillemans (1990, vol. 1, 91–92 and vol. 2, 132–33). 22 Tillemans (1999, 42). Cf. Dharmakīrti’s Svavṛtti to 1.318 (ed. Gnoli 1960, 167): na kvacid askhalita iti sarvaṃ tathā/vyabhicāradarśanāt, “It is not the case that when one is unmistaken on something, all the rest is similarly [unmistaken], for we see that this [implication] is deviant.” 23 Cf. Pramāṇavārttika 4.106 (Tillemans 2000, 148–50). Human understanding of subtle karmic facts and the like invariably depends on “acceptance” (abhyupagama), i.e., on having previously accepted (iṣṭa) scriptures that talk about them as reliable, because otherwise “no [human] deliberation will occur at all (na cintaiva pravartate).” 24 Cf. what is known as “Kantian humility,” i.e., human beings’ inevitable incapacity to understand things in themselves (Ding an sich). See Langton (1998; 2004).

References Bhattacharya, Vidhushekhara. 1931. The Catuḥśataka of Āryadeva: Sanskrit and Tibetan Texts with Copious Extracts from the Commentary of Candrakīrti [Chapters 8–16.] Part 2. Calcutta: Visvabharati Bookshop. Bhikkhu Bodhi. 2012. The Numerical Discourses of the Buddha: A Translation of the Aṅguttara Nikāya. Somerville, MA: Wisdom Publications. Boucher, Daniel. 2008. Bodhisattvas of the Forest and the Formation of Mahāyāna: A Study and Translation of the Rāṣṭrapālaparipṛcchāsūtra. Honolulu: University of Hawaii Press. Cabezón, José I. 2008. “Buddhist Narratives of the Great Debates.” Argumentation 22: 71–92. ———. 2017. Sexuality in Classical South Asian Buddhism. Studies in Indian and Tibetan Buddhism. Somerville, MA: Wisdom Publications. Chimpa, Lama and Alaka Chattopadhyaya. 1970. Tāranātha’s History of Buddhism in India. Simla: Indian Institute of Advanced Study. Cowherds (= Georges Dreyfus, Bronwyn Finnigan, Jay L. Garfield, Guy Newland, Graham Priest, Mark Siderits, Koji Tanaka, Sonam Thakchoe, Tom Tillemans, and Jan Westerhoff). 2011. Moonshadows: Conventional Truth in Buddhist Philosophy. New York: Oxford University Press. Finnigan, Bronwyn. 2015. “Madhyamaka Buddhist Meta-ethics: The Justificatory Grounds of Moral Judgments.” Philosophy East and West 65 (3): 765–85. ———. 2018. “Madhyamaka Ethics.” In The Oxford Handbook of Buddhist Ethics, edited by Daniel Cozort and James Mark Shields, 162–83. Oxford: Oxford University Press. Gnoli, Raniero. 1960. The Pramāṇavārttikam of Dharmakīrti: The First Chapter with the Autocommentary: Text and Critical Notes. Serie Orientale Roma 23. Rome: Istituto Italiano per il Medio ed Estremo Oriente. Gokhale, Vasudev Vishwanath and Budhiruci. 1930. Akṣara-Çatakam: The Hundred Letters, a Madhyamaka Text. After Chinese and Tibetan Materials Translated by Vasudev Gokhale. Materialen zur Kunde Buddhismus 14. Heidelberg: O. Harrassowitz. Haraprasād Shāstrī. 1914. “Catuḥśatika of Ārya Deva.” Memoirs of the Asiatic Society of Bengal 3 (8): 449–514 [Sanskrit fragments of CS and CST]. Hoornaert, Paul. 2004. “The Dharmapāla-Bhāvaviveka Debate as Presented in Dharmapāla’s Commentary to Catuḥśataka XVI.23.” Kanazawadaigaku bungakubu ronshū, Kōdōkagaku tetsugaku hen 24 – Studies and Essays, Behavioural Sciences and Philosophy, Faculty of Letters. Kanazawa University 24: 119–49. Johnson, Dennis A. 2012. “Refuting the Conditioned  – The Saṃskṛtārthapratiṣedha of Candrakīrti’s Catuḥśatakaṭīkā: Introduction, Translation and Summary Together with a Critical Edition of the Tibetan Translation.” MA Thesis, University of Vienna. Accessed January  18, 2019. https://othes. univie.ac.at/20891/ [Translation of chapter 15 of CS and CST].

249

Tom J. F. Tillemans Kajiyama, Yūichi. 1968/1969. “Bhāvaviveka, Sthiramati and Dharmapāla.” Wiener Zeitschrift für die Kunde Süd- und Ostasiens und Archiv für Indische Philosophie 12–13: 193–203. Keenan, John P.  1997. Dharmapāla’s Yogācāra Critique of Bhāvaviveka’s Mādhyamika Explanation of Emptiness. The Tenth Chapter of Ta-ch’eng Kuang Pai-lun Shih Commenting on Āryadeva’s Catuḥśataka Chapter Sixteen. Lewiston: Edwin Mellen. La Vallée Poussin, Louis de. 1970. Mūlamadhyamakakārikās (Mādhyamikasūtras) de Nāgārjuna, avec le commentaire de Candrakīrti. First published in St. Petersburg, Bibliotheca Buddhica IV, 1903–1913. Reprinted Osnabrück: Biblio Verlag. ———. 1971. L’Abhidharmakośa de Vasubandhu: traduction et annotations, 6 vols. Paris: Geuthner. Louvain: J.-B. Istas, 1923–31. New edition prepared by É. Lamotte, Mélanges chinois et bouddhiques 16. Brussels: Institut Belge des Hautes Études Chinoises. Lang, Karen C. 1986. Āryadeva’s Catuḥśataka: On the Bodhisattva’s Cultivation of Merit and Knowledge. Indiske Studier VII. Copenhagen: Akademisk Forlag [Translation of CS]. ———. 2003. Four Illusions:. Candrakīrti’s Advice to Travelers on the Bodhisattva Path. Translation and Introduction by Karen C. Lang. New York: Oxford University Press [Translation of chapters 1–4 of CS and CST]. Langton, Rae. 1998. Kantian Humility: Our Ignorance of Things in Themselves. New York: Oxford University Press. ———. 2004. “Elusive Knowledge of Things in Themselves.” Australasian Journal of Philosophy 82 (1): 129–36. Lindtner, Christian. 1982 [1990]. Nāgārjuniana. Studies in the Writings and Philosophy of Nāgārjuna. Indiske Studier 4. Copenhagen: Akademisk Forlag. Reprinted Delhi: Motilal Banarsidass. Luo, Hong. 2018. “Is Ratnākaraśānti a gZhan stong pa?” Journal of Indian Philosophy 46: 577–619. May, Jacques. 1979. “Chūgan.” In Hōbōgirin, Dictionnaire encyclopédique du bouddhisme d’après les sources chinoises et japonaises V, edited by Sylvain Lévi, J. Takakusu, Paul Demiéville, Hubert Durt et al., 470–93. Paris: Librairie d’Amérique et d’Orient. Tokyo: Maison franco-japonaise. ———. 1980. “Āryadeva et Candrakīrti sur la permanence (I).” In Indianisme et Bouddhisme, Mélanges offerts à Mgr Étienne Lamotte, 214–32. Louvain: Publications de l’Institut orientaliste de Louvain [French translation of chapter 9 of CS and CST]. ———. 1981a. “Āryadeva et Candrakīrti sur la permanence (II).” Bulletin de l’École française d’Extrême-Orient 69: 75–96. ———. 1981b.“Āryadeva et Candrakīrti sur la permanence (III).” Études Asiatiques 35 (2): 47–76. ———. 1982. “Āryadeva et Candrakīrti sur la permanence (IV).” Études de Lettres, University of Lau­ sanne 3: 45–76. ———. 1984. “Āryadeva et Candrakīrti sur la permanence (V).” In Acta Indologica, Studies of Mysticism in Honor of the 1150th Anniversary of Kōbō Daishi’s Nirvāṇam, 115–44. Narita: Naritasan Shinshoji. May, Jacques, and Katsumi Mimaki. 1979. “Chūdō.” In Hōbōgirin, Dictionnaire encyclopédique du bouddhisme d’après les sources chinoises et japonaises V, edited by Sylvain Lévi, J. Takakusu, Paul Demiéville, Hubert Durt et  al., 456–70. Paris: Librairie d’Amérique et d’Orient. Tokyo: Maison franco-japonaise. Mimaki, Katsumi. 1976. La Réfutation bouddhique de la permanence des choses (sthirasiddhidūṣaṇa) et la preuve de la momentanéité des choses (kṣaṇabhaṅgasiddhi). Publications de l’Institut de Civilisation Indienne, fascicule 41. Paris: Institut de Civilisation Indienne. ———. 1987. “Āryadeva.” In The Encyclopedia of Religion, edited by Mircea Eliade et  al., vol. 1, 431a–432a. New York: Macmillan Publishing. Nattier, Jan. 2003. A Few Good Men. Honolulu: University of Hawaii Press. Oetke, Claus. 2003. “Some Remarks on Theses and Philosophical Positions in Early Madhyamaka.” Journal of Indian Philosophy 31: 449–78. Robinson, Richard. 1967 [1976]. Early Mādhyamika in India and China. Madison: University of Wisconsin Press. Reprinted by Motilal Banarsidass, New Delhi. Seyfort Ruegg, David. 1981. The Literature of the Madhyamaka School of Philosophy in India. Wiesbaden: O. Harrassowitz. ———. 2010. The Buddhist Philosophy of the Middle: Essays on Indian and Tibetan Madhyamaka. Studies in Indian and Tibetan Buddhism. Somerville MA: Wisdom Publications.

250

Āryadeva Sonam, Ruth. 2008. Āryadeva’s Four Hundred Stanzas on the Middle Way, with Commentary by Gyeltsap. Translated by Ruth Sonam with additional commentary by Geshe Sonam Rinchen. Ithaca, NY: Snow Lion Publications [Translation of CS, CST and rGyal tshab rje’s commentary]. Sponberg, Alan. 1992. “Attitudes toward Women and the Feminine in Early Buddhism.” In Buddhism, Sexuality, and Gender, edited by José I. Cabezón, 3–36. Albany: State University of New York Press. Strawson, Peter F. 1959. Individuals: An Essay in Descriptive Metaphysics. London: Methuen. Suzuki, Kōshin. 1994. Sanskrit Fragments and Tibetan Translation of Candrakīrti’s Bodhisattvayogācāracatuḥśataka­ṭīkā. Tokyo: Sankibo [Sanskrit fragments of CS and CST with their corresponding Tibetan]. Thomas, Frederick W., and Hakuju Ui. 1918. “The Hand Treatise, A Work of Āryadeva.” Journal of the Royal Asiatic Society: 267–310. Tillemans, Tom J. F. 1990. Materials for the Study of Āryadeva, Dharmapāla and Candrakīrti. The Catuḥśataka of Āryadeva, Chapters XII and XIII, with the Commentaries of Dharmapāla and Candrakīrti: Introduction, Translation, Sanskrit, Tibetan and Chinese Texts, Notes. 2 vols. Wiener Studien zur Tibetologie und Buddhismuskunde 24.1 & 24.2. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. Reprinted in one volume by Motilal Banarsidass Publishing Co., New Delhi 2008 [Translation of chapters 12 and 13 of CS, CST and GBLSL]. ———. 1999. Scripture, Logic, Language: Essays on Dharmakīrti and His Tibetan Successors. Studies in Indian and Tibetan Buddhism. Somerville, MA: Wisdom Publications. ———. 2000. Dharmakīrti’s Pramāṇavārttika: An Annotated Translation of the Fourth Chapter ( parārthānumāna), vol. 1 (k. 1–148). Vienna: Verlag der Österreichischen Akademie der Wissenschaften. ———. 2016. How Do Mādhyamikas Think? And Other Essays on the Buddhist Philosophy of the Middle. Studies in Indian and Tibetan Buddhism. Somerville, MA: Wisdom Publications. ———. 2019. “Deflating the Two Images and the Two Truths: Bons baisers du Tibet.” In Wilfrid Sellars and Buddhist Philosophy: Freedom from Foundations, edited by Jay L. Garfield, 80–96. Routledge Studies in American Philosophy. New York and London: Routledge. Tucci, Giuseppe. 1925. “Studi Mahāyānici. La versione cinese del Catuḥçataka di Āryadeva confrontata col testo sanscrito e la traduzione tibetana.” Rivista degli Studi Orientali 10: 521–90 [Italian translation of chapters 9–16 of CS and some extracts of GBLSL]. ———. 1929. Pre-Diṅnāga Buddhist Texts on Logic from Chinese Sources. Gaekwad’s Oriental Series, no. XLIX. Baroda: The Oriental Institute [Translation of Āryadeva’s Śatakaśāstra (Bai lun), Taishō 1569]. Wedemeyer, Christian K. 2007. Āryadeva’s Lamp that Integrates the Practices (Caryāmelāpakapradīpa): The Gradual Path of Vajrayāna Buddhism According to the Esoteric Community Noble Tradition. American Institute of Buddhist Studies. New York: Columbia University Press. Wilson, Liz. 1996. Charming Cadavers: Horrific Figurations of the Feminine in Indian Buddhist Hagiographic Literature. Chicago: University of Chicago Press.

251

15 ˙ GA ASAN Great Systematizer of Yogācāra Thought Jowita Kramer

The Indian scholar-monk Asaṅga (Chinese: Wuzhuo 無著; Tibetan: Thogs med) was one of the most influential masters of the Yogācāra tradition. Almost all information concerning his life and works available to us today derives from Chinese and Tibetan sources whose historicity is uncertain. Most modern scholars date Asaṅga to the fourth or fifth century.1

Asan˙ga’s Life The four main traditional sources that provide details of Asaṅga’s life and works are the Biography of the Master of the Doctrine Vasubandhu (Posoupandou fashi zhuan 婆藪槃豆法師傳, T 2049) by Paramārtha (Zhendi 眞諦; 499–569), the Records of the Western Regions (Datang xiyuji 大唐西域記, T 2087) by Xuanzang 玄奘 (600[?]–664), the History of the Doctrine (Chos ’byung) by Bu ston (1290–1364), and the History of the Doctrine in India (rGya gar chos ’byung) by Tāranātha (1575–1634).2 The oldest of these texts is the Biography of the Master of the Doctrine Vasubandhu (sixth century), according to which Asaṅga – who is said by Paramārtha to be Vasubandhu’s brother – was born in Puruṣapura (located in Gandhāra; present-day Peshawar) as the eldest son of a brahmin and ordained as a monk in the Sarvāstivāda tradition early in his life. Unsatisfied with the teachings, he is said to have visited Tuṣita heaven by means of his spiritual power, where he received instruction in the Mahāyāna sūtras from the future Buddha Maitreya. Moreover, Paramārtha explains that Maitreya also came down to earth in order to teach the Foundation of the Yoga Practitioners (Yogācārabhūmi) to Asaṅga, who subsequently composed several commentaries on the Mahāyāna sūtras Maitreya had explained to him. Paramārtha also states that Vasubandhu later became a Mahāyāna proponent himself after Asaṅga, under the pretext of an illness, had Vasubandhu visit him and attributed the cause of his illness to Vasubandhu speaking poorly of the Mahāyāna.3 The second important biographical source, Xuanzang’s travelogue Records of the Western Regions, additionally states that Asaṅga was active in Ayodhyā and that his disciple was Buddhasiṃha. Furthermore, Asaṅga is said to have received not only the Foundation of the Yoga Practitioners but also the Ornament of Mahāyāna Sūtras (Mahāyānasūtrālaṃkāra, Dasheng zhuangyanjing lun 大乘莊嚴經論) and the Distinguishing the Middle from the Extremes (Madhyāntavibhāga, Bian zhong bian lun song 辯中邊論頌) from Maitreya. Unlike Paramārtha, Xuanzang assigns Asaṅga to the Mahīśāsaka school, not the Sarvāstivāda.4 Some 252

DOI: 10.4324/9781351030908-24

Asan˙ga

modern scholars follow Xuanzang’s assumption and state that Asaṅga’s Abhidharma scriptures are based on many Mahīśāsaka ideas. However, others argue that there is not enough ground to determine which school affiliation Asaṅga had before he became a Mahāyāna teacher.5 In contrast to Chinese sources, Asaṅga’s Tibetan biographies do not speak of a conversion from Mainstream to Mahāyāna Buddhism. Bu ston reports that Asaṅga was the son of a woman named Gsal ba’i tshul khrims (reconstructed as “Prasannaśīlā” in Obermiller [1932, 37]), who wanted to prevent the downfall of the Buddhist doctrine through the birth of three extraordinary sons. According to Bu ston’s report, Asaṅga was the son of this woman and a “warrior” (kṣatriya) father, while his (half-)brother Vasubandhu is said to have been the son of a brahmin. Further, Bu ston reports that Asaṅga, after many years of unsuccessful practice, which he began at a cave in Ri bya rkang (reconstructed as “Kukkuṭapādaparvata” in Obermiller [1932, 37]) and through which he intended to propitiate Maitreya, met a sick dog whose body was partially eaten by worms. In order to free the dog from the worms, Asaṅga is said to have carved a slice of meat from his own body. But at the very moment he wanted to use his tongue to remove the worms from the dog’s body and put them on his amputated flesh, the dog disappeared and Maitreya appeared in his place. Maitreya is said thereafter to have taken Asaṅga to Tuṣita heaven, where he instructed him in various Mahāyāna sūtras, the Foundation of the Yoga Practitioners and the “five treatises of Maitreya” (i.e., according to the Tibetan tradition, the Ornament of Clear Realization [Abhisamāyālaṃkāra], the Ornament of Mahāyāna Sūtras, the Distinguishing the Middle from the Extremes, the Distinguishing the Factors from Their True Nature [Dharmadharmatāvibhāga] and the Distinguishing the Germ of the [Three] Jewels [Ratnagotravibhāga]).6 On another occasion, Bu ston states that Asaṅga composed, after his return to earth, the Foundation of the Yoga Practitioners, the Compendium of the Abhidharma (Abhidharmasamuccaya), the Summary of the Mahāyāna (Mahāyānasaṃgraha),7 the Determination of Reality (*Tattvaviniścaya, a commentary on the Ornament of Clear Realization; Delhey 2019, 75), and commentaries on the Distinguishing the Germ of the [Three] Jewels and the Sūtra of the Explanation of the Profound Meaning (Saṃdhinirmocanasūtra). He moreover is said to have written down the “five treatises of Maitreya.”8 Bu ston also tells the story of how Asaṅga converted Vasubandhu to the Mahāyāna, after the latter had sarcastically remarked: “Alas! Though Asaṅga meditated for twelve years in the forest, instead of attaining success in his meditation, he has composed a work useless in sense but heavy enough to be an elephant’s load.”9 In his History of the Doctrine in India, Tāranātha presents a similar story of the life and works of Asaṅga, but adorns it with more details, relating, for example, that Asaṅga established a residence called “Dharmāvikuravihāra” in “Veluvana,” Magadha. There he is said to have written down the five works of Maitreya, as well as composed the Foundation of the Yoga Practitioners, the Summary of the Mahāyāna, the Compendium of the Abhidharma, and a commentary on the Ornament of Clear Realization. Tāranātha further reports that under King Gambhīrapakṣa, there was a monastic gathering in the monastery of “Uṣmapuri” in the city of “Sagari,” “near the west,” where Asaṅga gave teachings. The king is also said to have invited Asaṅga to his court to test his abilities through philosophical questions about the Sūtra of the Perfection of Wisdom (Prajñāpāramitāsūtra). The king is said to have been so pleased with Asaṅga’s answers that he had 25 Mahāyāna monasteries built. During this time, Asaṅga is also supposed to have converted to the Mahāyāna his brother Vasubandhu, who had previously commented upon Asaṅga’s Foundation of the Yoga Practitioners with the sarcastic remark mentioned already by Bu ston.10 253

Jowita Kramer

Furthermore, Tāranātha ascribes to Asaṅga the establishment of monastic centers in “Kṛṣṇarāja,” “Urgyana,” and Magadha. When Asaṅga taught near Ayodhyā, he and his followers are said to have been attacked by an army of fighters. After Asaṅga instructed everyone to continue with contemplative practice and not to be disturbed by the attackers, they withdrew without harming anyone. Tāranātha also states that Asaṅga taught not disrespecting any of the Buddhist traditions, for which reason the adherents of mainstream Buddhism (śrāvaka) respected him and studied the sūtras and the Abhidharma with him. In the later part of his life, Asaṅga is said to have spent twelve years in Nālandā and to have died in *Rājagṛha, where his disciples are said to have erected a shrine (caitya) with his relics.11 In other parts of his report, Tāranātha names *Buddhadasa and *Dharmadasa as Asaṅga’s disciples.12 The questions of whether Maitreya(nātha) (Chinese Cizun 慈尊; Tibetan Byams pa, Byams pa mngon po) is to be considered a historical person or a deity, and whether the works Asaṅga received from him should be regarded as his or as the works of a separate author, have sparked much discussion among contemporary scholars. The traditional Chinese and Tibetan accounts do not agree on this question, either.13

Asan˙ga’s Works In total, the following nineteen works are traditionally associated with Asaṅga as primary author: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15.

Foundation of the Yoga Practitioners (Yogācārabhūmi)14 Summary of the Mahāyāna (Mahāyānasaṃgraha)15 Compendium of the Abhidharma (Abhidharmasamuccaya)16 Treatise on Propagating the Teaching of the Noble One (Xianyang shengjiao lun 顯揚聖 教論)17 Treatise on Six Aspects of Meditation Instruction (Liumen jiaoshou xiding lun 六門教授 習定論)18 Treatise on the Perfection of Wisdom That Cuts [Like] a Vajra (Jingang bore lun 金剛般 若論, *Vajracchedikā Prajñāpāramitāśāstra)19 Treatise Conforming with the Middle (Shunzhong lun 順中論)20 Seventy Stanzas on the Perfection of Wisdom in Three Hundred Stanzas (Triśatikāyāḥ Prajñāpāramitāyāḥ Kārikāsaptatiḥ)21 Hymn on the Qualities Based on the Body [Consisting in the True Reality] of Phenomena (Chos kyi sku la gnas pa’i yon tan la bstod pa, *Dharmakāyāśrayāsāmānyaguṇastotra),22 Meditation Manual of the Perfection of Wisdom (Shes rab kyi pha rol tu phyin ma’i sgrub thabs, *Prajñāpāramitāsādhana),23 Noble Maitreya Meditation Manual (’Phags pa byams pa’i sgrub thabs, *Āryamaitreyasādhana),24 Noble Commentary on [the Sūtra of] the Explanation of the Profound Meaning (’Phags pa dgongs pa nges par ’grel pa’i rnam par bshad pa, *Āryasaṃdhinirmocanabhāṣya),25 Commentary on the Recollection of the Buddha (Sangs rgyas rjes su dran pa’i ’grel pa, *Buddhānusmṛtivṛtti) Commentary on the Recollection of the Doctrine (Chos rjes su dran pa’i ’grel pa, *Dharmānusmṛtivṛtti) Commentary on the Recollection of the Monastic Community (Dge ’dun rjes su dran pa’i bshad pa, *Saṃghānusmṛtivyākhyā)26 254

Asan˙ga

16. Commentary on Distinguishing the Germ of the [Three] Jewels (*Ratnagotravibhāgavyākhyā)27 17. Instruction Called Lamp of Absorption (bSam gtan gyi sgron ma zhes bya ba’i man ngag, *Dhyānadīpanāmopadeśa)28 18. A (lost) commentary on the Ornament of Clear Realization (Abhisamayālaṃkāra)29 19. Commentary on the Ornament of Mahāyāna Sūtras (Mahāyānasūtrālaṃkārabhāṣya)30 In trying to evaluate Asaṅga’s contribution to Buddhist philosophy, we first face the difficulty that it remains controversial which works can actually be attributed to him as their original author. Second, with regard to many Yogācāra texts, as with earlier Buddhist texts in general, there is often no consensus regarding the chronological order. This means that for many texts and passages, we can know neither exactly who wrote them nor which of the texts’ passages that contain a particular teaching are to be considered the earlier. Among the nineteen texts ascribed to Asaṅga, only the Compendium of the Abhidharma, the Summary of the Mahāyāna, and the Treatise on Propagating the Teaching of the Noble One are widely accepted by modern scholars to have been composed by a single author called Asaṅga. In contrast, most modern scholars take the Foundation of the Yoga Practitioners to be not the work of an individual author but rather a compilation that grew into its present shape over the course of many decades or even centuries. If Asaṅga is to be associated with the Foundation of the Yoga Practitioners, then it is only as its compiler, not as its original author. As Lambert Schmithausen notes, the text appears to be a compilation of different strands and materials that have been transmitted by an early Yogācāra community, which seems not to have been entirely homogeneous (Schmithausen 2014, 11). The Treatise on Propagating the Teaching of the Noble One31 is a summary of the teachings of the Foundation of the Yoga Practitioners, which is partially reproduced verbatim (see Delhey 2019, 78). The work is only available in Xuanzang’s Chinese translation. Most modern scholars accept the attribution of this work to Asaṅga, some of them assuming that only the verses were composed by this author, while regarding the prose as the work of Vasubandhu (Delhey 2019, 74). The text seems to be closely related to the Compendium of the Abhidharma (von Rospatt 1995, 85). The authorship of the other fifteen works ascribed to Asaṅga remains doubtful.32 While the short commentary on the Sūtra of the Explanation of the Profound Meaning might indeed have been written by him, it can be stated with some certainty that the same Asaṅga who composed the three Yogācāra works just mentioned is not the author of the two tantric ritualistic meditation manuals (sādhana) ascribed to Asaṅga (Delhey 2019, 79). It is also not very probable that this Asaṅga is the composer of the commentaries on the Distinguishing the Germ of the [Three] Jewels and the Root Verses on the Middle Way (Mūlamadhyamakakārikā) (i.e., the Treatise Conforming with the Middle), which shows no relation to Yogācāra thought, as well as the two works related to the perfection of wisdom ( prajñāpāramitā).33 The two meditation manuals, the Instruction Called Lamp of Absorption and Treatise on Six Aspects of Meditation Instruction, as well as the three works on the recollection of the Buddha, the doctrine (dharma), and the monastic community (saṃgha) have not been investigated thoroughly enough in modern scholarship to come to any final conclusions regarding their authorship. The hymn (stotra) on the “body consisting in the true reality of phenomena” (dharmakāya) was probably extracted from the Ornament of Mahāyāna Sūtras, while the commentary Illumination of the Ornament of Clear Realization (Abhisamayālaṃkārālokā) is not available to us and therefore nothing can be said about its composer (Delhey 2019, 78). The authorship of the Commentary on the Ornament of Mahāyāna Sūtras is commonly ascribed to Vasubandhu. The Yogācāra tradition is characterized by a number of specific concepts. The one that has probably attracted most attention and is subject to controversial debates until the present day is the doctrine of “nothing but presentation” (vijñaptimātra). According to this tenet, the 255

Jowita Kramer

whole world is nothing but mind (cittamātra), and external objects exist not as material entities but only as mental manifestations. The second central theory of the Yogācāras is the doctrine of eight forms of mind, including the five sense perceptions – already propagated in the early Buddhist schools – plus the mental perception (manovijñāna), the defiled notion [of “I”] (kliṣṭaṃ manas), and the store mind (ālayavijñāna) (Schmithausen 2018, 264). The third doctrine for which the Yogācāras are famed is the set of three natures (svabhāva) or characteristics (lakṣaṇa) describing three aspects of all phenomena.34 The first refers to the assumption that we superimpose a false, conceptualized ( parikalpita) nature on the phenomena we perceive, which does not reflect their real existence. The second aspect of all entities is that their nature is “dependent on causes and conditions” ( paratantra). The final aspect describes the perfect ( pariniṣpanna) nature of phenomena, which means that, in actuality, manifestations are completely free of the conceptualizations ascribed to them. Another doctrine that was developed in Yogācāra circles is the concept of the three bodies (kāya) of a Buddha. The body through which a Buddha makes his appearance in the world is called the “manifestation body” (nirmāṇakāya), while the one through which he teaches advanced Bodhisattvas in the pure lands is referred to as his “enjoyment body” (saṃbhogakāya). The true reality realized by a Buddha is his “essential body” (svābhāvikaḥ kāyaḥ) or his “body consisting in the true reality of phenomena” (dharmakāya).35 The concept of the three bodies of the Buddha seems to appear for the first time in the Ornament of Mahāyāna Sūtras.36 The question of whether the three Yogācāra tenets of “nothing but presentation” (vijñaptimātratā), “store mind” (ālayavijñāna), and the three “natures” (svabhāva) are inextricably connected has given rise to a number of controversial discussions. Some scholars assume that they were developed interdependently and complemented each other from the very beginning.37 In contrast, Lambert Schmithausen tried to show in several of his publications (1987, 2014, 2018) that such a logical inseparability of the three concepts is neither necessary nor reflected in early Yogācāra sources. In most of the earlier Yogācāra texts, the three tenets are treated in separate contexts, and it is often the case that one is even completely absent (Sch­ mithausen 2018, 263). According to Schmithausen (2018, 266), the concepts of the store mind and the three natures originate from two different sections of the Foundation of the Yoga Practitioners. There, the tenet of “nothing but presentation” (vijñaptimātratā) is not mentioned except for a quotation from the Sūtra of the Explanation of the Profound Meaning, which was almost certainly added at the very end of the compilation process of the Foundation of the Yoga Practitioners (Schmithausen 2014, 31) – and may indeed be the text in which the idea of “nothing but presentation” (vijñaptimātra) occurs for the first time.38 Notably, the store mind (ālayavijñāna) is not mentioned in what we might consider the “Māhāyanistic” parts of the Foundation of the Yoga Practitioners, such as the section called Stage of the Bodhisattva (Bodhisattvabhūmi), but only in the “traditional” parts of the text (Schmithausen 1987, 98ff.). In the treatises Ornament of Mahāyāna Sūtras, Distinguishing the Middle from the Extremes, and Distinguishing the Factors from Their True Nature, which are commonly ascribed to Maitreya(nātha), the store mind is not mentioned at all. The doctrine of “mind only” occurs in the Ornament of Mahāyāna Sūtras in the form of cittamātra (not vijñaptimātra). If we assume that Asaṅga is not the original author of the Foundation of the Yoga Practicioners (and, if at all, only its compiler) or of the three works ascribed to Maitreya(nātha), but that he is the composer of the Compendium of the Abhidharma, Summary of the Mahāyāna, and the Treatise on Propagating the Teaching of the Noble One, we have to conclude that Asaṅga is not the originator of the central Yogācāra concepts of the store mind, mind only, the three natures, or the three bodies of a Buddha. Rather, what we can credit to this scholar is the combination of these doctrines into a coherent system and a nuanced elaboration of that system. 256

Asan˙ga

The Compendium of the Abhidharma and the Summary of the Mahāyāna are considered Asaṅga’s key works by modern scholars (Delhey 2019, 77). Both these texts summarize and systematize Yogācāra thought and have had a lasting impact on the development of Mahāyāna Buddhism in India and beyond. According to Dan Martin (2002, 344), the Compendium of the Abhidharma played an important role in the early transmission of Abhidharma teachings in Tibet and was only later superseded by the Commentary on the Treasury of Abhidharma (Abhidharmakośabhāṣya). Compared to the (three) works ascribed to Maitreya (mentioned previously in this chapter), both the Compendium of the Abhidharma and Summary of the Mahāyāna show a much closer affiliation with Abhidharmic thought (Schmithausen 1987, 100).

The Compendium of the Abhidharma The Compendium of the Abhidharma is extant in some Sanskrit fragments, as well as in Tibetan and Chinese translation.39 One of the notable aspects of the Compendium of the Abhidharma is that it is a combination of the Abhidharmic thought of mainstream Buddhism with innovative Yogācāra ideas. Drawing mainly on old sūtras and elements of earlier Yogācāra texts, such as the Foundation of the Yoga Practitioners, Asaṅga thus creates a distinctive system of Yogācāra Abhidharma (Kritzer 1999, 13). While the Compendium of the Abhidharma is closely related to the Foundation of the Yoga Practitioners, it further systematizes and develops its ideas (Kritzer 1999, vii).40 The theory of the store mind (ālayavijñāna) and the scheme of three natures (svabhāva) appear already well established in the Compendium of the Abhidharma. At the same time, as noted by Bayer (2010, 19), the store mind (ālayavijñāna) is mentioned only six times in the Compendium of the Abhidharma, and only in its first part called Compendium of Characteristics (Lakṣaṇasamuccaya). Remarkably, the theory of “nothing but presentation” (vijñaptimātratā) is not mentioned at all (Schmithausen 2018, 269ff.). The doctrine of the three bodies (kāya) of a Buddha, which is absent from the Foundation of the Yoga Practitioners (Deleanu 2006, 227 n. 171), is mentioned in the Compendium of the Abhidharma (Bayer 2010, 22). Another doctrine that we do not find in the Foundation of the Yoga Practitioners but in the Compendium of the Abhidharma is the concept of the “non-abiding nirvāṇa” (apratiṣṭhitanirvāṇa), which refers to the Buddha’s quality of neither being fixed in nirvāṇa nor in cyclic existence (saṃsara).41 Many parts of the Compendium of the Abhidharma seem to follow the traditional doctrine of mainstream Buddhism in affirming the nonexistence of a permanent self (ātman) and do not propagate or presuppose the idea of an ultimate unreality of all factors (dharma) or the nonexistence of external material objects expressed in other Yogācāra texts (Schmithausen 1987, 190). The concept of emptiness (śūnyatā) is also not explained in terms of the nonexistence of the factors of existence, but instead as their lack of a substantial and unchangeable self. The true essence (tathatā) of these factors, which is equated with this emptiness, is therefore – according to the Compendium of the Abhidharma – not their total unreality but the fact that they do not have such a self (Schmithausen 1972, 155ff.). Notably, the Compendium of the Abhidharma also explains the three natures (lakṣaṇa = svabhāva) in accordance with this understanding. The conceptualized (vikalpita) nature is understood as the wrong ascription of the idea of a self to the five constituents (skandha), the eighteen elements (dhātu), and the twelve bases (āyatana). The dependent ( paratantra) nature is explained to be the constituents, elements, and bases themselves, whereas the true nature (dharmatālakṣaṇa = pariniṣpannaḥ svabhāvaḥ) is defined as the absence of a self in the constituents, elements, and bases.42 As mentioned by Bayer (2010, 20), the term “mind only” (cittamātra) is only used in the Compendium of the Abhidharma in connection with mental factors, such as feeling (vedanā), etc., which are assumed to be associated with only the mind (citta) and not with a self (ātman). 257

Jowita Kramer

At the same time, some scholars have pointed out that there are a few statements to be found in the Compendium of the Abhidharma that seem to indicate the unreality of matter.43 Schmithausen (1972, 158) refers to a passage stating that atoms ( paramāṇu) are “without a body” (niḥśarīra); that is, they do not exist as particles of a specific size. Another passage of the text to suggest an idealistic point of view appears within an explanation of the meditative practice concerned with realizing that the previously visualized object is nothing but one’s own mind (Bayer 2010, 21). Although the doctrine of the store mind is mentioned only a few times in the Compendium of the Abhidharma, it appears in a well-developed shape in the text. However, it is important to note that it only has the function of a substratum, which ensures that the mind can continue uninterruptedly (Bayer 2010, 19), and of a container of all mental seeds or the collection of these seeds (sarvabījaka). We do not find any statements in the Compendium of the Abhi­ dharma, which would define the store mind as having the role of cognizing an object.44 One of the remarkable parts of the Compendium of the Abhidharma is the discussion of “Vai­pulya,” a synonym of the Mahāyāna (Skilling 2013, 95), defined by Asaṅga as the “canon of the Bodhisattvas” (*bodhisattvapiṭaka).45 In this context, the author gives reasons why some are afraid of these teachings and lists twenty-eight “wrong views” that result in the fact that some, even though they follow the Mahāyāna, are not liberated by it. Asaṅga also explains some central statements of the Mahāyāna, such as “all dharmas are without substance” and all dharmas are “unarisen, undestroyed, at peace from the beginning” (Skilling 2013, 88). Another notable section is the Determination of the Talk (Sāṃkathyaviniścaya), which discusses various aspects of “talk” (sāṃkathya), one of them being the topic of “speech” (vāda).46 In addition to forms of everyday talk, such as “common talk” (sarvalokavacana) and “rumor” ( pravāda), Asaṅga deals with various components of a (formal) debate in this passage. He lists the possible locations (adhikaraṇa), including the royal court (rājakula); the subjects (adhiṣṭhāna), including the various forms of proofs (sādhana); and the qualities of the debater, such as the perfection of eloquence (vākkaraṇasampad) and self-confidence (vaiśāradya). Another remarkable category mentioned in this context is “defeat” (nigraha), which can involve ceding (tyāga) a debate in recognition of one’s own faults or because of manifesting anger or pride. The text enumerates nine kinds of defects (doṣa) in a discussion, such as it being confusing (ākula), unmeasured (amita), meaningless (anarthaka), or disconnected (aprabaddha). According to Asaṅga, there is also the possibility of withdrawing (niḥsaraṇa) from the debate – for instance, after having realized the opponent’s inability, the assembly’s inferiority, or one’s own incompetence. Finally, Asaṅga states that one should only engage in a debate in order to develop one’s own knowledge, and not only for the purpose of debating with others. In this context, he quotes the Abhidharmasūtra, which lists twelve reasons why a Bodhisattva should not get involved in a discussion with others. It is noteworthy that in contrast to Abhidharmic works such as the Commentary on the Treasury of Abhidharma, the Compendium of the Abhidharma does not usually deal with topics in a controversial way. While the Commentary on the Treasury of Abhidharma often presents a number of different and opposing positions on certain concepts and doctrines, the Compendium of the Abhidharma is, doctrinally, a rather uniform piece of work (Bayer 2010, 16). However, at the same time, scholars have also noticed features that indicate its compilatory character.47

The Summary of the Mahāyāna The Summary of the Mahāyāna, which is lost in its Sanskrit original and only available in one Tibetan (D 4048) and four Chinese translations (T 1592, 1593, 1594, 1596), can be regarded as one of the most important works of the Yogācāra school.48 The text consists of ten chapters 258

Asan˙ga

devoted to the exposition of all the main Yogācāra teachings. Compared to earlier Yogācāra treatises, such as the Ornament of Mahāyāna Sūtras, the text is written in a clearer and more systematic style and appears to represent a more mature stage of Yogācāra philosophy. As the main sources of the Summary of the Mahāyāna, scholars have identified the Sūtra of the Explanation of the Profound Meaning, the Abhidharmasūtra, the Sūtra(s) of Perfect Wisdom (Prajñāpāramitāsūtra), the Foundation of Yoga Practicioners, the Ornament of Mahāyāna Sūtras, and the Distinguishing the Middle from the Extremes.49 Notably, some arrive at the conclusion that the Summary of the Mahāyāna was probably composed before the Compendium of the Abhidharma,50 even though the Compendium of the Abhidharma seems closer to the teachings of traditional Abhidharma than the Summary of the Mahāyāna. The ten chapters of the Summary of the Mahāyāna deal respectively with the store mind (ālayavijñāna), the three natures (svabhāva), the theory of “nothing but presentation” (vijñaptimātratā), the six perfections ( pāramitā), the ten stages (bhūmi), the three kinds of training (śikṣā) – i.e., in highest discipline (adhiśīla), highest mind (adhicitta) and highest knowledge (adhiprajñā) – the “transformation of the basis” (āśrayaparāvṛtti), and Buddhahood (including the three Buddha bodies [kāya] and four awarenesses [  jñāna]).51 As pointed out by Lambert Schmithausen, the arrangement of the first three chapters of the Summary of the Mahāyāna follows the structure of chapters 5–8 in the Sūtra of the Explanation of the Profound Meaning. However, in contrast to earlier Yogācāra works, in the Summary of the Mahāyāna, the three Yogācāra concepts are explicitly integrated into a unified system (Schmithausen 2014, 27–40). At the same time, the concept of “nothing but presentation” (vijñaptimātratā) is fully established and becomes the central idea of spiritual practice and philosophical theory (Schmithausen 2018, 278). While in earlier texts, the store mind is taught independently from the theory that all matter is only a mental manifestation, the Summary of the Mahāyāna seems to be the first Yogācāra text that fully combines the store mind (ālayavijñāna) with the “mind only” system and regards the store mind as the basis of all phenomena of the whole world. The original meaning of the store mind, namely the mind that lies hidden in the material sense faculties, is modified in the Summary of the Mahāyāna to denote a mind to which all conditioned factors stick (ālīyante), in the sense that they are bound to it as their cause (Schmithausen 1987, 65). The first chapter of the text is one of the most important, providing a detailed explanation of the store mind that should be considered as the most elaborate of the older Yogācāra texts (Schmithausen 1985, 140). The store mind, on the one hand, is still ascribed its traditional function of “biological appropriation” and of keeping the corporeal basis of personal existence alive. On the other hand, it is also presented as a container of all seeds (bīja) and is called a “maturation mind” (vipākavijñāna) (Schmithausen 1987, 61ff. and 100ff.). Asaṅga, moreover, describes the store mind as being constituted by three kinds of imprints (vāsanā), namely the imprints of (1) expressions (abhilāpa); (2) the (false) view of the self (ātmadṛṣṭi); and (3) the limbs of existence (bhavāṅga) (Lamotte 1938, 22 [58]). As already noted by Erich Frauwallner (2010, 333), especially the first two classes of imprints reflect the Yogācāra’s emphasis on two specific doctrines: the first stresses the relevance of conventional expressions for the human conceptualization of phenomena; the second is related to the Yogācāra theory of the defiled notion [of “I”] (kliṣṭaṃ manas), which is also rooted (in the form of its seed) in the store mind. The theory of a supramundane (lokottara) insight, arising from the imprint (vāsanā) of hearing (śruta) the Buddhist teachings, is also proposed in the first chapter of the Summary of the Mahāyāna.52 According to this teaching, this type of insight cannot arise from seeds originating in the store mind. Instead, it is assumed that the first moment of the supramundane 259

Jowita Kramer

insight arises from an imprint left in a person’s mental continuum when hearing the Buddhist teachings, which is considered the outflow (niṣyanda) of the true reality (dharmadhātu). Asaṅga further explains that the imprint of the supramundane insight exists together with the store mind but nonetheless is not part of it (Schmithausen 2014, 585ff.), a state he compares to milk mixed with water. It cannot be the store mind itself as it actually forms its antidote. Frauwallner (2010, 332) points out that Asaṅga’s approach to the state of liberation and Buddhahood is more psychological than what we find in the works attributed to Maitreya(nātha): while the Maitreya treatises seem more concerned with the description of the highest reality (and what we might call its ontological status), Asaṅga aims at explaining the psychological complex underlying the process of liberation. In chapter three, and partly also chapter two, Asaṅga treats the topic of “nothing but presentation” (vijñaptimātratā) in detail. In this context, we find a list of arguments by means of which Asaṅga tries to prove that external objects are not existent.53 This list may be considered the earliest transmitted enumeration of this kind (Schmithausen 2014, 626). According to the first argument, the assumption that the same object appears in various ways to the different classes of beings – such as humans, animals or gods – can only be reasonable if the referents of their perception are not really existent. Second, Asaṅga states that all objects perceived in dreams, images visualized in meditation, past and future objects (appearing in thoughts), and reflections in mirrors do not presuppose a real (material) object as their basis. The third argument explains that if external objects were really existent, everyone would effortlessly see true reality when perceiving them. Furthermore, nonconceptual awareness would never be experienced as completely objectless. The following two arguments point out that if external objects really existed, it would be impossible for spiritually advanced practitioners to transform material things (for instance, earth into gold) or make material things appear according to their will. Finally, Asaṅga states that nonconceptual awareness could not be experienced with really existing external objects (Schmithausen 2014, 627ff.). As Schmithausen points out (2014, 628), most of these considerations regarding the nonexistence of external matter seem to be associated with (spiritual) experience in the first place and less with the solution of theoretical problems. Another two sets of proofs to be found in the Summary of the Mahāyāna that are of particular importance are those concerning the store mind (ālayavijñāna) and the defiled notion [of “I”] (kliṣṭaṃ manas) (Lamotte 1938, 5ff. [7] and 13–22 [29–57]). While the existence of the store mind has already been proved in (at least) one earlier Yogācāra source (namely the Foundation of Yoga Practitioners), the list of six proofs provided in the Summary of the Mahāyāna does not seem to rely directly on the nine proofs of the Foundation of the Yoga Practitioners and includes a number of innovative aspects.54 At the same time, the Summary of the Mahāyāna seems to be the earliest transmitted Yogācāra source that explicitly presents a list of arguments showing that the existence of the defiled notion of “I” is necessary because its rejection would result in unacceptable consequences (Kramer 2016, 162ff.). Asaṅga not only treats the topic of “nothing but presentation” (vijñaptimātratā) in detail, but also defines what constitutes basic categories of conceptualization (vikalpa) and mental presentation (vijñapti). Most notable in this context are two lists of ten kinds of vikalpa and of eleven kinds of vijñapti,55 both summarizing various categories of a human being’s experience of the world. These classes of concepts include not only diverse aspects of the cognitive process, such as the perceived objects, the sense faculties, and the sense perceptions, but also concepts of “time,” “numbers,” the “surrounding world” (bhājanaloka), and the “distinction between oneself and others.” All these categories are assumed to be only modifications of the mind. In order to explain how, under these circumstances, a subject experiences the perception

260

Asan˙ga

of (external) objects, Asaṅga differentiates between two parts of the mind, the “image” (nimitta) portion and the “vision” (darśana) portion.56 Thus, every moment of perception is constituted of three different kinds of mental presentations, namely (1) the mental presentation of the sense faculty; (2) the mental presentation of the object – that is, the “image portion” of the mind; and (3) the mental presentation of the perception of the object, that is, the perceiving subject or the “vision portion” of the mind (Frauwallner 2010, 330). An important topic of Asaṅga’s investigation of the Bodhisattva’s path to liberation is “mental verbalization” (manojalpa). It is notable that Asaṅga treats mental verbalization as an important part of the Bodhisattva’s path to awakening. Thus, mental verbalization is regarded to be the basis for realizing true reality.57 The term “mental verbalization” is usually applied as a synonym of mental perception (manovijñāna) in Yogācāra texts (Kramer 2020, 365) and as such might also be assumed to be a hindrance on the path to liberation. In the third chapter of the Summary of the Mahāyāna Asaṅga explains that the Bodhisattva penetrates “nothing but presentation” by means of this very mental verbalization in the form of, for instance, the four investigations ( paryeṣana) and the corresponding four kinds of knowledge.58 The tenth chapter of the Summary of the Mahāyāna also shows some peculiarities that might be considered specific to Asaṅga. After a short overview of the three Buddha bodies – that is, the manifestation body (nirmāṇakāya), the enjoyment body (sambhogakāya) and the dharma body (dharmakāya) – this chapter deals extensively with the dharma body in most of its sections.59 In this part of the Summary of the Mahāyāna, we also find an explanation of the transformation of the cognitive system into the four kinds of awareness (  jñāna) as known from the Ornament of Mahāyāna Sūtras (chapter 9, verse 67), including the mirror-like awareness (ādarśajñāna), awareness of equality (samatājñāna), discriminating awareness ( pratyavekṣājñāna), and the all-accomplishing awareness (kṛtyānuṣṭhānajñāna). Notably, Asaṅga also mentions a less known explanation of five kinds of “mastery” (vibhutva) resulting from a change of the five constituents (skandha) of the person. He states that the dharma body is realized by means of the transformation of all five constituents (Lamotte 1938, 85ff. [5]). With regard to the body, the practitioner perfects the ability to manifest pure Buddha fields in all possible bodies, equipped with the Buddha signs and with countless melodious voices. Regarding the Buddha’s feelings, it is said that he has perfected the ability to remain in countless blissful states (sukhavihāra). His ideation (saṃjñā) is changed so that he becomes a master of teaching through words (nāman), phrases ( pada), and phonemes (vyañjana). The transformation of his impulses (saṃskāra) leads to the mastery of the manifestation (nirmāṇa), transformation (bsgyur ba), and the accumulation of positive factors (śukladharmasaṃgraha). Finally, his perception (vijñāna) changes into the four types of knowledge ( jñāna) mentioned previously.60 In summary, Asaṅga was one of the most influential masters of the Yogācāra tradition. Even though it remains controversial which works can definitely be attributed to him as their original author, his two main texts – the Compendium of the Abhidharma and the Summary of the Mahāyāna, whose authorship is considered fairly certain – offer a sufficient basis for this statement. In the Compendium of the Abhidharma, Asaṅga creates a distinctive system of Yogācāra Abhidharma and systematizes the ideas of the Foundation of the Yoga Practitioners. His Summary of the Mahāyāna became one of the most important works of the Yogācāra school, summarizing all the main Yogācāra teachings, such as the “store mind” (ālayavijñāna), the concept of “nothing but presentation” (vijñaptimātra), the three “natures” (svabhāva), and the three bodies (kāya) of a Buddha. Asaṅga might not have been the originator of these central Yogācāra concepts, but he doubtlessly contributed greatly to the creation of a coherent doctrinal system of the Yogācāra school.

261

Jowita Kramer

Notes 1 For details, see Watanabe (2000, 7). 2 See, e.g., Watanabe (2000, 5). 3 T 2049, fol.188a11ff. and 190c14ff.; Dalia (2002, 39ff. and 51). See also Delhey (2019, 73). 4 T 2087, fol. 896b20ff.; Beal I, 226ff. 5 For an overview of this discussion, see Kritzer (1999, 7–13). 6 Lokesh Chandra 103b4–104b2; Obermiller (1932, 137–39). 7 Lokesh Chandra 21a4ff. and Obermiller 1931, (54–56). 8 Lokesh Chandra 104b7ff. and Obermiller (1932, 140). 9 Lokesh Chandra 105b7ff. and Obermiller (1932, 143). The translation of this passage is quoted from Chimpa and Chattopadhyaya (1970, 168) (i.e., the translation of Tāranātha’s History of the Doctrine in India) as it is more precise than Obermiller’s rendering. 10 Schiefner (1868, 88,1–90,9); Chimpa and Chattopadhyaya (1970, 160–64). 11 Schiefner (1868, 88,1ff., 91,2–92,2 and 92,13–18) and Chimpa and Chattopadhyaya (1970, 160, 165–67). 12 Schiefner (1868, 99,8ff. and 105,18ff.) and Chimpa and Chattopadhyaya (1970, 177 and 186). 13 For a detailed overview of the various aspects and controversial points of this discussion, see Delhey (2019, 74–77). 14 Lokesh Chandra 1971, 21a4ff.; Schiefner (1868, 88,7). The Foundation of the Yoga Practitioners is also ascribed to Asaṅga in some late Indian sources (see Delhey 2013, 502; 2019, 75). 15 Lokesh Chandra 1971, 21a4ff.; Schiefner (1868, 88,7). Asaṅga is also mentioned as the author in the colophons of the Tibetan translation in D 4048 and of the Chinese translation in T 1594, fol. 132c19. 16 Lokesh Chandra 1971, 21a4ff.; Schiefner (1868, 88,6ff.). Asaṅga is also mentioned as the author in the colophons of the Tibetan translation in D 4049 and of the Chinese translation in T 1605, fol. 663a05. 17 Asaṅga is mentioned as the author in T 1602, fol. 480b13. 18 Asaṅga is mentioned as the author of the verses in T 1607, fol. 774a13, while the prose is attributed to Asaṅga. See also Delhey (2019, 77). 19 Asaṅga is mentioned as the author in T 1607, fol. 774a13. 20 Asaṅga is mentioned as the author in T 1565, fol. 44c19. 21 Asaṅga is mentioned as the author in T 1514, fol. 885a06. For further references to studies discussing the authorship of this work, see Delhey (2019, 78). 22 See Chimpa and Chattopadhyaya (1970, 394) and D 1115. 23 Asaṅga is mentioned as the author in the colophon of the Tibetan translation in D 3550. 24 Asaṅga is mentioned as the author in the colophon of the Tibetan translation in D 3648. 25 Lokesh Chandra 1971, 105a1; Asaṅga is also mentioned as the author in the colophon of the Tibetan translation in D 3981. 26 Asaṅga is mentioned as the author in the Tibetan colophons of these three works in D 3982, D 3983, and D 3984. 27 The colophon of the Tibetan translation in D 4025 attributes it to Maitreya. As mentioned previously, Bu ston also states in his History of the Doctrine that Asaṅga composed a commentary on the Distinguishing the Germ of the [Three] Jewels (Lokesh Chandra 1971, 105a1). 28 Asaṅga is mentioned as the author in the colophon of the Tibetan translation in D 4073. 29 Schiefner (1868, 88,7ff.). Bu ston mentions this work as *Tattvaviniścaya (De nyid rnam nges; Lokesh Chandra 1971, 105a1; on this title, see Brunnhölzl [2014, 103, n. 102]). According to Delhey (2019, 75), Haribhadra also ascribes a commentary on the Ornament of Clear Realization to Asaṅga. 30 According to Delhey (2019, 74ff.), Ᾱryavimuktisena attributes this commentary to Asaṅga instead of Vasubandhu. 31 For the various possibilities of retranslating the title of this text to Sanskrit, see Schmithausen (1987, 261, n. 99). 32 For bibliographical references of various modern studies on several of these texts, see Watanabe (2000, 7–15). 33 For further details on the authorship of these three works, see Delhey (2019, 78), who points out that it is probably impossible to exclude the possibility of Asaṅga’s authorship in the case of the Seventy Stanzas on the Perfection of Wisdom in Three Hundred Stanzas. 34 See also Schmithausen (2018, 265ff.).

262

Asan˙ga 3 5 For further details, see, e.g., Williams 179ff. 36 Radich (2007, 1187); Bayer (2010, 22). 37 For further details of this discussion, see Schmithausen (2018, 266). 38 According to Schmithausen (2018, 264), this is one of two strands from which the concept of “mind only” emerged, the other being the Sūtra of the Concentration of the Direct Encounter with the Buddhas of the Present (Pratyutpannabuddhasaṁmukhāvasthitasamādhisūtra), where cittamātra is mentioned in connection with the visions of the Buddha Amitāyus. 39 The Sanskrit text of the Compendium of the Abhidharma was first edited by Gokhale in 1947 on the basis of a fragmentary manuscript preserved in Tibet. In 1950, the whole Sanskrit text was published by Pradhan, who filled in the passages missing in the manuscript with his own reconstructions of the Sanskrit from the Tibetan translation of the Compendium of the Abhidharma (D 4049). For the Chinese translation of the Compendium of the Abhidharma, see T 1605. Recently also a Sanskrit manuscript of the Commentary on the Compendium of the Abhidharma (Abhidharmasamuccayavyākhyā), which contains the whole of its root-text – that is, the Compendium of the Abhidharma – came to light and is currently being edited by Li Xuezhu and Kazuo Kano (see Li 2012, 2013, 2014, 2015; Li and Kano 2014). For a French translation of the Compendium of the Abhidharma, see Rahula (1980), whose translation has been rendered into English by Sara Boin-Webb (2001). 40 Kritzer (1999, vii). 41 Bayer (2010, 22). As noted by Bayer, the idea of the apratiṣṭhitanirvāṇa is also present in the treatises attributed to Maitreya, such as the Mahāyānasūtrālaṃkāra. 42 See Schmithausen (1972, 156); Bayer (2010, 21). 43 See Bayer (2010, 20ff.); Schmithausen (2014, 30ff.). 44 See Schmithausen (1987, 100); Schmithausen (2014, 302ff.). 45 The Sanskrit text of this passage is not extant. For the Tibetan version, see D 4049, 101b1. 46 The Sanskrit text of this passage is not extant. For the Tibetan version, see D 4049, 118a6ff. 47 Kritzer (1999, 5); Bayer (2010, 37ff.). 48 The Summary of the Mahāyāna has been translated to French in Lamotte (1938) and to English in Brunnhölzl (2018). 49 Watanabe (2000, 16); Brunnhölzl (2018, 4ff.). 50 Schmithausen (1987, 403, n. 713); Bayer (2010, 32). 51 For a detailed summary of the ten chapters, see Brunnhölzl (2018, 7–21). 52 Lamotte (1938, 19ff., 45–49). See also Schmithausen (1987, 79ff.). 53 Lamotte (1938, 30ff., 14). For the Sanskrit text, see ASBh 41,21–42,15. 54 For details, see Kramer (2016). 55 Lamotte (1938, 25 [2] and 34ff. [20]). See also Frauwallner (2010, 329); Kramer (2018, 327, 329). 56 Lamotte (1938, 29 [11]). See also Frauwallner (2010, 329). 57 See also Watanabe (2000, 77ff.). 58 Lamotte (1938, 51ff. [7]). See also Brunnhölzl (2018, 13). 59 For a detailed study of this chapter of the Summary of the Mahāyāna, see Griffiths et al. (1989). 60 See also Brunnhölzl (2018, 237).

Bibliography Bayer, Achim. 2010. The Theory of Karman in the Abhidharmasamuccaya. Tokyo: The International Institute for Buddhist Studies. Beal, Samuel. 1884. Si-yu-ki: Buddhist Records of the Western World, vol. 2. London: Trübner & Co. Boin-Webb, Sarah. 2001. Abhidharma Samuccaya: The Compendium of the Higher Teaching (Philosophy). Fremont: Asian Humanities Press (English translation of Rahula 1980). Brunnhölzl, Karl. 2014. Gone Beyond: The Prajñāpāramitā Sūtras, The Ornament of Clear Realization, and Its Commentaries in the Tibetan Kagyu Tradition. Boulder: Snow Lion Publications. ———. 2018. A Compendium of Mahāyāna, Asaṅga’s Mahāyānasaṃgraha and Its Indian and Tibetan Commentaries, vol. 1. Boulder: Snow Lion Publications. Chandra, Lokesh. 1971. The Collected Works of Bu-ston. Part 24 (ya). Śata-Piṭaka Series 64. New Delhi: International Academy of Indian Culture. Chimpa, Lama, and Alaka Chattopadhyaya. 1970. Tāranātha’s History of Buddhism in India. Simla: Indian Institute of Advanced Study.

263

Jowita Kramer Dalia, Albert A. 2002. “Biography of Dharma Master Vasubandhu.” In Lives of Great Monks and Nuns, translated by Albert A. Dalia and Li Rongxi, 31–53. Berkeley: Numata Center for Buddhist Translation and Research. Deleanu, Florin. 2006. The Chapter on the Mundane Path (Laukikamārga) in the Śrāvakabhūmi. Tokyo: The International Institute for Buddhist Studies of the International College for Postgraduate Buddhist Studies. Delhey, Martin. 2013. “The Yogācārabhūmi Corpus Sources, Editions, Translations, and Reference Works.” In The Foundation for Yoga Practitioners: The Buddhist Yogācārabhūmi Treatise and Its Adaptation in India, East Asia, and Tibet, edited by Ulrich Timme Kragh, 498–561. Cambridge: Harvard University Press. ———. 2019. “Asaṅga/Maitreya(nātha).” In Brill’s Encyclopedia of Buddhism. Vol. II: Lives, edited by Jonathan Silk et al., 73–80. Boston: Brill. Frauwallner, Erich. (1956) 2010. Philosophie des Buddhismus. Berlin: Akademie Verlag. Gokhale, V. V. 1947. “Fragments from the Abhidharmasamuccaya of Asaṅga.” Journal of the Bombay Branch of the Royal Asiatic Society 23: 13–38. Griffiths, Paul et al. 1989. The Realm of Awakening: Chapter Ten of Asaṅga’s Mahāyānasaṃgraha. New York: Oxford University Press. Kramer, Jowita. 2016. “Some Remarks on the Proofs of the ‘Store Mind’ (Ālayavijñāna) and the Development of the Concept of Manas.” In Text, Philosophy, and History: Abhidharma across Buddhist Scholastic Traditions, edited by Bart Dessein and Weijen Teng, 146–68. Leiden: Brill. ———. 2018. “Conceptuality and Non-Conceptuality in Yogācāra Sources.” Journal of Indian Philosophy 46 (2): 321–38. ———. 2020. “Concepts of the Spiritual Path in the *Sūtrālaṃkāravṛttibhāṣya (Part II): The Eighteen Manaskāras and the Adhimukticaryābhūmi.” In Mārga: Paths to Liberation in South Asian Buddhist Traditions, edited by Cristina Pecchia and Vincent Eltschinger, 329–61. Wien: Österreichische Akademie der Wissenschaften. Kritzer, Robert. 1999. Rebirth and Causation in Yogācāra Abhidharma. Wien: Arbeitskreis für Tibetische und Buddhistische Studien, Universität Wien. Lamotte, Étienne. 1938. La somme du grand véhicule d’Asaṅga (Mahāyānasaṃgraha). Louvain: Université de Louvain, Institut Orientaliste. Li, Xuezhu. 2012. “ ‘Abhidharmasamuccayavyākhya no jobun ni tsuite’ Abhidharmasamuccayavyākhya の序文について (‘Sanskrit text of the Opening Section of Sthiramati’s Abhidharmasamuccayavyakhya’)”. Indogaku Chibettogaku Kenkyu インド学チベット学研究 (Journal of Indian and Tibetan Studies) 16: 1–7. ———. 2013. “Diplomatic Transcription of Newly Available Leaves from Asaṅga’s Abhidharmasamuccaya.” Annual Report of the International Research Institute for Advanced Buddhology at Soka University 16: 241–53. ———. 2014. “Diplomatic Transcription of Newly Available Leaves from Asaṅga’s Abhidharmasamuccaya – Folios 29, 33, 39, 43, 44.” Annual Report of the International Research Institute for Advanced Buddhology at Soka University 17: 195–205. ———. 2015. “Diplomatic Transcription of the Sanskrit Manuscript of the Abhidharmasamuccayavyākhyā – Folios 2v4–8v4.” Annual Report of the International Research Institute for Advanced Buddhology at Soka University 18: 275–83. Li, Xuezhu, and Kazuo Kano. 2014. “Restoration of Sanskrit Text in Missing Leaves (Fols. 2, 6, 7) of the Abhidharmasamuccaya Manuscript on the Basis of the Abhidharmasamuccayavyākhyā Manuscript.” China Tibetology 23 (2): 53–63. Martin, Dan. 2002. “Gray Traces: Tracing the Tibetan Teaching Transmission of the Mngon-pa Kun-btus (Abhidharmasamuccaya) Through the Early Period of Disunity.” In The Many Canons of Tibetan Buddhism, edited by Helmut Eimer and David Germano, 335–57. Leiden: Brill. Obermiller, Eugene. 1931. History of Buddhism by Bu-ston (Part I: The Jewelry of Scripture). Heidelberg: Harrassowitz. ———. 1932. History of Buddhism by Bu-ston (Part II: The History of Buddhism in India and Tibet). Heidelberg: Harrassowitz. Pradhan, Pralhad. 1950. Abhidharma Samuccaya of Asanga. Santiniketan: Visva-Bharati. Radich, Michael. 2007. “The Somatics of Liberation: Ideas about Embodiment in Buddhism from Its Origins to the Fifth Century C.E.” PhD diss., Harvard University.

264

Asan˙ga Rahula, Walpola. 1980. Le compendium de la super-doctrine ( philosophie) (Abhidharmasamuccaya) d’Asanga. Paris: École Française d’Extrème Orient. Schiefner, Anton. 1868. Tāranāthae de Doctrinae Buddhicae in India Propagatione Narratio. St. Petersburg: Academia Scientiarum Petropolitana. Schmithausen, Lambert. 1972. “The Definition of Pratyakṣam in the Abhidharmasamuccayaḥ.” Wiener Zeitschrift für die Kunde Südasiens 16: 153–63. ———. 1985. “Once Again Mahynasaṃgraha I.8.” In Buddhism and Its Relation to Other Religions: Essays in Honour of Dr. Shōzen Kumoi, 139–60. Kyoto: Heirakuji Shoten. ———. 1987. Ālayavijñāna: On the Origin and Early Development of a Central Concept of Yogācāra Philosophy. Tokyo: The International Institute for Buddhist Studies. ———. 2014. The Genesis of Yogācāra-Vijñānavāda: Responses and Reflections. Tokyo: The International Institute for Buddhist Studies. ———. 2018. “Some Remarks on the Genesis of Central Yogācāra-Vijñānavāda Concepts.” Journal of Indian Philosophy 46 (2): 263–81. Skilling, Peter 2013. “Vaidalya, Mahāyāna, and Bodhisatva in India: An Essay Towards Historical Understanding.” In The Bodhisattva Ideal: Essays on the Emergence of Mahāyāna, edited by Bhikkhu Nyanatusita, 69–119. Kandy: Buddhist Publication Society. von Rospatt, Alexander. 1995. The Doctrine of Momentariness. Stuttgart: Franz Steiner Verlag. Watanabe, Chikafumi. 2000. “A Study of Mahāyānasaṃgraha III: The Relation of Practical Theories and Philosophical Theories.” PhD diss., University of Calgary. Williams, Paul. 1989. Mahāyāna Buddhism: The Doctrinal Foundations. London: Routledge.

265

16 VASUBANDHU Mainstream and Mahāyāna Jonathan C. Gold

Introduction: Vasubandhu’s Identity Problem Vasubandhu (fourth century CE) sits at the center of the Buddhist philosophy curriculum. His Commentary on the Treasury of Abhidharma (Abhidharmakośabhāṣya), especially its ninth chapter, contains (among other things) many of the most compelling and influential arguments in defense of the definitive Buddhist view of no-self.1 In India, this work was debated by Buddhists of different schools, as well as between Buddhists and non-Buddhists, and its ideas were fundamental to the development of Buddhist epistemology. Its influence expanded with Buddhism into China and later Tibet, where it remains central to scholastic education. In recent decades, it has found a new, admiring readership among Western philosophers such as Derek Parfit. Similarly influential was Vasubandhu’s Twenty Verses (Viṃśikā), also called the Establishment of Appearance Only (Vijñaptimātrasiddhi), which contains the most widely repeated arguments in defense of the Buddhist understanding that everything in experience is only mind, like an illusion or a dream.2 Many other works, some of them also extremely influential, are attributed to Vasubandhu (Skilling 2000). But just looking at these two, we see a strange fact: The first is a work of mainstream Abhidharma philosophy, analyzing the nature of the foundational elements (dharmas) found within the Buddha’s early teachings. The second, however, expresses doctrines of the Great Way (Mahāyāna) and denies the ultimate reality of the very elements decisively defended in the first. The works establish quite different perspectives. Vasubandhu’s traditional biographies account for the difference between these works with a story of his having converted from mainstream Buddhism to Mahāyāna late in his life, under the influence of his elder brother, Asaṅga (Takakusu 1904). The scholar Erich Frauwallner (1898– 1974) proposed that the difference could be accounted for by the realization that there were, in fact, two different scholars named Vasubandhu – an earlier one who wrote the Mahāyāna treatises and commentaries and a later one who wrote the Treasury Commentary (Frauwallner 1951). This idea has drawn much discussion, and the “two Vasubandhus” hypothesis lives on (Buescher 2013; Kapstein 2018). Yet, a key point for our purposes is that Frauwallner was compelled to clarify his thesis and acknowledge that the two doctrinally conflictual texts mentioned previously were, in fact, most likely written by the same person named Vasubandhu (Frauwallner 1961, 131–32). According to the revised “two Vasubandhus” account, the texts that the “earlier” Vasubandhu wrote are now thought to include commentaries on Mahāyāna 266

DOI: 10.4324/9781351030908-25

Vasubandhu

scriptures and treatises, but not all of the Mahāyāna treatises (and, notably, not the Twenty Verses itself). The “two Vasubandhus” hypothesis, whatever its legitimacy, does not explain why both mainstream and Mahāyāna positions are defended in the same author’s works. To add to the problem, tradition ascribes Vasubandhu authorship of not just the  Trea­ sury’s ­Commentary, but also its root verses, the Treasury of Abhidharma Verses (Abhidharmakośakārikā). This root text provides a comprehensive systematization of the Abhidharma system according to the Vaibhāṣika (also called the Sarvāstivāda), a position that the Commentary tends to challenge and critique from many perspectives, primarily one that the Commentary calls Sautrāntika. So Vasubandhu apparently represented the Vaibhāṣika and its opponents, respectively, in the root text and commentary. Traditional biographies and surviving criticisms of his work attest that Vasubandhu was, indeed, considered a traitor by Vaibhāṣikas (Park 2014). That means he was not just a flip-flopper from mainstream to Mahāyāna; he switched within the mainstream of Abhidharma, as well, which is why his extant work represents three distinct positions. The tradition that followed from Asaṅga, which his Mahāyāna works defend, is called Yogācāra. So we have Vasubandhu writing as Vaibhāṣika, Sautrāntika, and Yogācāra. At the risk of yet further complication, we may note that when the term Yogācāra names a system of Indian Mahāyāna philosophy, whether in traditional or contemporary scholarship, it is commonly contrasted with Madhyamaka. But Vasubandhu never used the term “Yogācāra” to refer to his own work, and he made use of some of the key argumentative moves that are seen in the Madhyamaka works of Nāgārjuna. He never disparaged that system of thought, except where it was misread as an overly zealous negation of traditional doctrines (a disparagement that is contained within Nāgārjuna’s works, as well) (Gold 2015b). So of the four standard schools of philosophy that Tibetans list in the introductory textbooks – Vaibhāṣika, Sautrāntika, Yogācāra, and Madhyamaka – Vasubandhu appears to encompass them all. A solution to the problem comes when we understand Vasubandhu not as a representative of a particular school or even a set of schools but as a philosopher dealing in turn with a range of inherited traditions. For in spite of the diversity of the various identities adopted therein, the previously mentioned texts display a notable consistency of method and style. Other works ascribed to Vasubandhu diverge in style but often extend a coherent set of ideas. We may doubt whether one person wrote the various texts attributed to him – but a unified, and unifying, intention resonates across them. When they are read together, they exhibit a distinctive approach that weaves Mahāyāna through the full fabric of mainstream Buddhism. Traditional biographies of Vasubandhu (e.g., Takakusu 1904) depict a brilliant and combative scholar whose career seems to have been animated by controversy. He was the middle of three sons of a prominent brahmin in the city of Puruṣapura in Gandhara, all officially named “Vasubandhu” – though his younger brother went by Viriñcivatsa and his elder by Asaṅga. At the time, the Sarvāstivāda system of the Vaibhāṣikas was the pinnacle of Buddhist learning and it was officially taught only in Kashmir. Yet a handful of scholars had surreptitiously brought these teachings out of the valley, and Vasubandhu was able to master them, possibly through his teacher Buddhamitra. He composed a verse summary of the Vaibhāṣika system in 600 verses, which the elders of Kashmir apparently approved, but then wrote his commentary, which they considered traitorous. At one point, Buddhamitra was publicly humiliated by a defender of the Sāṃkhya school while Vasubandhu was away, so after he returned, he wrote a devastating critique of the Sāṃkhyas. This earned him praise and royal support, and he became the personal instructor of the queen and the crown prince. Subsequently, after being challenged by a Vyākaraṇa (grammar) scholar, he wrote a similarly ruinous critique of the Vyākaraṇas. This defeated enemy 267

Jonathan C. Gold

drew Vasubandhu to the attention of the Vaibhāṣika master Saṅghabhadra, who in turn wrote two extensive criticisms of Vasubandhu’s Treasury Commentary and challenged Vasubandhu to a debate. This time, Vasubandhu refused the contest since he believed that he had already made a strong case and did not believe that his views had been undermined. The final challenge to Vasubandhu’s views, however, came from his brother Asaṅga. Asaṅga had spent his life propounding the Mahāyāna teachings he had solicited directly from the future Buddha Maitreya in heaven, and he worried that his brilliant brother would compose a treatise countering the legitimacy of the Mahāyāna. We are told that Asaṅga, feigning illness, sent for Vasubandhu and, when his brother arrived, personally taught him the Mahāyāna. Vasubandhu saw the depth of the teachings and changed his perspective entirely, regretting his previous doubts. He became, thenceforth, a defender of the Mahāyāna and spent his later years composing his various Mahāyāna commentaries and treatises. In this chapter, I will use a few of Vasubandhu’s key arguments to draw out a unifying theme in his approach to causality and scripture. Vasubandhu’s preference for empirically grounded causal stories over conceptual-linguistic abstractions, we will see, implies that the words of the Buddha must often be considered figurative or under a secondary, special intention. Vasubandhu’s understanding of the Buddha’s indirect use of language is grounded in his early reading of the doctrine of no-self as an expression of dependent origination, which, when understood to be paradigmatic for all doctrines, provides him an opening to Mahāyāna in a Yogācāra mode. Central to our reading of Vasubandhu’s view is his recognition that Yogācāra’s distinctive position – the mind-only nature of reality – is both an ontological thesis about the impossibility of perceived entities and a reading of how concepts generate false appearances in the mind. Yogācāra in Vasubandhu’s hands never loses sight of the fact that even false concepts are the most important and most effective method used by the Buddha to move disciples forward on the path to enlightenment – that is to say, the concepts of the Dharma. To acknowledge the mind-only nature of all dharmas is not to reject the Buddha’s teachings, but rather to accept that all of the Buddha’s teachings are unified in the common project of revealing the ineffable nature of reality, to whatever degree living beings are able to see it. As we will see, an appeal to dependent origination as the causal basis of both reality and illusion is formalized in the Yogācāra doctrine of three natures, but it was already implicit in the argument that the false view of the self is superimposed on the aggregates. The following two sections trace this argument first in Vasubandhu’s mainstream Abhi­ dharma and then through his Mahāyāna works.

Perception, Causality, and Scripture in the Treasury Commentary The final chapter of Vasubandhu’s Treasury Commentary begins by stating that it is a key teaching of the Buddha that the self is a false construction imposed upon the ever-changing continuum of five aggregates: forms (rūpa), feelings (vedanā), cognitions (saṃjñā), conditionings (saṃskāra), and consciousness (vijñāna). Then, in response to an objector’s request for an explanation of how he can know this, Vasubandhu throws down the gauntlet by stating a proof that there is no self: And how is this to be understood, that the word “self ” indicates only the continuum of aggregates, and does not apply elsewhere? Because of there being neither perception nor inference. For where there are entities (dharmas), their perception is apprehended when there is no obstruction. Such is the case for the six sensory objects (viṣaya) and the mind. And there is an inference in the 268

Vasubandhu

case of the five sensory organs. Here the inference is that with a cause in place, when another cause does not exist, no result is seen, and when it does exist the result too exists, as in the case of a sprout. That is, with the cause in place that consists in the sensory object with a particular appearance together with mental effort, the grasping of the sensory object is seen to not exist and to exist, respectively, for people who are blind or deaf, etc. and people who are not blind or deaf, etc. So the nonexistence and existence, respectively, of another cause are determined. And that other cause is the sensory organ: that’s the inference. And this does not exist for the self, so there is no self.3 Vasubandhu declares that there are two means for ascertaining whether or not something is real – perception and inference – and for each, he provides a model and a set of objects of awareness. In one of his main categorical systems, the Buddha explained that all of experience was divisible into twelve so-called “spheres” or “domains” (āyatana): six sensory organs and six sensory objects, six to include the five ordinary senses plus the mind and its objects. Vasubandhu answers the question, then, by switching from the aggregate system to the sphere system. All twelve spheres, Vasubandhu says, can be known by perception and inference, whereas there is no such argument for the self. Thus, there is no self. Perception is the direct acquisition of knowledge about perceptual objects. Vasubandhu does not defend the idea that the self is unperceived, but this is an important premise of the no-self view. On the one side, it is in line with prevalent non-Buddhist positions (“By what means can one perceive the perceiver?” BAU 4.5.15, Olivelle 1996, 71), and on the other side, it challenges potential opponents to define just what aspects of the self may be perceived, over and above the aggregates (or, in this case, the spheres). He says that what is perceived consists in, naturally, the sensory objects – but also, interestingly, the mind itself. This adds a twist to an otherwise straightforward claim. The great Buddhist epistemologist Dignāga, who is said to have been a follower if not a disciple of Vasubandhu, argued that the mind is intrinsically selfperceiving (svasaṃvedana), a view that became a crucial topic in later Buddhist philosophy. This is not Vasubandhu’s view; in order to understand what he is saying, we need to delve into what he means by considering the mind as mental organ (manas). As previously mentioned, the Buddha classified experience into twelve spheres, as well as into five aggregates. In addition, he provided a third comprehensive classificatory system, the eighteen “realms” (dhātu). Abhidharma analysis works to bring these three systems – and many other subsidiary systems – into alignment. The “realms” classification keeps all twelve of the twelve spheres and adds six consciousnesses: five sensory consciousnesses and one mental consciousness. Now, an Abhidharma philosopher, charged with making sense of Buddhist doctrine, needs to be able to explain: if the Buddha already said that all of reality may be classified among the twelve spheres, how can there be six further elements in the realms classification? Was the spheres classification not really comprehensive? Vasubandhu addresses this issue in his discussion of the nature of the mind as mental organ (I.17; Pradhan 1975, 11–12). What he says is that the six consciousnesses in the realms classification are just an expansion of the mind as understood in the spheres classification. The mind as mental organ just is the sensory consciousnesses to the extent that they serve as a basis for awareness. This, after all, is how sensation works: the eye comes into contact with a visual object, which in turn generates a visual consciousness. If that visual consciousness is taken up in a subsequent awareness (in a memory, for instance), it is playing the role of a “sensory organ” for mental awareness, analogous to the eye with its visual object. So this is why the mind is included among the objects of perception: not because the mind perceives itself in its every act of awareness, but because sensory consciousnesses commonly appear to subsequent awarenesses. 269

Jonathan C. Gold

When a visual consciousness comes into being, there is perception of the visual object. When a mental consciousness comes into being, there is a perception of a mental object, as when one has a thought. When the mental object happens to be a previous mental consciousness – for instance, when a visual consciousness is remembered – there is perception of the mind. This is how the mind becomes an object of perception. It is not the case that all mental events are objects of mental perception. But sometimes, the mind is perceived. This accounts for the internal consistency of the Buddha holding two different classificatory systems because they are two ways of speaking of the same thing. The realms classification with its six consciousnesses allows you to supply a causal basis and object of awareness for the mental organ. But if you do not need to detail its operations, the mental organ can be considered under a spheres classification as including seven distinct realms (six consciousnesses and the mental organ). In fact, Vasubandhu allows that if one is considering only substantial entities, the realms classification ought to list only seventeen, not eighteen, because if you list the six consciousnesses that are an expansion of the mental organ, you do not also need to list the mental organ. It is only from the perspective that distinguishes the mental organ as a support for subsequent awareness that you need to list them all. The Buddha’s classificatory systems are all fully comprehensive, but they serve distinct discursive purposes.4 To return to the argument against the self under the spheres classification, then, if we have covered all sensory objects and the mental organ, what is left unperceived are the five material sensory organs. These, Vasubandhu says, can be established by inference. The inferential structure he provides is original and consequential for our understanding of his approach to reality. It is a causal inference: one may infer the reality of an unperceived entity if it is a necessary component – the unseen cause – in a causal relation. We know that when a blind person and a person who is not blind are put in the same visual situation, the blind person cannot see what the other person can see. This is formalized as, “with a cause in place, when another cause does not exist, no result is seen, and when it does exist the result too exists.”5 In such a situation, the thing that either does not exist or does exist – in this case, the faculty of the visual organ – is ascertained with certainty. I am wary of those who profess Buddhism to be “science” as opposed to a religion; it is not. But Vasubandhu’s emphasis on causal reasoning with a generally empirical bent resonates with modern models for scientific reasoning. He emphasizes the evidential power of a “seen” (dṛṣṭa) result even when inferring an unperceived object, and he only allows that causality is inferable once proper controls are in place (quasi-experimental controls, because this is a thought experiment). Most importantly, reality is founded, most basically, on causality. Vasubandhu does not say that this is the only valid inferential means to an unperceived entity, but he does say that all the unperceived spheres can be proven through this means, and across his work, he often demands precise definitions and disallows reality ascription until an empirically grounded causal argument is found. Vasubandhu also has his own version of the principle of scientific reasoning commonly known as Occam’s Razor, or explanatory simplicity, which he applies to eliminate causal overdetermination. He mocks an opponent who accepts a visible cause but still holds onto an invisible one, saying, “For him, once his laxative is taken, the goddess causes the purge – so what else will he imagine?”6 Together, he uses these methods to disprove the eternal self and other non-Buddhist eternal entities such as a creator God and atoms, as well as a host of categories defended in the Vaibhāṣika system, from invisible materiality to substantially existent past and future. Vasubandhu’s requirement that you provide the causal matrix for every entity is also central to his famous argument proving universal momentariness (Pradhan 1975, 193.2–8). If everything is defined according to what causes it and what it causes, to exist is just to be a 270

Vasubandhu

cause for specifiable results. This means that a thing can only be said to properly exist when it is producing its subsequent results. Consequently – and this is Vasubandhu’s innovation (Von Rospatt 1995) – if a thing does not have it in its nature to pass out of existence, there is nothing that could make it pass out of existence; conversely, a thing that has it in its nature to continue to exist would have to always exist. The only thing that could stop it from continuing to exist would be something that was the cause of its non-existence. However, a “non-existence” is just the kind of thing that we think is real because of linguistic construction but does not really exist. A non-existence is just what we think we see when we expect to see something that is not present. So no entity could be defined as having a substantial nature that brings about another entity’s non-existence; that would be like saying it causes a bachelor’s wife. The passage we have been discussing initiates the chapter with the challenge to his opponent that no self can be proven, and it is followed by a variety of responses that attempt to find a causal justification for the self. Much of the chapter is taken up with answers to questions about how the no-self view – that is, the view that all there is to the self is the aggregates – can account for accepted empirical realities such as memory, karma, or the perception of an apparent person. But although the refutation of self is Vasubandhu’s central purpose in the chapter, his opening point is not just that there is no self; it is that the self is a false construction imposed upon the aggregates, and that this was central to the Buddha’s teaching. I will address in turn these two crucial additions, which are important to Vasubandhu over and above the idea that there is no self: the self as imposed upon the aggregates and the use of this truth as a guide to the Buddhist scriptures. First, it is one thing to say that there is no essential reality to the self, but it is something else entirely to say that the self is a false construction imposed upon the aggregates, spheres, and realms. To say that is to admit more effective, more useful, measurable entities that deserve our recognition because they can provide a truer reality than our false belief in the self. There are substantial entities which, unlike the self, exist in a nexus of causes and results that we can see and deduce. This nexus allows us to advance on the path to liberation and ultimately free ourselves from the false construction which has no causal basis. The key distinction here is that between how things appear, on the one hand, and how things actually operate, causally, on the other. Now, this is a natural distinction that takes no special instruction to affirm, but it will be of use to note that the two contrasting concepts make use of two distinct frames of reference. Things look one way from a certain perspective, and things are really happening, causally, in a different way from another perspective. It is important to note that there is a hierarchy here between perspectives: the false construction is grounded within the substantial perspective, whereas the opposite is not true. The belief in the self is grounded in a misperception of the five aggregates – the body, first, and also the four mental aggregates. These aggregated components are in constant flux; they have no unifying “self ” or essence or core – but it is with reference to the aggregates that people imagine themselves to have a self. I sometimes call this a distinction between viewing-of and viewing-as-if. One experiences the aggregates, but takes them to be a self. Now, several paragraphs on from the section we have been discussing, Vasubandhu allows an imagined opponent to question him on this: why, the opponent asks, are you so sure that it is the aggregates that ground the mistaken idea of the self? This is no longer an argument that there is no self; it is an argument that instead of the self, there is something else, specifically the aggregates, which we mistake for a self. He answers as follows: People of ordinary intellect come to believe, “I am white; I am dark; I am fat; I am thin; I am old; I am young.” They identify themselves with these things. Souls are not of this type. Therefore, the identification of the self has the aggregates as its object.7 271

Jonathan C. Gold

We all know that an eternal soul is not “old” or “young,” it is not “thin” or “fat,” it is not “white” or “dark.” These are qualities that might be attributable to a body, but not a soul. If a soul can be reborn, it can take on different kinds of bodies, so it could never be “white” or “dark,” “fat” or “thin” in itself; and being “old” or “young” indicates not just a particular body, but a particular time of life for a particular body. So Vasubandhu’s argument is simple and straightforward: people do, of course, think of themselves as “white” or “dark” – he is acknowledging that – but those are qualities of aggregates, not a soul. It is a mistake to attribute this kind of quality to a self. The fallacious construction of self, then, is fundamentally a mistake whereby we grasp onto and identify with impermanent components – the aggregates. Notice that the translator here, Charles Goodman, has made something of a clever choice in rendering gaura and śyāma as “white” and “dark.” These are color terms, and they could just be rendered “white” and “black”; but that translation would imply, in this context in English, the racial identities associated with those terms in a way that would be anachronistic and culturally inappropriate. In addition to their use as literal color terms, the words are used in Sanskrit to indicate lightness or darkness of skin color, both as terms of beauty. The gaura beauty shines brightly like the moon, and the śyāma beauty has a glossy shine, like a raven. Needless to say, in those contexts, they do not literally mean the colors “white” and “black,” and this is what might motivate the alternate translation as “light” and “dark” here. Goodman’s rendering chooses one from each column: “white” and “dark.” For me, this elegantly foregrounds the difference between our social setting and the classical Indian one while, at the same time, displaying how in both cultural contexts, the false reification of skin color-related identity is implicit in the terms themselves. This foregrounding of culturally generated biases might be distracting, were it not the central point of the passage to argue that we delusively generate our identities in dependence upon temporary, conditioned appearances. The first several times I  read this passage, I  assumed that when he said we ground our affection for ourselves in the aggregates, Vasubandhu was speaking entirely about the material aggregate, the body. He mentions the aggregates all together, but the qualities seem to pick out the material body. I was amazed to notice that Vasubandhu and other classical Buddhists were aware of what we think of today as psychological constructions imposed upon the body. But I had not seen just how insightful he was. Recently, thinking about this passage in the light of the widely acknowledged social construction of race, it occurred to me that the form aggregate – the material body itself – is unsuitable for identification. The error in which we experience (vijñāna) an affectionate identification with qualities of the material body (rūpa) requires that we apply classifications (saṃjñā) that are based in inherited linguistic and conceptual patterns (saṃskāra) and that are always associated with specific feelings and attitudes (vedanā). Remarkably, Vasubandhu was aware of the psychological construction of identities (though he was not aware of the social construction of identities) that the aggregates impose upon the material body. He understood that the illusion of self involves all the aggregate groups. Yet even if all of the aggregates are involved in the process of self-construction, the point that Vasubandhu is making in this passage highlights the distinction between the aggregates as actively constructing identities, which happens in the background (so to speak), and the material body aggregate, with apparent qualities, with which the self is ostensibly identified. This is, again, viewing-of and viewing-as-if: to say “I am white” is to see not just the aggregates, but specifically the material body, as a self. The key difference between these two levels is that the basis, the material aggregate, has its own causal story that has no need for the causal story of the illusory view of self or the conceptual construction of “whiteness.” Their causal story – the causal story of these views – requires that one import a description of the other aggregates: karmic dispositions, emotional patterns and the imposition of false constructs. But when one 272

Vasubandhu

describes those causes, there is no longer anything one can literally describe as “self ” or “whiteness.” They are just appearances. Vasubandhu uses the distinction between useful causal descriptions (aggregates, spheres, and realms) and false constructions not only to articulate the view of no-self, but also to provide a method for analyzing Buddhist scripture. This is the second significant purpose in the chapter, over and above the proof of no-self, that I mentioned previously. In fact, much of the chapter is taken with scriptural citation, both in defense of the scriptural basis for no-self and in reply to opponents’ interpretations of scriptures that challenge that notion. To take a wider view, Buddhist philosophy is most fundamentally an attempt to make sense of Buddhist scripture. Vasubandhu himself may have been more aware of this than most philosophers; his Proper Mode of Exposition (Vyākhyāyukti) is the only known Indian Buddhist treatise dedicated to scriptural interpretation. But the philosophical traditions in Buddhism are all, directly or indirectly, derived from Abhidharma, which simply means “true Dharma” or “higher Dharma” – that is, it is the practice dedicated to clarifying the reality known and expressed by the Buddha.8 But granting that Abhidharma – and, by extension, all Buddhist philosophy – must consider scriptural evidence, Vasubandhu wants to distinguish between legitimate citations and those that are foolish and misleading. Vasubandhu provides an extensive set of citations that lay out the no-self view directly. But he is not a literalist in scriptural interpretation. On the contrary, one of his key criticisms of his opponents is that they deduce the reality of entities based on scriptural passages that are simply using conventional terminology to speak about complex entities. That is, there is a tendency to take structures that are natural grammatical expressions or common-sense expressions and read them at their surface level as reified entities. To pick a telling example from earlier in the Treasury Commentary (I.42), this kind of erroneous reification of entities appears to follow naturally from the Buddha’s statement that “the eye sees, the mind cognizes,” which suggests that the eye is an agent of seeing. In truth, as Vasubandhu argues, the eye is not an agent. It is a causal factor that, in the presence of the appropriate kind of object, generates a visual consciousness. It would be more accurate to say that it is the visual consciousness which “sees” – but even this is not entirely correct, for the visual consciousness contains the visual information within it; it is not distinct from that visual appearance. So it is not doing anything that could be considered “seeing.” What appears to us as “seeing” is a causal event with several distinct elements operating in a series. As he argues elsewhere, momentariness means that the visual object is actually already gone – it is in the past – in the moment the visual consciousness arises, so there is no question of the awareness acting on the object. When the opponent asks for clarification on whether, then, it is cognition that “sees,” Vasubandhu reacts with exasperation, and points out that the opponent is seeking precision with regard to merely conventional language: On this, the Sautrāntika says: Why carve the ether? For, conditioned by the eye and forms, the eye consciousness comes about. In that case, what sees, and what is seen? For it is passive (nirvyāpāra), merely dharmas, and merely cause and effect. With regard to this, figurative terms are used by choice with a conventional meaning: “Eye sees, consciousness cognizes” – one should not be attached to them. For the Lord said, “Do not be attached to the popular etymology, nor rush to accept the world’s ideas.”9 This appeal to the notion that what appears to be agent and action is really “merely dharmas, and merely cause and effect” is one of Vasubandhu’s key philosophical moves for interpreting scriptures in ways that generate unnecessary entities and relations out of grammatical conventions. Possession, agency, and temporal identity are described as merely figurative, common-sense linguistic 273

Jonathan C. Gold

constructs that have been reified into ontologies. In chapter two of the Treasury Commentary, Vasubandhu argues that if you start accepting elements this way, you end up reifying entities like “number, size, separation, conjunction, disjunction, distance, nearness, and existence,” entities that he insultingly designates “non-Buddhist conceptual constructions” (tīrthakaraparikalpita).10 In the Treasury Commentary’s final chapter, it is the notion of a real but ineffable “Person” ( pudgala) that is the target of this criticism. A variety of scriptures are passed back and forth until the Personalist claims that one must acknowledge and make sense of how the Buddha spoke of a person transmigrating and taking on the aggregates of a new life. According to the Personalist, the “person” takes on the aggregates like a person “becomes a grammarian” or “becomes a monk” by taking on a new status or quality. It is in response to this that Vasubandhu cites the following quotation which proves, again, that the causal action of the aggregates must be understood to be operating without any additional agent: Monks, there is action and there is fruit, but no agent is perceived who casts off these aggregates and appropriates other aggregates, because this is counter to the stipulated meaning (saṃketa) of “dharma.”11 This very quotation is used in chapter three as a response to the very same issue (against a different opponent). There, he goes on to argue: In this situation, the stipulated meaning of “dharma” is just what is dependently originated – elaborated as “when this exists, that arises.”12 Again, it is causality – the rule-governed, patterned transition from one set of elements in one moment to another in the next – that is the foundation of the teachings and the world that the teachings describe. Both are encompassed in the term “dharma.” For Vasubandhu, the doctrine that all things are dependently originated serves as the foundational interpretive principle for all of the Buddha’s teachings. In fact, since the teachings themselves are dependently originated, based in the many encounters the Buddha had with a wide variety of disciples, the language is often not only “figurative” but requiring a secondary reading as containing a “special intention” behind a basic teaching for a lesser disciple. This was why, for instance, the Buddha sometimes spoke of a self being reborn – not because he did not really believe in no-self, but to prevent people who were irreparably attached to the concept of a self from thinking that there would be no afterlife consequences of their actions. The principle of dependent origination, explained with the expression, “when this exists, that arises,” brings us full circle to the argument at the opening of the chapter that established the structure of inference and denied the possibility of self. With the nature of dependent origination in mind, we can read Vasubandhu’s opening argument to be saying not just that there is no causal basis for a self, but that the very idea of a self is counter to the key doctrine of dependent origination. To express things causally in a way that undermines false reification is the quintessence of the Buddha’s Dharma.

Mind-Only, Three Natures, and Scripture (Again) in Vasubandhu’s Yogācāra Three principles that we have revealed in Vasubandhu’s Treasury Commentary point toward named, thematized concepts in Mahāyāna. The causal view of perception, whereby sensory

274

Vasubandhu

organs and – especially – the mind are denied “agency” with regard to their respective objects and are understood as dependently originated, is called “mind-only.” The distinction between real causal entities on the one hand, and the false way that things appear – which is generated by and superimposed upon those causal entities – is explained in terms of “three natures.” The necessity that scripture be taken as largely “figurative,” with its truths only properly understood through a reasoning that sees the dependently originated nature of all things, is referred to by declaring all dharmas “figurative” and ultimate reality “ineffable” (acintya). These parallels might make one think that Vasubandhu was a Mahāyāna advocate all along, and that is a position held by some scholars today (Kritzer 2005). But Vasubandhu makes a point of arguing for these positions, and this allows us to follow a process of development into Yogācāra. The most important doctrinal innovation that distinguishes the Mahāyāna from mainstream Buddhism, apart from the adoption of the bodhisattva path, is the claim that all entities (dharmas) are as “empty” of reality as the self. Vasubandhu’s Twenty Verses on Appearance Only is the most influential philosophical defender of this view in its Yogācāra form. At the opening of the work, he states the position: “In the Mahāyāna the threefold world is defined as nothing but appearance” – and appearance is said to be equivalent to “mind” or “consciousness,” together with the qualities that minds can have.13 Everything non-mental is as unreal as the illusions of hairs created by ailing eyes. The work is dedicated to proving this view and explaining its consequences. I can only sketch the argument here, but we will see that the Twenty Verses extends the ideas established in the Abhidharma analyses previously discussed and grounds the conceptual system of the Yogācāra worldview within these principles. At the start of the argument, Vasubandhu allows an objector to point out how, in various ways, the world seems to be quite different from illusions. Specifically, things in the world have spatiotemporal and intersubjective reality, and action is effective. That seems not to happen with illusions. Vasubandhu points out that all of this could still be true of an illusory world; all of those qualities are sometimes true even in dreams. Changing tactics, then, the objector turns to scripture and asks why, if everything is only a dream, the Buddha spoke of the material spheres. In reply, Vasubandhu appeals to the notion of “special intention” mentioned earlier. The aggregates and spheres are useful, he says, for countering the false view of self. Those who are attached to the notion of self can see through this explanation that selfhood is merely a superimposition on the spheres. But this is only the conventional, surface meaning of the doctrine. The Buddha spoke as if a real relation of perceiver and perceived obtained between sensory organs and their objects. The real meaning of the sensory organs and objects – what the Buddha was speaking of when he listed these pairs – can be known by advanced practitioners: It is these two – the seed from which an appearance arises, and what appears in the appearance – that the Sage spoke of as the twofold spheres. (9)14 Sensory organs and their objects (“the twofold spheres”) are properly understood in a causal sequence as, respectively, the “seed” of perception and the fruit that arises from that seed. The formulation here is initially confusing. How is it that the sensory object – such as a visual object – is the result in the causal story of sensation, rather than its source? Ordinarily, we would place the visual object first, along with the visual organ, and the contact of these would cause a visual consciousness. Here the key is noticing that the result named as the reality behind the Buddha’s talk of an “object” is not a “thing” seen, but the visual object as it appears within an experienced perception. The causes of a sensory experience, whatever they are, bring about some appearance,

275

Jonathan C. Gold

some consciousness, and that experience has content; it has what we called earlier the “information” within a consciousness. The parallel with the view of the spheres in the “Why carve the ether?” passage discussed previously is quite close. There he argued that it is not the eye but the consciousness that “sees,” and an unspoken but implicit conclusion was that although perception seems to be taking place in perceptual organs, in fact to the extent that there is actually perception, it is mental (see Gold 2015a, 80). Here, in an effort to explain the Buddha’s language of sensation in a mind-only perspective, the argument is made that in reality there is no need for a real sensory object appearing to awareness; perception only requires the appearance of a sensory object. The sensory object is therefore the causal result of whatever dreams or illusions brought it about. The metaphor of a seed as cause of mental events will be addressed in what follows. It may be useful to pause to notice the parallels here between the reinterpretation of doctrine under an indirect reading and the reduction of agency to causality. In the earlier passage, the goal was to prove that there was no such thing as “agency” ascribable to an eye; the eye does not “see.” And in order to prove that, Vasubandhu appealed to the “figurative” nature of the Buddha’s scriptural language. Here, the example of the unitary causal line is used to justify the idea that the Buddha’s apparently literal language about sensory “agency” is in fact concealing a viable “special intention.” Among the more contentious questions in recent Buddhist philosophy has been the debate over whether Yogācāra philosophers were making an ontological claim with their mind-only doctrine, or simply an epistemological one. It is clear that Vasubandhu’s Twenty Verses is an ontological argument; one of its central arguments disproves materiality. Vasubandhu runs through a series of arguments that prove, to his satisfaction, that no consistent account of materiality is possible, which proves, once again, that the Buddha could not possibly have truly believed in materiality. But his parallel argument from the Treasury Commentary discussed previously would seem to be arguing the epistemological version of the argument, saying that even when we assume that there are material sensory organs and objects, our experience of them is always only mental. It is possible, once again, to read the Treasury Commentary as teeing up Abhidharma for the Twenty Verses’ drive into the Yogācāra position. After the opponent’s defenses of materiality are spent, a final objection is raised as to the karmic, which is to say moral, significance of reality under a mind-only perspective. Vasubandhu says that there is no reason why causal relations and moral significance cannot be effected in a mental mode. But in the end, he says, the reality of things under the Mahāyāna perspective is known only by Buddhas; it is “beyond the scope of reason.”15 This is a remarkable statement to find at the end of an elegant, compactly argued work by one of the great philosophers of all time. (In this, it resembles nothing if not Wittgenstein’s Tractatus.) Yet, it expresses not defeat but the doctrinal position of the ultimate ineffability of reality in the Yogācāra. The conclusion of the Thirty Verses on Appearance (Triṃśikāvijñaptikārikā), which is considered Vasubandhu’s companion work to the Twenty Verses, explains that the mind-only position is only the first in a two-step process that leads to the attainment of Buddhahood.16 The final stage entails a relinquishment of all cognition: “It is without thought, without apprehension: and that is supramundane awareness.”17 The awareness of Buddhas is an inconceivable state caused by, but beyond, the attainment of the state of mind-only. Like the Twenty Verses, then, Vasubandhu’s Thirty Verses on Appearance (Triṃśikāvijñaptikārikā) culminates in the ineffable. The Thirty Verses also opens like the Twenty Verses, indicating that things are not as they seem. It says that selves and entities (dharmas) – that is, things we can name – function to the extent that they do only within the illusory play of experience called the “transformation of consciousness.” This seems to parallel the earlier discussions. But a key aspect seems to have been added when the verse says that things exist figuratively:

276

Vasubandhu

A manifold figurative predication of “self ” and “things” is what functions with respect to the transformation of consciousness.18 The Thirty Verses thus begins with a linguistic explanation of the doctrine of mind-only. Is this something new? In fact, it may be the Twenty Verses, with its express interest in arguments about the material world that ground mind-only in Abhidharma disputes, that is new. Before turning to the Thirty Verses, then, let us look at other, earlier Yogācāra scriptures and treatises – upon many of which we have commentaries attributed to Vasubandhu. There we see that one of the most important themes of this literature was to establish a view of language that would make sense of the Buddha’s teachings in light of universal emptiness. One of the foundational Yogācāra scriptures is the Scripture Clarifying the Intent (Saṃdhinirmocanasūtra), which introduces a hermeneutic structure by which the Buddha’s teachings may be classified into “Three Turnings of the Dharma Wheel.”19 The idea is that the Buddha taught doctrines in three categories for beings of different capacities, and associated with the three turnings are, respectively, three natures or three qualities of reality: the first turning is associated with the traditional reading of doctrines as expressing a kind of surface meaning, which is still an expression of the emptiness of ordinary reality; the second turning is associated with the Mahāyāna scriptures that declare the emptiness, or negation, of the principles and entities that populate the first turning; and the third and final turning is associated with a recognition that all realities, including the expressions of the first and second turnings, are inconceivable. Each of the three natures, then, represents a particular turn of the wheel of the doctrine, understood as a distinct conventional expression of the same basic inexpressible reality, which is emptiness: the Buddha taught the same emptiness in three different forms. This doctrine of the three approaches to the nature of emptiness, also called the three natures, serves as an explanation of the Buddha’s teachings as unified, but multi-faceted – like a three-sided crystal. The three natures are three ways that the Buddha’s teachings have emerged, in answer to the three categories in which the doctrinal needs of sentient beings can be satisfied. Since the Dharma has these three natures, there is no single way to understand any given doctrine in ordinary language. A theme of the Scripture Clarifying the Intent is that disputation over doctrines is always foolish, since those who see things rightly (“practitioners of yoga,” yogācāras) know that all views are illusory. In the treatises of the Yogācāra, this multi-faceted view of reality, of truth, provides a method for interpreting the Dharma in all of its aspects. For instance, in the Commentary on Distinguishing the Middle from the Extremes (Madhyāntavibhāgabhāṣya), one of those commentaries attributed to Vasubandhu, the three natures serves exactly this purpose, as a method for explaining and integrating different approaches to Buddhist doctrine – different facets of each doctrinal module.20 Chapter three works this out in great detail under the rubric of “reality” or “truth” (tattva). The fundamental root reality described in verse 3.3 explains the threefold nature in terms of existence: the first nature is truly non-existent, the second is existent but not truly existent, and the third is truly both existent and non-existent. The three natures should be understood as eternally non-existent, falsely existent, and truly existent and non-existent. (3.3b–d)21 After establishing this rubric, the chapter runs through every category and system in Buddhist doctrine, explaining how all three of these aspects are present in every teaching. Just to give

277

Jonathan C. Gold

one example that will resonate with what we have discussed above, the Buddha’s teaching of how the false view of self is projected onto the five aggregates is explained as enacting teachings of all three kinds of emptiness, all three turnings: It has the meanings of multiplicity, being thrown together, and being separated. (3.17a–b)22 The Distinguishing Commentary explains “multiplicity” as the teaching of “non-existence” in the first turning by saying that the teaching of the aggregates shows how the self lacks the unity it appears to have. This makes sense of early Buddhist doctrine as a kind of teaching of “emptiness” of the self. The second turning, expressed as “being thrown together,” is glossed as providing a causal description that tells how the appearance of the self comes about from its parts, and is therefore falsely existent. Here, dependent origination – specifically, dependent origination of the appearance of self based in the aggregates, is a narrative of how false natures come to appear to exist. The third turning, expressed as “being separated,” makes the point that the separation of the aggregates, which follows from that fact that they came together through causes and conditions, leaves a background fundamental nature of emptiness, which is the existent truth of the non-existence of the self in each false construction. This is why the third turning is said to be both existent and non-existent.23 This is complex and deserves further analysis; indeed, the whole chapter needs careful analysis. But it is already clear from the present example that, when read in this mode, the teaching of the aggregates, which is basic Buddhism, becomes not just a teaching on no-self but also emptiness, and even more than that: when we take all three together, it becomes a teaching on the inconceivable, simultaneous nature of appearance and emptiness. To clarify why Buddhists would want to say that all of the teachings are, in this way, ultimately inconceivable, I would note an interesting side point from this same commentary (again, attributed to Vasubandhu) on the fifth verse of the fifth chapter. Chapter five explains the bodhisattva path, and the fifth verse lists the ten perfections. The final perfection on the tenth bodhisattva stage is awareness (  jñāna), and the commentary tells us that this awareness consists in transcending the delusion that comes from adhering to the letter of the doctrine.24 Of course, we need to know the meaning and not the letter – but this is the last stage of the path. How can a bodhisattva who never understood the meaning have ever gotten that far? I hope the reader is already thinking of Vasubandhu’s view that doctrines need to be taken as figurative, and an overly literal reading often leads one astray. In this case, though, on the tenth bodhisattva level, we can see the culmination of this skepticism about the word of the dharma expressed by the quintessential tenth-level bodhisattvas Avalokiteśvara and Mañjuśrī. In the Heart Sutra, Avalokiteśvara speaks of his vision which denies the literal meanings of the doctrine: there is “no eye, no ear, no nose, etc.” The final stage bodhisattva, with wisdom perfected, sees that the dharma is not to be taken literally. A key motivation for what we call early Yogācāra – the third turning of the dharma wheel – was to counter the delusion that arises from an overly literal reading of scripture. This rehabilitates the doctrines of early Buddhism against the potential onslaught of nihilistic readings that use Avalokiteśvara’s and Mañjuśrī’s scriptural negations of dharmas as a justification for discounting their utility. Until we are Buddhas, we cannot hold all of these perspectives in our awareness at once; until the tenth bodhisattva level, we are attached to the literal meanings of the teachings, and we need to use language – in particular the language of the Dharma – as a crutch. But that does not discount the fact that from the ultimate perspective, all of the Buddha’s teachings are unified, emanating from the same, inconceivable source, which is the vision of 278

Vasubandhu

the true nature of reality. Yogācāra texts provide creative, alternative structures and strategies to display the Dharma’s kaleidoscopic multiplicity of meanings and preclude any single, too direct, reading of the Buddha’s teachings – whether it is an overly literal set of non-Buddhist conceptual constructions, or a simplistic, nihilistic reading of the perfection of wisdom doctrine that might lead one to neglect the beneficial “emptinesses” taught in the first turning.25 This is our context, then, for the opening statement in the Thirty Verses, which describes every­thing as merely figurative (upacāra). Vasubandhu’s declaration ties the figurative nature of scripture that we have seen to be a natural method for interpreting the teachings with the wider, Yogācāra principle that all of the Buddha’s teachings serve their purposes within the nexuses of constructed mental appearances of sentient beings. What seem to be selves and dharmas at the level of appearance are actually based in an ever-changing flow of consciousness. From there, the work lays out the key terms of Yogācāra doctrine under this rubric. Consciousness is constantly transforming itself in, specifically, three different ways. The first kind of conscious flow is what is called the store consciousness (ālayavijñāna). This is the subconscious and equanimous mentation that, according to the text, flows like a river, and contains the seeds of the other transformations (verses 2–4). I mentioned we would see the “seed” metaphor again, and here the “store consciousness” is the “store of seeds” of the other kinds of consciousness. The second kind of transformation is the mental consciousness (mano-vijñāna), which takes the first as its mental object, and is associated with the afflictions and the view of the self. This is the kind of mental flow that is eliminated by those at the stage of stream-enterer or above; in the transcendent (lokuttara) mind, there is no mental organ (manas) (5–8a). When this “mental consciousness” is said to take the “seeds” of the “storehouse consciousness” as its mental object, we have a direct recapitulation of the ninth verse of the Twenty Verses discussed previously as an exemplification of the unitary causal line. The fundamental error is here described as the activity of the organ that believes itself an agent of perception, a subject of experience. The third kind of mental flow, finally, is the awareness of sense objects – and in this context, Vasubandhu provides a catalog of the full efflorescence of mental characteristics: virtuous and nonvirtuous, subtle and refined (kleśa and upakleśa) (8b–14). Thus, the detailed analysis of the causes and conditions of the flow of consciousness provides the structure for the first half of the text. The middle of the text then integrates these three kinds of mental transformation, and shows how they are interconnected. Verse 15, which is the opening of the middle section, brings together the first and the third of the kinds of mental transformation: The five consciousnesses arise according to conditions in the root consciousness Together or not, like waves in water. (15)26 Sensory awarenesses can occur together or separately: you can have a smell and a taste together, or just have one – and what you get depends upon what is happening in the root mind. The currents in the water generate one or more waves. The analogy of waves in water expresses a relationship between the root mind and sensation that is both cause-and-result and something more. Because as much as waves come from the water, they also are the water. The waves are a description of the state of the water itself – that it is wavy. At the very least, they are not separate from it. So to say that sensory objects can be separate or together is to allow for different sensory occurrences, but also to point out that sensation is a matter of perspective. Counting waves is an act of constructive isolation, an activity of picking out and naming components from what is an undifferentiated, ongoing flow. The five sensory consciousnesses, which are the awarenesses of sensation, arise from – but are also a part of – the root mind. This interconnectedness is a key, 279

Jonathan C. Gold

sensory exemplification of the doctrine of non-duality in Yogācāra. It is an accurate description of – a view of – what the deluded “mental consciousness” imagines and constructs about perception – the view as if – the mind were the “perceiver” and the sensation the “perceived.” In the next two verses, Vasubandhu says that, unlike the sensory consciousnesses, the mental consciousness is omnipresent in the ordinary waking mind and is always engaged in the discrimination, the selective reification, of what appears. That is to say, mental consciousness is always counting waves and failing to notice that the waves just are the water. This leads to the next verse, where this same arising of appearances in the mind is given a more technical analysis by calling it “construction” (vikalpa) and attributing it to the mutual influence of the “seeds” that make up the root consciousness in its various transformations: The consciousness that holds all the seeds is a transformation, this way and that, Which proceeds through mutual influence, by which this and that construction comes about. (18)27 This verse is discussing the same thing as verse 15, saying both that there are causes that bring about appearances, and that those appearances are not properly separate from the root mind of which all of them consist. Here, the point is that all the seeds that make up the “store” of seed consciousnesses (ālayavijñāna) undertake transformations by influencing one another in countless ways, and it is this mutual influence itself that brings about constructions, however they appear. The mutual influence of these consciousnesses just is the constructions. There is nothing else that comes about over and above those consciousnesses. They are the currents and waves on the water, interacting with one another, constantly forming new shapes. This is the point of the echoing “this and that” (sa sa) which picks up the “this way and that” (tathā tathā) in the first line. They are two different ways of talking about the same activity. However things come about causally, that is what is coming about as appearances. These two levels of description are parallel and interconnected. The next verse (verse 19) points out the karmic consequences of this activity of reification enacted by the mental consciousness, saying that the “grasping” that separates out waves from the water in our metaphor is what keeps saṃsāra going. It causes further maturations in the root consciousness. This is a very important point because it reveals, conversely, how the path to liberation is supposed to work. In order to get the grasping to stop, awareness has to rest in the perfected nature, which is the awareness of appearance only: you have to see that the waves just are the water. This is what happens when you, like Avalokiteśvara, attain the transcendent awareness that constitutes a stream-enterer. In the ending verses of the Thirty Verses, as we’ve already discussed, the point is made that this awareness cannot be discursive – you can’t become a stream-enterer just by thinking the thought, “this sensation is appearance only.” The transcendent (lokuttara) mind is without any apprehension; it is devoid of thoughts (acintya). The final stage is beyond conceptual knowledge. But before he gets there, Vasubandhu provides a very interesting line that suggests why discursive thought is always still tangled up with mistaken construction. The reason is that even accurate thoughts are, like sensory appearances, waves on the water. In the last half of verse 22, Vasubandhu compares the three natures, which are interconnected, with the basic doctrines of Buddhism. The complex, dizzying doctrines of Yogācāra may seem abstruse, but they share yet another quality with the basic doctrines of Buddhism: As is said of impermanence, etc., the one is not seen where the other is absent.

280

(22)28

Vasubandhu

Here the point is that impermanence, and the other central Buddhist doctrines of suffering and no-self, only appear, when they do, along with the conditioning factors that they are intended to describe and counteract. Impermanence needs to be impermanence of something  – such as a self. Impermanence is thus tied up with appearances; when no self is seen, neither is the self’s impermanence. Impermanence is a wave on the water of the self. What we take to be a Mahāyāna theme, the notion of non-duality, is here assimilated to three of the most central, definitive Buddhist doctrines. This, I take it, is one of the key points of Yogācāra thought and one that we may see expressed throughout Vasubandhu’s work. Setting aside the worry about the unity of Vasubandhu, then, we find common cause across these texts in the use of dependent origination understood in a causal mode to streamline Buddhist ontology, legitimate and coordinate Buddhist scripture, and sharpen Buddhist thinking on the interconnectedness of mind, language, and world. The project of pointing behind the curtain of our delusions – the emptiness of our false conceptions – was key to Buddhist doctrine all along. Every doctrine was meant to help us see through and counteract this or that false construction; every doctrine was meant for our benefit, figuratively.

Notes 1 The ninth chapter of the Commentary on the Treasury of Abhidharma (Abhidharmakośabhāṣya) has been translated into English from the Sanskrit by Matthew Kapstein (2001a), James Duerlinger (2003), and Charles Goodman (2009). The French translation of the full text from Chinese by La Vallée Poussin (1923–1925) was translated into English by Leo Pruden (1988–1990) with reference to the Sanskrit. See Gold (2011; 2015a) and sources noted therein for fuller treatments of Vasuban­ dhu’s arguments mentioned in this chapter. 2 Translations of the Twenty Verses (Viṃśikā) in English are available by Kapstein (2001b), Anacker (1998), and Silk (2016) and in French by La Vallée Poussin (1928–1929). 3 Pradhan (1975, 461.2–20). 4 In chapter nine of the Treasury Commentary, Vasubandhu in fact cites several passages to prove, against the Personalist, that the Buddha declared these systems to be comprehensive (Pradhan 1975, 464.20–66.17). 5 Pradhan (1975, 461.8–9). 6 Ibid., 79.11–12. 7 This is Goodman’s translation (2009, 304). 8 I should acknowledge that I differ from most interpreters in intentionally neglecting the well-known distinction between the “Dharma” as the Buddha’s teachings and the dharmas as elements in lists in Abhi­ dharma philosophical analyses. It is clear that both terms refer to the fundamental principles or realities that the Buddha sees, understands, and articulates in various forms. To distinguish the Dharma from dharmas is to hypothesize that the Buddha’s teachings might be expressions of something other than fundamental truths, or perhaps to confuse genres such as Sūtra and Vinaya genres with the Dharma expressed through them. In any case, Vasubandhu’s definition of abhidharma (Pradhan 1975, 2.3–2.10) refers to the fundamental reality and its analysis in treatises like his own, but not its expression in lists or teachings. 9 Pradhan (1975, 31.12–16). 10 Ibid., 79.24–25. 11 Ibid., 468.23–24. 12 Ibid., 129.11–12. The actual citation glossed there is Pradhan (1975, 129.10–11). 13 Silk (2016, 30). 14 Ibid., 67. My translations of the Twenty Verses and the Thirty Verses are greatly indebted to the ongoing partnership of the Vasubandhu Translation Group (Mario D’Amato, Richard Nance, Parimal Patil, Trina Jones, Dan Arnold, and myself). 15 Ibid., 145: tarkkāviṣaya. 16 For translations of the Thirty Verses, see Anacker (1998) for English, and Lévi (1932) for French. 17 Buescher (2007, 138): acitto ’nupalambho ’sao jñānaṃ lokottarañ ca tat |. 18 Ibid., 40: ātmadharmopacāro hi vividho yaḥ pravartate | . . . vijñānapariṇāme ’sau . . . Note: I am not discussing Sthiramati’s commentary here.

281

Jonathan C. Gold 19 Translated into English in Powers (1995), into French in Lamotte (1935). My discussion focuses on chapter seven. 20 See D’Amato (2012) for an English translation of the Distinguishing the Middle from the Extremes, together with Vasubandhu’s commentary. 21 Nagao (1964, 38.1–2): asac ca nityaṃ sac cāpy atattvataḥ | sad-asat-tattvataś ceti svabhāva-traya iṣyate || III.3 22 D’Amato (2012, 153); Nagao (1964, 45.1): anekatvābhisaṃkṣepa-paricchedārtha. 23 Nagao (1964, 45.2–5); D’Amato (2012, 153): “With regard to the first” refers to the aggregates, understood according to three meanings. With respect to the meaning of “multiplicity”: as it is said, “Form is dispersed across the past, present and future”[;] hence it is not a unity. With respect to the meaning of “being thrown together”: all of the aggregates are thrown together at once hence they are not a unity. And the aggregates may also be understood in terms of the meaning of “being separated,” through laying out the characteristics of form, feeling, conceptualization, dispositions, and consciousness according to their respective fundamental natures wherein they would be seen as distinct. 2 4 D’Amato (2012, 176); Nagao (1964, 62.18): yathāruta-dharmma-saṃmohāpagamāt. 25 For a fuller exposition of the point of these two paragraphs, see Gold (2015b). 26 Buescher (2007, 102). 27 Ibid., 110. 28 Ibid., 126.

References Anacker, Stefan. (1984) 1998. Seven Works of Vasubandhu, the Buddhist Psychological Doctor. Corrected Edition. New Delhi: Motilal Banarsidass. Buescher, Hartmut. 2007. Sthiramati’s Triṃśikāvijñaptibhāṣya: Critical Editions of the Sanskrit Text and its Tibetan Translation. Philosophisch-Historische Klasse Sitzungsberichte 768. Vienna: Verlag der Österreichische Akademie der Wissenschaften. ———. 2013. “Distinguishing the Two Vasubandhus, the Bhāṣyakāra and the Kośakāra, as YogācāraVijñānavāda Authors.” In The Foundation for Yoga Practitioners: The Buddhist Yogācārabhūmi Treatise and Its Adaptation in India, East Asia, and Tibet, edited by Ulrich Timme Kragh, 368–96. Cambridge, MA: Harvard University Department of South Asian Studies and Harvard University Press. D’Amato, Mario. 2012. Maitreya’s Distinguishing the Middle from the Extremes (Madhyāntavibhāga) Along with Vasubandhu’s Commentary (Madhyāntavibhāga-bhāṣya): A Study and Translation. Treasury of the Buddhist Sciences. New York: American Institute of Buddhist Studies. Duerlinger, James. 2003. Indian Buddhist Theories of Persons: Vasubandhu’s “Refutation of the Theory of a Self.” London: RoutledgeCurzon. Frauwallner, Erich. 1951. On the Date of the Buddhist Master of the Law Vasubandhu. Rome: Istituto Italiano per il Medio ed Estremo Oriente. ———. 1961. “Landmarks in the History of Indian Logic.” Wiener Zeitschrift für die Kunde des Süd- und Ostasiens 5: 125–48. Gold, Jonathan C. 2011. “Vasubandhu.” In Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta. https://plato.stanford.edu/entries/vasubandhu/. ———. 2015a. Paving the Great Way: Vasubandhu’s Unifying Buddhist Philosophy. New York: Columbia University Press. ———. 2015b. “Without Karma and Nirvāṇa, Buddhism is Nihilism: The Yogācāra Contribution to the Doctrine of Emptiness.” In Madhyamaka and Yogācāra: Allies or Rivals? edited by Jay L. Garfield and Jan Westerhoff, 213–41. New York: Oxford University Press. Goodman, Charles. 2009. “Vasubandhu’s Abhidharmakośa: The Critique of the Soul.” In Buddhist Philosophy: Essential Readings, edited by William Edelglass and Jay L. Garfield, 297–308. New York: Oxford University Press. Kapstein, Matthew T. 2001a. “Vasubandhu and the Nyāya Philosophers on Personal Identity.” In Reason’s Traces, 347–91. Somerville, MA: Wisdom Publications. ———. 2001b. “Mereological Considerations in Vasubandhu’s ‘Proof of Idealism’.” In Reason’s Traces, 181–204. Somerville, MA: Wisdom Publications.

282

Vasubandhu ———. 2018. “Who Wrote the Trisvabhāvanirdeśa? Reflections on an Enigmatic Text and Its Place in the History of Buddhist Philosophy.” Journal of Indian Philosophy 46: 1–30. Kritzer, Robert. 2005. Vasubandhu and the Yogācārabhūmi: Yogācāra Elements in the Abhidharmakośabhāṣya. Studia Philological Buddhiaca XVIII. Tokyo: International Institute for Buddhist Studies. La Vallée Poussin, Louis de. 1923–1925. L’Abhidharmakośa de Vasubandhu, 6 vols. Société Belge D’Études Orientales. Paris: Geuthner. ———. 1928–1929. Vijñaptimātratāsiddhi: la siddhi de Hiuan-Tsang, traduite et annotée par Louis de la Vallée Poussin. Paris: P. Geuthner. Lamotte, Étienne. 1935. Saṃdhinirmocana sūtra: l’explication des mystères. Louvain: Université de Louvain. Lévi, Sylvain. 1932. “La Trentaine.” In Matériaux pour l’étude du système Vijñaptimātra, 61–123. Paris: Librairie Ancienne Honoré Champion. Nagao, Gadjin M., ed. 1964. Madhyāntavibhāga-bhāṣya: A Buddhist Philosophical Treatise Edited fror the First Time from a Sanskrit Manuscript. Tokyo: Suzuki Research Foundation. Olivelle, Patrick, trans. 1996. Upaniṣads. Oxford World’s Classics. New York: Oxford University Press. Park, Changhwan. 2014. Vasubandhu, Śrilāta, and the Sautrāntika Theory of Seeds. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien, Universität Wien. Powers, John, trans. 1995. Wisdom of Buddha: The Saṁdhinirmocana Mahāyāna Sūtra; Essential Questions and Direct Answers for Realizing Enlightenment. Berkeley, CA: Dharma Publishing. Pradhan, Prahlad, ed. (1967) 1975. Abhidharmakośabhāṣyam of Vasubandhu. Tibetan Sanskrit Works 8. Patna: K. P. Jayaswal Research Institute. Pruden, Leo, trans. 1988–1990. Abhidharmakośa Bhāṣyam by Louis de la Vallée Poussin. Berkeley, CA: Asian Humanities Press. Silk, Jonathan A. 2016. Vasubandhu’s Viṁśikā (I): Sanskrit and Tibetan Critical Editions of the Verses and Autocommentary; An English Translation and Annotations. Harvard Oriental Series 81. Cambridge, MA: Harvard University Department of South Asian Studies and Harvard University Press [Open Access Edition 2018]. Skilling, Peter. 2000. “Vasubandhu and the Vyākhyāyukti Literature.” Journal of the International Association of Buddhist Studies 23 (2): 297–350. Takakusu, J., trans. 1904. “The Life of Vasu-bandhu by Paramārtha (A.D. 499–569).” T’oung Pao 5: 269–96. Von Rospatt, Alexander. 1995. The Buddhist Doctrine of Momentariness: A Survey of the Origins and Early Phase of This Doctrine Up to Vasubandhu. Alt- und Neu-Indische Studien 47. Stuttgart: Franz Steiner Verlag.

283

17 DIGNĀGA Early Innovator in Buddhist Epistemology1 Kei Kataoka

Introduction The Buddhist logician Dignāga (c. 480–540/470–530 CE)2 is one of the foremost figures to introduce full-fledged philosophical discussions of the means of valid cognition ( pramāṇa) in Indian philosophy. A difficulty in studying Dignāga, however, is that many of his works are not available in the original Sanskrit.3 Scholars therefore have studied his texts mainly through Chinese and Tibetan translations, as well as some Sanskrit fragments. Portions of several of his works have also been reconstructed using various forms of evidence. Dignāga’s final and most influential work is the Compendium of the Means of Valid Cognition (Pramāṇasamuccaya, PS, hereafter the Compendium), a verse treatise on which he also has written an autocommentary (PSV). The recent discovery of Sanskrit manuscripts of the Indian sub-commentary on the Compendium, namely that of Jinendrabuddhi (PSṬ), has been particularly fortunate in that it has made possible the reconstruction of a good portion of Dignāga’s most famous work. Among the Compendium’s six chapters, the first and the fifth chapters are now restored in Steinkellner (2005) and Pind (2015), respectively.4 Dignāga himself describes Compendium as based on his own previous writings including his Gateway to Logic (Nyāyamukha),5 a work on dialectics clearly modeled on his predecessor Vasubandhu’s Method for Argumentation (Vādavidhi). A comparison of the chapter structure of the Compendium to that of the Gateway indicates that Dignāga reorganized his philosophical system to prioritize the presentation of the means of valid cognition ( pramāṇa) – that is, epistemology – rather than dialectics or the art of debating with others (vāda), which is concerned primarily with proof (sādhana) and refutation (dūṣaṇa). This new focus can be seen by considering the chapter structure in the Compendium, where the six chapters deal with the following topics: (1) perception ( pratyakṣa); (2) inference-for-oneself (svārthānumāna); (3) inference-for-others ( parārthānumāna); (4) example (dṛṣṭānta); (5) the exclusion of what is other (anyāpoha); and (6) sophistic rejoinder ( jāti). The emphasis on the means of valid cognition is clear by their priority in the order of presentation. In his Compendium, Dignāga emphasizes that he accepts only two means of valid cognition, namely perception ( pratyakṣa) and inference (anumāna). He then further subdivides inference into two kinds, inference-for-oneself and inference-for-others, thus enabling him to treat argumentation under the heading of inference. He reduces verbal testimony (śabda) to inference 284

DOI: 10.4324/9781351030908-26

Dignāga

and thus denies its independent status as a separate means of valid cognition. Finally, he treats polemical or sophistic techniques called jāti in an appendix to inference. In this way, Dignāga unifies three domains in his Compendium: epistemology ( pramāṇa-tradition), dialectics (vādatradition), and semantics (the theory of word-referents). Understanding proof and verbal testimony as forms of inference constitute the hallmark of his philosophical genius and have had a lasting influence in Buddhist philosophy. Beyond this, his particular contribution lies in treating important received ideas in a well-organized, consistent system of epistemology rather than inventing insightful concepts out of nothing. Indeed, scholars have successfully located the seeds of what seem, at first sight, to be Dignāga’s quite original ideas in earlier works of the Abhi‑ dharma, Sautrāntika, Yogācāra, and grammarian traditions. Vasubandhu was the most influential previous thinker for Dignāga, though this did not prevent Dignāga from critically engaging with him.6 Some of Dignāga’s insights can also be traced back to the important grammarian Bhartṛhari (Ogawa 2017); indeed, many of Bhartṛhari’s verses found their way into Dignāga’s Investigation of the Three Times (Traikālyaparīkṣā) (Frauwallner 1959). Given that its explicit goal is both to refute theories about the means of valid cognition elaborated by other scholars and to demonstrate the correctness of his own theories, Dignāga’s Compendium offers insight into the epistemological debates of his time. A textual analysis of the structure of the first chapter, for example, illustrates Dignāga’s strategy in which he first elucidates his own doctrine and then refutes other presentations. These other presentations derive from, in order: (1) Vasubandhu’s Method for Argumentation (Vādavidhi); (2) the Aphorisms on Logic (Nyāyasūtra) attributed to Akṣapāda Gautama; (3) the Aphorisms of Vaiśeṣika (Vaiśeṣikasūtra) attributed to Kaṇāda; (4) the Manual of Sixty Principles (Ṣaṣṭitantra) attributed to Vārṣagaṇya together with the views of Dignāga’s elder contemporary, the Sāṁkhya theorist Mādhava; and (5) the Aphorisms of Hermeneutical Inquiry (Mīmāṁsāsūtra) of Jaimini, together with the views of a “commentator,” probably Bhavadāsa. Dignāga’s Compendium provoked strong positive and negative reactions from later scholars, whether from inside or outside the Buddhist tradition; among works by the latter, Uddyotakara’s Commentary on Logic (Nyāyavārttika) and Kumārila’s Commentary in Verse (Ślokavārttika) are the most representative. Answering Kumārila’s severe criticism of Dignāga is probably one of the main reasons why Dharmakīrti reorganized Dignāga’s system according to his own principles.7

Types of Cognition Dignāga divides cognition into two basic types: nonconceptual and conceptual. The first of these is equivalent to perception ( pratyakṣa), which Dignāga defines as cognition “devoid of conceptualization” (kalpanāpoḍha). In the Compendium (PS I 4–6), he recognizes four distinct kinds of perception (Katsura 1984; Funayama 2000). These are: (i) perception by the five sense-faculties; (ii-a) mental perception that deals with external objects such as color/form; (iib) mental perception, a kind of self-awareness, that deals with internal feelings such as desire; (iii) yoga practitioners’ pure intuitive perception that is not mixed with scriptural conceptualization; and (iv) cognition’s self-awareness, including the case of conceptual cognition. As for conceptual cognition, Dignāga recognizes three kinds, all of which are mental (mānasa), meaning that they are primarily caused by a mental organ (manas). According to Dignāga, all cognitions, whether nonconceptual or conceptual, are self-aware (svasaṃvedana) with regard to themselves, because even a conceptual cognition of an inferred fire, for example, has direct access to the cognition itself. The Yogācāra scheme of the three natures (trisvabhāva) – i.e., (1) the perfected nature ( pariniṣpanna-svabhāva); (2) the dependent nature ( paratantra-svabhāva); and (3) the 285

Kei Kataoka

imagined nature ( parikalpita-svabhāva) – can be used as a framework for understanding Dignāga’s classification of different types of cognition.8 This is true even though explicit reference to the three natures is absent in the Compendium, which can be regarded primarily as a Sautrāntika work. The most crucial difference between the Yogācāra and the Sautrāntika schools is that the latter accepts external reality, whereas the former holds that there is nothing outside consciousness (i.e., everything is “consciousness-only,” or vijñāna-mātra). Additionally, Sautrāntikas do not posit three natures but rather only two: (1) a substantially existent (dravyasat) reality, which they regard as ultimately true; and (2) a nominally existent ( prajñaptisat) reality, which they regard as only conventionally true or true merely as designation. If we consider the Sautrāntika position in terms of the three natures, then, we might say that Sautrāntikas distinguish the dependent (substantially existent) and the imagined (nominally existent) natures but leave aside any explicit treatment of the perfected nature. Nevertheless, while the three natures play only an implicit role in most of Dignāga’s writings,9 they still can be a useful tool for considering his general approach to the types of cognition since we know that Dignāga embraced a Yogācāra perspective in some of his work. Here we can take some help from the later Buddhist philosopher Ratnākaraśānti (c. 1000 CE), who interprets the three natures respectively in his Instruction on the Perfection of Wisdom (Prajñāpāramitopadeśa, or PPU) as: (1) that which exists ultimately ( paramārthasat); (2) that which exists substantially (dravyasat); and (3) that which exists merely as designation ( prajñaptisat).10 Although it may seem obvious that the three natures of the Yogācāra system correspond to these three levels of existence beginning with paramārthasat, previous scholars of the Buddhist logic-epistemological tradition probably felt hesitant to explicitly state this equivalency due to the fact that only two levels are distinguished in the Sautrāntika system, where the perfected nature, i.e., consciousness-only, has no room. Taking our inspiration from Ratnākaraśānti, though, we can match the basic types of cognition recognized by Dignāga in his Compendium (i–iv discussed previously) with the three natures as Ratnākaraśānti presents them (1–3 discussed previously) as follows: (1 = iv) (2a = i) (2b = ii-a) (2c = ii-b) (3abc)

Yoga practitioners’ intuitive perception of religious truths.11 Ordinary sense perception of color and form, smell, sound, taste, and touch. Mental perception of color/form and so on that immediately follows 2a.12 Mental perception (i.e. self-awareness) of desire and so on. Various kinds of conceptual mental cognition (discussed in what follows).

When we classify Dignāga’s various types of cognition in terms of the three natures like this, we can see that the border between (1) and (2) marks the difference between sages (“those who have seen reality as it is”) and ordinary people; namely, the border between cognitions which fall into Ratnākaraśānti’s first category (1), the perfected nature, and cognitions that fall into the second category (2a, 2b, and 2c), the dependent nature. We can also see that all the cognitions in the first and second categories are nonconceptual cognitions; i.e., they are all perceptions. The line of demarcation between the second and third categories is therefore the difference between nonconceptual and conceptual cognitions. As mentioned earlier, this difference is the fundamental division in types of cognition for Dignāga. Dignāga refers to three kinds of conceptual cognition, which we can consider as three divisions within the imagined nature: (3a) various types of conceptual cognition that rely on previous experiences; (3b) the cognition of empirical reality (saṃvṛtisajjñāna) that superimposes a different form onto something else; and (3c) erroneous cognition, e.g., mistaking a mirage for “water.” In the first of these subcategories (3a), Dignāga lists inference together with its 286

Dignāga

result (ānumānika), memory, and desire. The second subcategory (3b), which does not rely on a previous experience, can probably be called perceptual judgment (Katsura 1984, 116; 1986b, 226), e.g., the judgment “pot,” which superimposes general characteristics (sāmānyalakṣaṇa) onto a particular characteristic (svalakṣaṇa), e.g., the absolutely unique instants and multiple parts that are collectively subsumed under the judgment “pot.” The third subcategory (3c) would consist of errors which do not reflect or correspond to real particular characteristics, such as mistaking a rope for a snake. An exceptional case of error – i.e., the erroneous cognition of what seems to be bundles of hair and so on due to optical floaters or cataracts (taimira) – does not involve conceptualization and therefore cannot be treated under the final subcategory (3c). This type of error is mentioned by Dignāga as a type of pseudo-perception ( pratyakṣābhāsa). Later, Dharmakīrti adds the word “non-erroneous” (abhrānta) to Dignāga’s previous definition of perception as “devoid of conceptualization” (kalpanāpoḍha) in order to explicitly exclude this inconvenient case from the category of perception (NB I 4). Returning again to our classification of types of cognition according to the three natures (1, 2abc, 3abc), we can see that Dignāga groups valid inference together with invalid conceptual cognition, such as memory, under the same category (3a). Furthermore, we note that he has difficulty handling the exceptional case of nonconceptual erroneous cognition, which does not fit easily under the category of 3c but also does not fit easily into the categories 2a, 2b, or 2c. These facts indicate all the more that the demarcation line between nonconceptual and conceptual, i.e., the line between 2 and 3, has a binding force even in his (predominantly Sautrāntika) epistemology of valid cognition that consists of perception and inference. In other words, one way of reading Dignāga is to see his main framework as based on the old system of the three natures, in which conceptualization demarcates the line between 2 and 3. But his new system of the means of valid cognition does not fit with this old scheme, because inference – although valid in the new system – has to be grouped under the third category of the old system, just like other types of invalid mental conceptualization, in as much as it is a kind of conceptual cognition. The case of optical disease is similarly embarrassing because its caused cognition is erroneous but nonetheless nonconceptual. We thus see that Dignāga represents a transitional point in the development of Buddhist philosophy in which the primary divisions of conceptual/nonconceptual and nominal/substantial do not easily line up.

Classifying Dignāga’s Works in Light of the Types of Cognition By relating Dignāga’s hierarchical classification of the various types of cognition to the Yogācāra scheme of the three natures, we can now shed light on his entire epistemological project and arrange all of his works in a logical order, including those known only by name and those extant in fragments or partial translation.13 Starting with Level 1, corresponding to the perfected nature or the ultimately existent, as well as to the perception of sages, are his Summary of Perfection of Wisdom (Prajñāpāramitāpiṇḍārtha) and Entrance to Yoga (Yogāvatāra), both of which deal with the path of cultivation for realizing the ultimate truth. Moving on to Level 2, corresponding to the dependent nature or the substantially existent, as well as to the various forms of ordinary perception, is the Lamp of Vital Points (Marmapradīpa), an abridgment of Vasubandhu’s Commentary on the Treasury of the Abhidharma (AKBh), which deals mainly with substantially existent entities. Considered in terms of the transition from Level 1 to Level 2, we have the Investigation of the Three Times (Traikālyaparīkṣā), which borrows Bhartṛhari’s idea of the one brahman and its proliferation and which discusses how the absolute, i.e., cognition (vijñāna), appears as diverse phenomena. This text explains a shift from the first to the second kind of 287

Kei Kataoka

cognition insofar as it moves from the perception of sages to that of ordinary persons. As for the reverse transition, from Level 2 to Level 1, we have the Investigation of the Cognitive Object (Ālambanaparīkṣā) with its autocommentary, which denies external reality (atoms and their aggregate) in order to prove consciousness-only. We can regard the function of this text as expounding a progression from the second level of cognition to the first insofar as it moves from the perception of ordinary persons to that of sages. The remaining texts all correspond to the imagined nature or the existent merely as designation, along with the various forms of conceptual cognition. Looking at Level 3a, which deals primarily with inference and its object, there are the Two-Headed Drum of the Wheel of Reasons (Hetucakraḍamaru) and the Gateway to Logic, as well as the Commentary on the Precepts of Argumentation (Vādavidhānaṭīkā), and the Gateway to Logical Reason (Hetumukha). These texts all treat inference (including argumentation) primarily from the perspective of the valid means of cognition, whereas the Investigation of General Characteristics (Sāmānyalakṣaṇaparīkṣā) and the Investigation of Universals (Sāmānyaparīkṣā) discuss general characteristics as word-referents. Turning to Level 3b, corresponding to cognition of empirical reality or perceptual judgment is the Treatise on Dependent Designation (Upādāyaprajñaptiprakaraṇa), which handles what is ordinarily taken to be empirically existent – e.g., the object of a conceptual cognition “pot” – in order to justify the Buddha’s reference to such things in his teachings. Level 3c, corresponding to erroneous cognition, is represented by the Hair in the Hand Treatise (Hastavālaprakaraṇa), which discusses the famous case of the rope in a dark room erroneously taken to be a snake in order to show the nature of conceptualization in general. The existence of the snake – i.e., the object that erroneously appears to cognition – is denied and instead the existence of the rope is affirmed; this corresponds to the transition from Level 3c to Level 3b. The existence of the rope, in turn, is denied by reducing it into parts (aggregated atoms) that really exist in the second level, corresponding therefore to the transition from Level 3b to Level 2. Ultimately, however, as the Investigation of the Cognitive Object also proves, substantially existent entities (such as atoms) can also be discarded as false, as is seen in the transition from Level 2 to Level 1. In this way, the text covers the issues related to all three stages. The Compendium mainly deals with the issues relevant to Level 2 and Level 3a, i.e., perception and inference, in addition to the articulation of criticism of other schools as carried out previously in his now lost works such as the Investigation of the Nyāya (Nyāyaparīkṣā), the Investigation of the Sāṁkhya (Sāṁkhyaparīkṣā), the Investigation of the Vaiśeṣika (Vaiśeṣikaparīkṣā), and the Investigation of Universals, judging from their known titles.

Self-Awareness The notion of self-awareness (svasaṃvedana/svasaṃvitti) refers to both the fact that and the process by which cognition feels, senses, or is aware of itself by means of itself. It is one of the characteristic features of Dignāga’s system. Modern scholars distinguish between two subcategories of self-awareness: reflexive and intentional (Williams 1998; Kellner 2010), although Dignāga himself does not explicitly distinguish between the two. The idea of reflexive self-awareness, in which a cognition simultaneously grasps itself as well as its targeted object, can be traced, before Dignāga, to the Buddhist school of the Mahāsāṁghikas (Katsura 1976a; Yao 2005, 15). These earlier Buddhists argue that just as a lamp simultaneously illuminates objects and itself, the Buddha’s cognition – in as much as he is omniscient – can grasp everything, including cognition itself (T1545, 42c12–14). This model of self-awareness is described in PS I 11cd (Kellner 2010). It is noteworthy that Dignāga himself does not appeal to the example of the lamp when he explains the self-awareness 288

Dignāga

of ordinary cognition.14 However, his Mīmāṁsaka opponents, such as the “Commentator” Vṛttikāra, who is quoted by Śabara and identifiable as Upavarṣa, and Kumārila, mention the example of the lamp and fire, respectively, when they criticize the idea of self-awareness.15 Intentional self-awareness, in contrast to reflexive self-awareness, is the process by which, within the same cognition, one aspect (ākāra) of cognition, called the grasper (grāhaka), apprehends another aspect, called the grasped (grāhya). This intentional structure is what makes it possible to say that it is the same cognition that knows itself. This model fits Dignāga’s explanation in PSV I 9a (Williams 1998). A precursor of this idea – namely, the notion of the two internal aspects of cognition – is found in Yogācāra works before Dignāga. References to the grasped and the grasper are very common, particularly when they refer to the phenomenal world consisting of objects and subjects.16 What is the strongest motivation for Dignāga to claim that cognition feels or senses itself by means of itself? A clue is given in PSV I 8cd, where Dignāga denies that cognition involves any operation (vyāpāra), i.e., any dynamic process involving an action (kriyā) and factors of action (kāraka). This idea of negating operation was explicitly advanced by Vasubandhu (AKBh 31,12–15) when he proclaimed his Sautrāntika view that the cognitive process is not an action carried out by an agent, but only a causal process in which “an optical cognition arises depending on the eye and colors/forms.” In other words, Vasubandhu claims that a cognitive process should be regarded as an instance of dependent origination ( pratītyasamutpāda) among causal factors and their effect and not as a kind of action accomplished by its factors of action (such as a separate agent, object, instrument, and so on). According to the Vaibhāṣikas, mundane expressions such as “the eye sees” can literally be interpreted as capturing the fact that the eye, i.e., the faculty of sight, is the agent of cognizing colors/forms. According to the Sautrāntikas, however, this is not a literal expression but only a figurative one (upacāra). Dignāga accepts this and develops Vasubandhu’s Sautrāntika view even further. In Dignāga’s theory of self-awareness, according to which cognition feels or senses itself by means of itself, all elements of cognition are nothing other than cognition and are not really distinct from each other. His arguments in support of this view can best be understood as an attempt to convince his Brahmanical opponents – such as the Naiyāyikas – of this fact. Naiyāyikas, for example, interpret a cognition as an action together with its factors of action, such as the object and the instrument of knowing. Yet for Dignāga, ultimately speaking, what happens in a cognitive process is not an action but merely the arising of an instance of cognition endowed with two aspects, that of the grasper and that of the grasped. The action-model of the Vaibhāṣikas, e.g., “The eye sees colors/forms,” should be revised, as Vasubandhu states, to the dependent arising model of the Sautrāntikas, e.g., “An optical cognition arises depending on the eye and colors/forms,” which in turn can be interpreted according to the Yogācāra model of consciousness-only. Now, when speaking of visual cognition, all that is being asserted is that from a preceding cognition1, a subsequent instance of cognition2 arises. Ultimately, in the Yogācāra view, there are only successive series of instances of cognition; no factors external to cognition are necessary to explain it. What people call “apprehending” or “grasping” something is not an operative process with a subject and an object, but the mere origination of an instance of cognition. Adopting the position common to the latter two models (which is called sākāra-vāda or the theory that a cognition is necessarily endowed with the form/image of an object), Dignāga criticizes the first model in which the object of cognition is outside of cognition (and which is therefore classified as nirākāra-vāda, or the theory that a cognition is not endowed with the form/image of an object). Thus, in contrast to the typical Brahmanical framing of the process producing a valid cognition as involving a means of valid cognition ( pramāṇa), its object ( prameya), and valid 289

Kei Kataoka

cognition ( pramiti) as a result ( phala),17 Dignāga offers two different analyses of the cognitive process based on two different models, the Yogācāra and the Sautrāntika. From the standpoint of the Yogācāra, when the internal grasped-aspect of cognition is the object,18 the graspedaspect of cognition is self-cognized by its grasper-aspect, and the result is self-awareness (PS I 9ab, 10). From the standpoint of the Sautrāntika view, however, i.e., when an external object is the object, cognition cognizes it by “having its appearance” (viṣayābhāsatā). Sautrāntrikas understand that it is simply this fact of a cognition having the form or appearance of an external object that is the “means” by means of which valid cognition of an external object is accomplished as the result (PS I 9cd). On this second view, the fact of self-awareness, in the intentional sense that a cognition cognizes an aspect of itself and does not directly cognize an external object, is disregarded, Dignāga claims, because the form that enters cognition and the form of an external object are similar.19 One can legitimately say that an external object is cognized, although – even in the Sautrāntika view – in reality, it is the internal aspect of cognition and not an external object that is directly cognized. These two are ontologically different, though alleged to be similar, due to the fact that the image is understood to be the direct effect of the external object. Thus, while cognition actually knows an aspect of itself, it is stated to know an external object in the Sautrāntika view. Kumārila (c. 600–650 CE), one of the prominent Mīmāṁsā philosophers, revealed Dignāga’s trick and severely criticized him by pointing out that in Dignāga’s Sautrāntika view, the means ( pramāṇa) and the result ( phala) would have different objects (bhinnārthatva).20 This is because the former deals with an external object, but the latter, in fact, deals only with the internal aspect of cognition, at least according to Dignāga. As we shall see, it is only with Dharmakīrti’s further revision of Dignāga’s theory that Kumārila’s criticism was avoided (Kataoka 2016). As explained previously, it seems that Dignāga’s introduction of the notion of “selfawareness” is motivated by his desire to present his own Buddhist views using the Brahmanical terminology of the three factors of action (means, object, and result), which ultimately go back to the grammatical notions of action and its factors. But as Vasubandhu stated, the framing of cognition as an action is not apt for the Buddhist. Mundane expressions of cognitive operation such as “sees” and “cognizes” are only figurative. The cognitive process should be explained not as an action but rather in terms of the Buddhist framing of dependent origination. Dignāga wholeheartedly agrees with Vasubandhu in this regard and he himself makes it explicit that while cognition seemingly has an operation, in fact all the elements of a cognition are devoid of operation (nirvyāpāra). Nonetheless, in order to clarify his views in contrast to those of his opponents, Dignāga provisionally stands on the same ground as his opponents, most typically that of the Naiyāyikas. For this reason, Dignāga engages the language of means, object, and result – even while holding that these are ultimately not separate entities. Kumārila is harsh in his condemnation of this trick, and he also severely criticized Dignāga’s view of self-awareness21 by pointing out effectively that transitive and causative verbs, such as “to grasp,” “to illuminate,” “to cause to shine forth,” and so on, which require a grammatical object, necessarily imply a distinct subject and object (ŚV Śūnya 65–67ab). Insofar as transitive and causative verbs such as “to feel” (vedayate) clearly require an object, one cannot but admit the operation or action toward something else when using them. Likewise, cognition cannot know or “grasp” itself but must cognize or “grasp” something separate from itself. This is where Dharmakīrti’s revision – changing the model of self-awareness from transitive to intransitive, from dual to nondual – proved to be critical. For Dharmakīrti, it is not the case that cognition causes something else to shine forth ( prakāśayati) (PV III 329ab, PSṬ I 70,12–16); rather, it spontaneously shines forth (PV III 478d: svayam eva prakāśate). The self-luminosity or self-revelation of a cognition endowed with content is the actual nature 290

Dignāga

of the cognitive process (PV III 478, 480). The notion of self-luminosity (svayaṁprakāśa/ svaprakāśa) clarifies his view regarding cognition endowed with a form (sākārajñāna) without referring to the action-model. Dharmakīrti gives up figurative expressions such as “grasps (itself)” and “cognizes (itself)” to describe his own views and paves the way to a distinctively Buddhist epistemology oriented to consciousness-only by introducing the notion of self-luminosity. Cognition endowed with a form spontaneously shines forth without requiring a distinct object to be illuminated. Dignāga’s earlier attempts to articulate a Buddhist theory of selfawareness in which the three factors of action are not different were clearly critical in allowing Dharmakīrti to take the next step to a vision of self-awareness as pure luminosity.

Inference As explained previously, Dignāga’s system of inference, typically exemplified by the inference of fire from smoke,22 includes both argumentation (e.g., where one proves the property of being transient from the property of being produced in the case of sounds and other products23) and verbal testimony (e.g., where a hearer understands a cow from the word “cow”). As previously noted, Dignāga applies the same method of analysis to the three domains of epistemology, dialectics, and semantics, which in earlier Buddhist traditions had been conceived as distinct from each other. Dignāga is not completely free from the tradition in which the analogical – and therefore empirical – character of inference is retained, and this emphasis on the analogical and empirical can be seen in his continued emphasis on the necessity for similar and dissimilar examples in the presentation of a valid inference. At stake in these conversations is how to supply a warrant for the validity of inference. How, in other words, can we be certain that a particular reason property (e.g., being produced) is a reliable indication of a property we are trying to prove (e.g., being transient)? The emphasis on similar and dissimilar examples allows us to recognize Dignāga’s basic empirical orientation, which will implicitly become the target of criticism by Dharmakīrti, who favors a more analytical approach to inferential reasoning. The empirical attitude of Dignāga is, in fact, shared by other scholars prior to Dharmakīrti. Dharmakīrti also severely criticizes the empirical attitude of Kumārila, who appeals to repeated observations (darśana) for establishing the relation between reason and consequent. At the same time, Dharmakīrti equally criticizes the empiricism of the lesser-known Buddhist thinker Īśvarasena, who appeals to mere non-observation (adarśana) as a warrant for the absence of a given entity (Steinkellner 1997). From the perspective of Dharmakīrti, however, one cannot analogically infer x on the basis of thus-far repeated experiences of x; nor can one negate x just because one has never perceived x so far (Tillemans 2004). The development of further criteria beyond the analogical and empirical for establishing a warrant for valid inference is one of the hallmarks of Dharmakīrti’s contribution to Buddhist argumentation. Īśvarasena, who is placed after Dignāga and before Dharmakīrti, accepts nonperception (anupalabdhi/anupalambha) as an independent means of valid cognition, whereas Dignāga posits only two means of valid cognition, i.e., perception and inference. We should probably regard Īśvarasena’s revision to Dignāga’s theory, therefore, as an attempt to make manifest the implicit presupposition of his predecessors.24 As we will see, nonperception does play a critical role in Dignāga’s system of inference, in particular regarding the ascertainment of invariable concomitance (vyāpti). That is, from the perspective of Dignāga, one can conclude the absence of x if x has never been perceived.25 A deviating counterexample, such as smoke without fire, certainly does not exist because such a case has never been observed. This non-observation of deviation is the key to understand Dignāga’s system of inference (Katsura 1992). In developing his theory of nonperception as a separate means of valid cognition, Īśvarasena was probably trying to make further 291

Kei Kataoka

sense of Dignāga’s own theory of inference. Dharmakīrti’s further refinements would not have been possible without the pioneering contributions of these two predecessors.

The Three Conditions of an Inferential Reason Dignāga’s contribution to logic lies first in his sharpening of the three conditions of an inferential reason already well established in his time.26 These three conditions are: 1. The reason property (e.g., smoke) must belong to the subject (e.g., a mountain). This is called the pakṣadharmatā, or possession of the reason property. 2. The reason property occurs only in similar examples (e.g., a kitchen). This is called the anvaya, or coexistence. 3. The reason property does not occur in dissimilar examples (e.g., a lake). This is called the vyatireka, or co-absence. Taking these three conditions into account, Dignāga lists nine types of reasons and makes a complete chart in his earliest logical work, the Two-Headed Drum of the Wheel of Reasons (Katsura 1984, 140; 1998, 262). His adding the word “only” to the second condition is the most crucial and original point in his analysis of a genuine reason. Inference of fire from smoke requires the prior establishment of invariable concomitance that wherever there is smoke, there is fire. This invariable concomitance is certified by conditions 2 and 3 in the preceding list, the so-called coexistence and co-absence. Dignāga is the first scholar who thematized invariable concomitance through the concept of pervasion (vyāpti) so as to highlight that a genuine reason must showcase that the domain of the reason property is pervaded by the domain of the target property or the property to be proved. In other words, the domain of smoke must be pervaded by that of fire if smoke is to be a legitimate reason for inferring fire. Dignāga’s introduction of the concept of pervasion implies that he does not allow a mere analogical inference called śeṣavat (“just like the rest” or “possessing a remainder”), i.e., a probabilistic inference based on induction. The Essentials of Method (Fang-pien-hsin-lun) illustrates the śeṣavat type of inference with the example of inferring the saltiness of all sea water from tasting only a bit of it (T1632, 25b13–14). Another example, inferring that all the rice in a pot is cooked from sampling just a single grain, is mentioned in Piṅgala’s commentary (T1564, 24b10) on Nāgārjuna’s Middle Way and criticized later by Dharmakīrti as śeṣavad anumānam (in PV I 13–14; see Katsura 1986a, 41–42). As we will see in Dignāga’s semantic theory, Dignāga prefers co-absence (vyatireka), e.g., the fact that there is no smoke without fire, to coexistence (anvaya), e.g., the fact that smoke occurs with fire, as the primary method for establishing invariable concomitance. This is because the latter, coexistence, cannot cover all infinite cases, whereas he believes the former, co-absence, can. As mentioned previously, Dignāga presupposes, like other scholars in his time, that a counterexample, such as the place where we find smoke without fire, does not exist if it has never been perceived. He refers to the invariable concomitance of the reason property with the property to be established, e.g., of smoke with fire, by means of a double negative expression avinābhāva (literally “nonexistence without”). The idea is that without fire, there can be no smoke, thus solidifying the invariable nature of this concomitance. Dignāga inherits this preference from Vasubandhu’s Method for Argumentation (Vādavidhi) (Katsura 1986a, 52–60), and we see here a structural similarity to the semantic theory of exclusion of what is other (anyāpoha), to be discussed in what follows. In this connection, it is necessary to distinguish between contingent coexistence and invariable coexistence. Dignāga (PSV IV 4) refers to the former as “a reason merely existing in 292

Dignāga

(a limited number of) similar examples” (sapakṣe sadbhāvamātram) and the latter as “a reason existing only in similar examples (and not in dissimilar examples)” (sapakṣe eva sadbhāvaḥ).27 The former does not imply that a reason is excluded from dissimilar examples, whereas the latter does do so. The particle “only” (eva) highlights a concomitance that it is invariable and necessary (avaśyam), such as the case that wherever there is smoke, there is fire.28 He also etymologically interprets (in PSV IV 2) a probans’s “following” (anugama) a probandum as its “going everywhere” (sarvatra gamaḥ). In other words, smoke’s co-occurrence with fire is universal. In contrast to the case of coexistence, Dignāga does not distinguish between the two types of co-absence. He considers only invariable co-absence as the vyatireka and not contingent co-absence. For Dignāga, the idea that “without fire, there is no smoke” indicates a universal relationship and not a single case of co-absence by chance. It is necessarily the case that in the absence of a probandum there is no probans (sādhyābhāve’sattvam eva). In order to establish an invariable concomitance, Dignāga adopts the negative method and rejects the positive method based on limited experiences of coexistence, because the former method is universal. As mentioned earlier, Dignāga makes a chart in his Two-Headed Drum of the Wheel of Reasons to illustrate the existence or absence of a reason in similar and dissimilar examples either everywhere or partially (sarvatra ekadeśe vā). Altogether he lists nine (3 × 3) possible types of reasons (Katsura 1984, 140; 1998, 262), of which only the second and the eighth are valid. For example, as the ninth type, he refers to amūrtatva, or formlessness, as the reason (wrongly) applied to prove the eternality of sounds. This reason, being formless, exists in a portion of similar examples, such as ether, and it also exists in a portion of dissimilar examples, such as motion. It is to be noted that in this chart, the coexistence that Dignāga intends is contingent and not invariable. He does not intend existence of formlessness only in similar examples because he clearly accepts it as existing in dissimilar examples such as ether, too. In other words, the list of nine possible reasons presents an asymmetry between coexistence that is contingent and co-absence that is invariable. This asymmetry is the main reason that prompts him to give preference to (invariable) co-absence and to disregard (contingent) coexistence. But it also entails an exegetical problem that either of the two conditions of a proper reason, i.e., its existence in similar examples or its absence in dissimilar examples, well established in the Buddhist tradition, would be useless. As mentioned previously, he adds “only” to the second condition in his last work (PSV IV 4) when expressing invariable concomitance, e.g., of smoke with fire: smoke exists only in similar examples. For him, this expresses invariable co-absence, i.e., smoke never exists without fire. But then, the third condition, invariable absence in dissimilar examples (vipakṣe’sattvam eva), e.g., that smoke never exists wherever fire does not exist, would be useless, because it would be already understood by implication (arthāpatti). That smoke exists only where fire exists logically implies that smoke never exists without fire. We can depict Dignāga as caught between the traditional account of the three conditions of the reason and his own original contribution of adding “only” to the formulation of the second. Table 17.1  Dignāga’s nine possible types of reasons from the Two-Headed Drum of the Wheel of Reasons. In Dissimilar Examples In Similar Examples

existent everywhere

absent

partially existent

existent everywhere

I

II

III

absent

IV

V

VI

partially existent

VII

VIII

IX

293

Kei Kataoka

His earlier Two-Headed Drum of the Wheel of Reasons still retains the old notion of contingent coexistence, whereas the later Compendium already shows a new idea of invariable coexistence, but Dignāga does not totally reject the traditional account of the three conditions. He tries to find some use for either of the last two conditions of a proper reason, i.e., its existence in similar examples and its absence in dissimilar examples, by considering five different cases in PSV IV 4 as described in what follows (Kitagawa 1965, 251ff.; Katsura 2004, 159ff.). He first divides the cases into two: (1) when the third condition (i.e., absence in dissimilar examples) is demonstrated; and (2) when the second condition (i.e., existence in similar examples) is demonstrated:29 1. When the third condition is demonstrated, 1a. A similar example too should be stated to exclude a pseudo-reason. 1b. Stating a similar example is not required because invariable coexistence is implied. 2. When the second condition is demonstrated, 2a. Stating a dissimilar example is not required because invariable co-absence is implied. 2b1. The third condition should be additionally stated to confirm invariable concomitance. 2b2. The third condition is interpreted differently. Let us examine these cases one by one. 1a. In the first case, i.e., when the third condition is demonstrated with a dissimilar example to show the invariable concomitance, a similar example should also be stated in order to exclude a pseudo-reason called “unique-inconclusive” (asādhāraṇa-anaikāntika). This pseudo-reason does not prove a probandum, because it is unique and found only in the subject. In the case of the fifth type of the nine reasons listed previously, audibility (śrāvaṇatva) is (wrongly) applied to prove the eternality of sounds. The inference is formulated in this way: “sounds must be eternal, because they are audible, just like x.” But audibility is found only in sounds, the very target of the proof, and not in similar examples. At the same time, the reason fulfills the third condition, because anything transient, except the disputed sounds, is inaudible. (Only sounds are regarded as audible in this argument.) This proof, therefore, fulfills the first and the third conditions, but not the second. Although invariable concomitance is stated in reference to a dissimilar example, it is also necessary to mention at least one similar example in order to exclude this type of exceptional pseudo-reason. In this way, Dignāga shows some use of the second condition (contingent coexistence) with consideration for the tradition. 1b. He simply admits that it is not necessary to state a similar example, because the third condition that confirms invariable co-absence implies, via implication, the second one, i.e., invariable coexistence. In other words, 2a is the same as 1b. When a similar example is stated to show invariable coexistence, invariable co-absence is implied  – and therefore, stating a dissimilar example is not necessary. This also means that Dignāga accepts contraposition. Either of the two conditions would be superfluous. In both cases, Dignāga presupposes invariable coexistence. 2b1. When the second condition is demonstrated with a similar example, invariable concomitance cannot in fact be shown, because coexistence – if following the traditional idea – is contingent. Therefore, the third condition should be additionally stated to confirm invariable concomitance. 2b2. Following the new idea, however, invariable concomitance can be confirmed by the second condition. In other words, invariable coexistence is stated with a similar example. Then, the third condition, if normally interpreted, would be useless. Therefore, Dignāga changes the 294

Dignāga

position of “only” and tries to apologetically show the necessity of the third condition.30 His forceful explanation, which objectively speaking is nonsense, all the more indicates that he tries to rescue the third condition from being useless. In his PSV, Dignāga is well aware of contraposition between the second and the third conditions that entails redundancy of one of them. But he also gives consideration to the traditional idea that both are needed, and he tries to some extent to show some use when one of them seemingly becomes redundant. Again, it is only with Dharmakīrti that a paradigm change occurs with the introduction of the completely new idea of natural binding (svabhāvapratibandha) as the basis of a phenomenal invariable concomitance. This allows Dharmakīrti to claim that only one of the two conditions is sufficient to show the invariable concomitance if the reason in question fulfills the objective criteria. Without Dignāga’s innovations, however, it is unlikely that Dharmakīrti would have been able to conceptualize his breakthrough ideas.

Antinomy The science of dialectic had a long history before Dignāga, as gauged from Buddhist and nonBuddhist logical works. In actual debates, there occurs a situation where neither of the two sides – the proponent and the opponent – wins over the other. This is a case of antinomy, in which a proponent’s proof encounters an equally strong counterproof. It results in a dialectical impasse. In PS III 23, Dignāga claims that a proper reason should not consist of two mutually contradicting reasons. In other words, a proper argument should not allow an equally strong counterargument. Dignāga designates this type of inconclusive reason as “not deviating from a contradictory reason” (viruddhāvyabhicārin), i.e., a pseudo-reason that has an equally strong counter-reason. Effectively, he recognizes the possibility of antinomy (Kitagawa 1965, 193ff.; Ono 2010). Dignāga’s successor Īśvarasena clarifies Dignāga’s intention by adding three more conditions to the three previously mentioned. Among them, the fifth is the condition that a proper reason should not have a counterproof of equal argumentative weight. Debates showcase situations when a reason, even when invalid objectively, can remain “valid” until a counterreason actually is presented by an opponent. Of course, this sort of subjective validity is inadmissible for philosophers like Dharmakīrti, who insist on the discovery of truth, and not mere victory, as the only legitimate goal of debate. In his Drop of Reasons (HB 37,15–38, 3), Dharmakīrti implicitly criticizes Dignāga’s empirical attitude when he explicitly criticizes Īśvarasena’s notion of the fifth condition. For Dharmakīrti, if the proponent’s reason passes the required criteria, it is always valid and never changes its value. If it is invalid because of not fulfilling the required criteria, then it is invalid from the beginning and does not have to wait for an opponent’s counterproof to render it so. In this matter too, then, we can confirm Dignāga’s empirical attitude that a counterproof does not exist if it has not been witnessed so far. And we can see as well how Dignāga’s ideas become the stepping stones for Dharmakīrti’s further refinements.

The Exclusion of What Is Other Non-observation plays a key role also in Dignāga’s semantics, i.e., in his theory of the exclusion of what is other (anyāpoha), which he claims is parallel to inference (PS V 1). For Dignāga, a word denotes its own referent just as an inferential reason property (probans) indicates a target property (probandum). Dignāga’s claim implies that one establishes the connection of the word “cow” with its referent, just as one establishes the invariable concomitance of 295

Kei Kataoka

a probans, such as being produced, with a probandum, such as transiency. This connection or invariable concomitance is based on non-observation. One has never seen something produced without transiency. Similarly, one has never seen the word “cow” applied to non-cows. Coabsence, which he glosses as non-occurrence in the absence of a similar case (atulye’vṛttiḥ) or the absence of the reason property in the absence of the target property,31 can be shown only through non-observation (adarśanamātreṇa), as Dignāga claims in PSV V 34. Dignāga refutes four semantic theories of realist opponents and establishes his own by demonstrating that his theory fulfills five criteria necessary for any semantic theory to function. The five criteria are as follows: 1. One can connect a word with its referent. In other words, learning the word-referent relationship is possible. 2. There is non-deviation of a word from its own referent. A word is always learned with respect to its referent, and thus does not arouse any doubt when it is applied. It denotes its own referent with certainty. 3. The theory must be able to explain the linguistic phenomenon of coreferentiality (sāmānādhikaraṇya),32 such as the fact that the same thing can receive the qualifications “an existing substance” (sad dravyam) and “an existing pot” (san ghaṭaḥ). 4. Lower concepts (i.e., subcategories) must be implied (ākṣepa), included (vyāpti), and not excluded (anapohana). For example, the word “existing,” which denotes existing entities, includes and implies its lower concepts such as substances and pots. 5. A word should operate toward its referent directly (sākṣādvṛtti) and not via something else. Indirectness or dependency should be avoided so as to secure the characteristic feature of (primary) denotation that is distinguished from secondary or metaphorical usage. Dignāga utilizes these five criteria to criticize the following four semantic theories of his realist opponents: (1) the theory of particulars (bheda); (2) the theory of universals (  jāti); (3) the theory of relationship (sambandha); and (4) the theory according to which a word denotes anything that has a real universal (tadvat). We shall look at these each in turn. The first semantic theory, which claims that words denote particulars, does not meet the first and the second of the stated five criteria. It is impossible to actually connect the word “cow” with every individual cow due to their infinite number. The word “cow,” if connected with individual cows, would cause a doubt when applied, because a hearer would not be sure which particular cow the word denotes.33 The second and the third theories, which claim that words denote universals and their relationships, respectively, are both refuted by way of the third criterion, because neither theory allows coreferentiality. For instance, according to the second theory, the word “existing” denotes a property, namely “existence,” whereas the word “substance” denotes not a property but rather the locus of a property.34 “existing”

→

existence |

“substance”

→

substance

The two denotations cannot occur at the same level in a single entity co-referred to by the two words as, for example, an “existing substance” (sad dravyam, both in the nominative). Rather, the two denotations would have to be coordinated in an expression such as the “existing of a substance” (sad dravyasya, where the latter is in the genitive). 296

Dignāga

Having rejected the first three theories, Dignāga undertakes in PS V 4 to deny the fourth theory, which claims that a word denotes anything that has a real universal (tadvat/jātimat) and which also fulfills the first three criteria. The holder of this last view is no doubt the strongest opponent for Dignāga because the structure of this theory is parallel to Dignāga’s own theory of apohavat (literally “having exclusion”),35 i.e., the view that a word denotes anything that has the exclusion of what is other. For example, the word “cow” expresses any cow that is excluded from non-cows. In this way, Dignāga gets rid of the real universal cowness in his semantics. Dignāga refutes the fourth view first by appealing to the fifth criterion, showing that in this view, a word would be dependent on a real universal when denoting its own referents. In this view, for example, the universal “existence” would intervene between the word “existing” and actual existing entities. Therefore, the word would not denote its referents directly but mediately or secondarily through the universal (PS V 4b: upacārāt). Furthermore, due to this mediation, the word “existing” cannot imply the lower concepts (subcategories), e.g., substances and pots. Thus, the fourth view does not fulfill the fourth condition, either. By contrast, Dignāga’s own view does not exhibit these issues. The word “existing” directly denotes existing entities and implies substances and pots. Exclusion, which functions as a common property, is unreal and, ontologically speaking, an absence. Thus, because exclusion is ontologically nothing, coreferential expressions can be explained in Dignāga’s exclusion theory, whereas they cannot be explained in the fourth theory due to the reality of the universal that causes intervention. Another important reason why the exclusion theory is superior to the fourth theory is its different conception of the relation (sambandha) between the word and its referent and its resulting different conception of denotation (abhidhāna). Whereas Dignāga presupposes, so to speak, a weak relation and a weak denotation based on the negative method of co-absence (e.g., the word “cow” is not applied to non-cows), his realist opponents presuppose a strong relation and a strong denotation based on the positive method of coexistence. According to the positive theory, the word “existing” is connected only with its referent, and it is neither connected with higher general concepts such as something knowable (  jñeya) nor with lower particular concepts such as substances (dravya), earthen things ( pārthiva, being composed of the earth element), trees (vṛkṣa), and so on, in the ontological hierarchy.36 A word has a strong relation with its own definite referent, e.g., something that has the universal existence, and therefore it targets only a particular referent and does not go beyond that boundary. Consequently, this theory cannot explain co-referential expressions, because according to this theory, the word “existing” does not imply a pot. It denotes only existing entities as such. Furthermore, this theory cannot explain the linguistic phenomenon (i.e., the fourth criterion) that a word brings about a certainty with regard to higher general concepts and an uncertainty with regard to lower particular concepts. When one hears the word “tree,” for example, one is sure that a tree is earthen, a substance, existing, and so on; but the same person is uncertain whether the tree is a North Indian rosewood or a parrot tree. According to the realist account, however, there would be no adequate explanation for why we can know with certainty what a tree means in general but lack certainty about which particular individual tree actually is being denoted in any given utterance of the word “tree.” The realist theory of strong denotation cannot explain this phenomenon, because according to their theory, the denoting power of words reaches only as far as its own directly connected referent. According to the realist account, therefore, one would indeed have certainty about the two types of referent of the word “tree,” i.e., both the higher general and the lower particular concepts. Rather than being uncertain about whether the particular referent is a North Indian rosewood or a parrot tree, then, one would – due to the strong connections between words and particular referents – be able to ascertain that “The tree is surely a North Indian rosewood and a parrot tree,” just 297

Kei Kataoka

as one ascertains, “The tree is certainly earthen, a substance, existing, and so on.” But this is rather absurd. Dignāga, on the other hand, does not meet the same difficulty, because the relation and denotation that he presupposes are weak in that they are based on co-absence confirmed by non-observation. The word “tree” has never been seen applied to non-earthen things, and so on. Therefore, the word can cover the complement domain (remnants; e.g., earthen things and so on) left after the exclusion of non-earthen things and so on. The higher concepts of tree are not excluded and thus may be included as the target domain of the word. The negative method of confirming the semantic relationship leaves a wide semantic field possible as the referents of a word. Thus, it can explain why the word “tree” brings about a certainty with regard to the higher general concepts. It can also explain how it leaves in pace an uncertainty with regard to the lower particular concepts, because the word is indifferent to (upekṣate), i.e., neither totally excludes nor includes, its subtypes. For Dignāga, denotation (abhidhāna) is based on the non-exclusion (anapohana) of referents. This non-exclusion is possible only in the negative semantic theory, because according to this theory, the referent of a word is what remains after the exclusion of other referents. In other words, it is a remainder. Non-exclusion is the basis for the inclusion of a certain referent. Due to inclusion, a certain referent is denoted, or, more correctly speaking, it is made manifest by a word. Dignāga himself uses the term “make manifest” (dyotayati) instead of “denote” (abhidhatte) in PSV V 1. This word choice indicates Dignāga’s awareness of the difference between the realist type of strong denotation and his own theory of denotation that can also be called “making manifest.” He probably implies that a word makes manifest and specifies a potentially existing meaning among many that an object has, just as a prefix (upasarga) such as pra- makes manifest a particular meaning of a verbal base that has many potential meanings. It is not the case that a word conveys all aspects of an object.37 The word “existing” highlights a certain aspect by excluding other referents, i.e., nonexisting entities. A word highlights only a part of an object.

A Concluding Remark In the three domains  – epistemology, dialectics, and semantics  – dealt with in the Compendium, nonperception plays a central role. Dignāga is well aware of the defects of limited and contingent experiences of coexistence and instead emphasizes the function of invariable coabsence that can cover an infinite number of cases. Later, Dharmakīrti revises Dignāga’s view by introducing a new criterion of an inferential reason and adding a qualification “of a perceptible entity” to nonperception, criticizing Dignāga’s empirical attitude by way of criticizing Īśvarasena. Dignāga’s innovations and developments, difficulties and considerations, and inconsistencies and limitations, are all thrown into relief when we compare him with his predecessors, such as Vasubandhu and Bhartṛhari; his opponents, such as Kumārila; and his faithful or critical successors, such as Īśvarasena and Dharmakīrti. Ongoing studies of the six chapters of his Compendium will further elucidate his exact place in Indian intellectual history. Due to the limitation of primary material, the present chapter has to remain a preliminary overview of his contribution. What remains abundantly clear, however, is the critical role that Dignāga played in setting the stage for the further innovations of Dharmakīrti, his most famous intellectual successor.

Notes 1 I thank Hugo David, William Edelglass, Pierre-Julien Harter, Sara McClintock, Shinya Moriyama, John Taber, Toshikazu Watanabe, and Alex Watson for their comments.

298

Dignāga 2 For chronology, see Frauwallner (1959); Hattori (1961a; 1968). 3 For Dignāga’s works, Frauwallner (1959) is the first reference. Hattori (1968) contains an exhaustive list of his works. Up-to-date bibliographical details are given in the website of the EAST (Epistemology and Argumentation in South Asia and Tibet) project (http://east.uni-hd.de/). Tsukamoto, Matsunaga, and Isoda (1990) is also helpful for bibliographical information. For Dignāga’s philosophy, Katsura’s articles are most helpful. 4 Of these, the first and the second chapters of the PSṬ are published as PSṬ I and PSṬ II respectively. PSṬ V is being prepared by Horst Lasic and Patrick McAllister. A Japanese group headed by Shōryū Katsura (and including Motoi Ono, Yasutaka Muroya, and Toshikazu Watanabe) is preparing the third, fourth, and sixth chapters of PS(V) and PSṬ. Their recent articles and interim reports based on the ongoing projects offer new insights. 5 Dignāga’s system of the Gateway to Logic thrives in east Asia via Xuanzang’s Chinese translation of and Kuiji’s commentary on Śaṅkarasvāmin’s Entry into Logic (Nyāyapraveśa) (Radich and Lin 2014). 6 Dignāga is often critical of Vasubandhu’s Vādavidhi and prefers the system of the Vādavidhāna, another work attributed to Vasubandhu. 7 As a result of this reorganization, Jinendrabuddhi later had to make huge efforts to reinterpret Dignāga’s inconvenient passages in accordance with Dharmakīrti’s new system. See Pind (2015, I xxxii); Kataoka (2016, 235). 8 In the case of an erroneous cognition of a rope as a snake, a snake is the imagined nature, whereas the aggregated atoms of the rope is the dependent nature; and the absence of a snake in the aggregated atoms of the rope is the perfected nature, i.e., the ultimate truth. 9 We also see an implicit reference to the three natures in some of Dignāga’s other works, including his Hair in the Hand Treatise (Hastavālaprakaraṇa) and Treatise on Dependent Designation (Upādāyaprajñaptiprakaraṇa) (Hattori 1961b), 10 See § 3.4.1 of the synopsis by Katsura (1976b). Ratnākaraśānti’s PPU offers us a systematic overview of various traditions, both Buddhist and non-Buddhist, from the Yogācāra perspective as the highest point. 11 Dignāga’s description of yogins’ pure intuitive perception as “not mixed with the teacher’s teaching” and “not mixed with scriptural conceptualization” (PS[V] I 6cd) implies that it is primarily concerned with religious matters. 12 For the exact nature of this mental perception, see Funayama (2000). 13 Here, my perspective is synchronic. For a diachronic analysis of his works, see Frauwallner (1959). 14 This implies that Dignāga does not understand the process of cognition with the metaphor of illumination. For Dignāga’s possible intention in introducing the notion of self-awareness, see Kataoka (2017). 15 ŚBh ad 1.1.3–5 (Frauwallner 1968, 28, 20–30, 1); ŚV Śūnya 65–67ab. 16 It is to be noted that for the Yogācāras the cognitive process (the action of grasping) is not the focus; rather, the emphasis lies on the point that the two categories, the grasped (grāhya) and the grasper (grāhaka), represent everything, i.e., the entire universe. Dignāga also seems to avoid mentioning the dynamic operation of grasping. He merely states that cognition is “self-felt by itself ” (  jñānasvasaṃvedya). 17 NBh 1,12–15. Vātsyāyana also mentions pramātṛ or the agent of valid cognition. In Buddhism, however, the agent, i.e., self (ātman), is naturally omitted. For the date of NBh, see Franco and Preisendanz (1995). 18 For interpretations of this passage, see Kellner (2010); Kataoka (2016). 19 PSV I 9cd (4,13–14), Kataoka (2016, 232). Thus, the cognition of an external object fulfills the two criteria of a genuine perception prescribed in Ālambanaparīkṣā 1–2a. 20 ŚV Pratyakṣa 79cd (cf. Taber 2005 for translation). See Kataoka (2016). 21 “A tip of finger cannot touch itself by means of itself. Similarly, cognition too cannot cognize itself by means of its part.” A quote from Kumārila’s lost work, the Bṛhaṭṭikā (Schmithausen 1965, 133, n. 40). Cf. LAS X 568. See also T1545, 43a26–28, T1552, 953c16–17, and T1564, 17a10. 22 An ontological perspective is particularly necessary to understand Indian theories of inference. The probans (hetu, sādhana) and the probandum (sādhya) – e.g., smoke and fire, respectively – are regarded as properties (dharma) of the property-bearer (dharmin), e.g., the mountain, which is also referred to as the subject ( pakṣa). Different terms originally coming from different disciplines and traditions are used to refer to the same concepts as a consequence of unification. 23 “A sound is transient, because it is produced, just like a pot.” This is a typical example of inference for others with regard to the two aspects of the same object which are not causally connected like

299

Kei Kataoka fire and smoke. The invariable concomitance presupposed here is that whatever is produced has been experienced as transient, as demonstrated in familiar examples like pots. 24 We find the same attitude in Vasubandhu. See Kellner and Taber (2014, 709): “We suggest that Vasubandhu develops in the Viṃśikā an extended argumentum ad ignorantiam where the absence of external objects is derived from the absence of evidence for their existence.” Kumārila’s pramāṇābhāva, i.e., the sixth means of valid cognition in his system, retains a similar character. Ogawa (2014) attempts to trace Dignāga’s adarśanamātra to the Pāṇinian notion of adarśanaṁ lopaḥ (A1.1.60). 25 Dharmakīrti’s view of the non-perception of perceptible entities (dṛśya-anupalabdhi), however, directly contradicts Dignāga’s position. For Dharmakīrti, one cannot assert the nonexistence of x just because x is not perceived, because there are cases when x, though existing, is not perceived due to its imperceptibility in terms of place, time, and nature. Non-perception of x works only when x is perceptible. Non-perception, if valid, must be qualified as “that of a perceptible entity.” Dharmakīrti rejects the old idea of mere (unqualified) non-perception (adarśanamātra) of x, which does not consider whether x is perceivable or not. 26 The three conditions are known to Asaṅga in his Conforming to the Middle Way (Shun-chung-lun, T1565, 42a24). The Ru-shi-lun (T1633, 30c20–21), which predates the Vādavidhi, is another important Buddhist work before Dignāga that also mentions them (Katsura 1986a). 27 The sandhi in this phrase has been omitted for clarity. 28 PSV II 10: yatra dhūmas tatra agniḥ. 29 I thank Shōryū Katsura for generously sharing his unpublished reconstruction of PSV IV. 30 He interprets the third condition as sādhyābhāve eva asattvam (= asati eva nāstitā) and not sādhyābhāve ’sattvam eva so that “only” functions to restrict the meaning of the negative particle to one of its three meanings, i.e., absence, by excluding other two meanings, i.e., other and contrary. For the details, see Lasic (2009). 31 PS IV 2 (Katsura 2004, 141): sādhyābhāve ca nāstitā. 32 For coreferentiality, see Taber and Kataoka (2017). 33 Remember that in this first theory, the word denotes (ungrouped) individual cows and not their aggregate. Claiming “all (individual) cows” in a group is equivalent to claiming either the second or the fourth theory. 34 Dignāga does not take into consideration substance-ness in this context, probably for the sake of simplicity. 35 The designation apohavat is found in Kumārila’s critique. Dignāga’s own expression is arthāntaranivṛttiviśiṣṭa, or “qualified by the negation of other referents.” 36 For the Vaiśeṣika hierarchy of universals that Dignāga presupposes, see Katsura (1979). 37 Pind (2015, II 3, n. 14).

Bibliography and Abbreviations Primary Sources AKBh A JIBS T NB

NBh PPU PV

Abhidharmakośabhāṣya. Pradhan, Prahlad, ed. 1967. Abhidharma Kośabhāṣya of Vasubandhu. Patna: Kashiprasad Jayaswal Research Institute. Aṣṭādhyāyī. Katre, Sumitra M., ed. 1989. Aṣṭādhyāyī of Pāṇini. Delhi: Motilal Banarsidass. Journal of Indian and Buddhist Studies. Taishō Shinshū Daizōkyō. The SAT Daizōkyō Text Database. http://21dzk.l.u-tokyo. ac.jp/SAT/. Nyāyabindu. Malvania, Dalsukhbhai, ed. 1971. Paṇḍita Durveka Miśra’s Dharmottarapradīpa, Being a Sub-commentary on Dharmottara’s Nyāyabinduṭīkā, a Commentary on Dharmakīrti’s Nyāyabindu. Patna: Kashiprasad Jayaswal Research Institute. Nyāyabhāṣya. Thakur, Anantalal, ed. 1997. Gautamīyanyāyadarśana with Bhāṣya of Vātsyāyana. New Delhi: Indian Council of Philosophical Research, 1997. Prajñāpāramitopadeśa. Pramāṇavārttika. PV III: Tosaki (1985). 300

Dignāga

PS(V) PS(V) I PS(V) V PSṬ PSṬ I PSṬ II

LAS ŚBh ŚV HB

Pramāṇasamuccaya(vṛtti). Steinkellner (2005). Pind (2015). Pramāṇasamuccayaṭīkā. Steinkellner, Ernst, Helmut Krasser, Horst Lasic, and Anne MacDonald, eds. 2005. Jinendrabuddhi’s Viśālāmalavatī Pramāṇasamuccayaṭīkā Chapter I. Part I: Critical Edition. Part II: Diplomatic Edition. Beijing, Vienna: China Tibetology Publishing. Lasic, Horst, Helmut Krasser, and Ernst Steinkellner, eds. 2013. Jinendrabuddhi´s Viśālāmalavatī Pramāṇasamuccayaṭīkā Chapter  2: Part 1: Critical Edition. Part 2: Diplomatic Edition. Beijing, Vienna: China Tibetology Publishing House and Austrian Academy of Sciences Press. Laṅkāvatārasūtra. P. L. Vaidya, ed. 1963. Saddharmalaṅkāvatārasūtram. Darbhanga: The Mithila Institute. Śābarabhāṣya. See Frauwallner 1968. Ślokavārttika. Dvārikadāsa Śāstrī, ed. 1978. Ślokavārttika of Śrī Kumārila Bhaṭṭa with the Commentary Nyāyaratnākara of Śrī Pārthasārathi Miśra. Varanasi: Tara Publications. Hetubindu. Steinkellner, Ernst, ed. 2016. Dharmakīrti’s Hetubindu. Beijing, Vienna: China Tibetology Publishing House and Austrian Academy of Sciences Pres.

Secondary Literature Franco, Eli, and Karin Preisendanz. 1995. “A Note on Bhavadāsa’s Interpretation of Mīmāṃsāsūtra 1.1.4 and the Date of the Nyāyabhāṣya.” Berliner Indologische Studien 8: 81–86. Frauwallner, Erich. 1959. “Dignaga, sein Werk und seine Entwicklung.” Wiener Zeitschrift für die Kunde Süd- und Ostasiens 3: 83–164. ———. 1968. Materialien zur ältesten Erkenntnislehre der Karmamīmāṃsā. Wien: Hermann Bölaus Nachf. Funayama, Toru. 2000. “Kamarashīra no Chokusetsu Chikaku ron niokeru I niyoru Chikaku (mānasa).” Tetsugaku Kenkyū 569: 105–32. Hattori, Masaaki. 1961a. “Digunāga oyobi Sono Shūhen no Nendai. Fu Sanji no Kōsatsu Wayaku.” Bukkyō Shigaku Ronshū, Tsukamoto Hakase Shōju Kinen, 79–96. Kyoto: Tsukamoto Hakase Shōju Kinenkai. ———. 1961b. “Digunāga ni okeru Kashō to Jitsuzai.” FAS 60: 16–28. ———. 1968. Dignāga, on Perception. Cambridge, MA: Harvard University Press. Kataoka, Kei. 2016. “Dignāga, Kumārila and Dharmakīrti on the Potential Problem of pramāṇa and phala Having Different Objects.” Journal of Indian Philosophy 44 (2): 229–39. ———. 2017. “Jikoninshiki no Seisei, Haikei, Henshitsu.” South Asian Classical Studies 12: 191–214. Katsura, Shōryū. 1976a. “Sarvālambanajñāna ni tsuite.” Journal of International Buddhist Studies 24 (2): 678(160)–679(161). ———. 1976b. “A Synopsis of the Prajñāpāramitopadeśa of Ratnākaraśānti.” Journal of International Buddhist Studies 25 (1): 487(38)–484(41). ———. 1979. “TheApoha Theory of Dignāga.” Journal of International Buddhist Studies 28 (1): 493(16)–489(20). ———. 1984. “Digunāga no Ninshikiron to Ronrigaku.” In: Kōza Daijō Bukkyō, edited by Akira Hirakawa, Yūichi Kajiyama, and Jikidō Takasaki, vol. 9, 104–52. Tokyo: Shunjūsha. ———. 1986a. “Indo Ronrigaku ni okeru Henjū Gainen no Seisei to Hatten.” Hiroshima Daigaku Bungakubu Kiyō 45: 1–122. ———. 1986b. “Dharmakīrti’s Theory of Truth.” Journal of Indian Philosophy 12: 215–35. ———. 1992. “Dignāga and Dharmakīrti on adarśanamātra and anupalabdhi.” Asiatische Studien 46/1: 222–31. ———. 1998. Indojin no Ronrigaku. Tokyo: Chūō Kōron sha. ———. 2004. “The Role of dṛṣṭānta in Dignāga’s Logic.” In The Role of the Example (Dṛṣṭānta) in Classical Indian Logic, edited by Shōryū Katsura and Ernst Steinkellner, 135–73. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien.

301

Kei Kataoka Kellner, Birgit. 2010. “Self-Awareness (svasaṃvedana) in Dignāga’s Pramāṇasamuccaya and -vṛtti.” Journal of Indian Philosophy 38: 203–31. Kellner, Birgit, and John Taber. 2014. “Studies in Yogācāra-Vijñānavāda Idealism I.” Asiatische Studien 68 (3): 709–56. Kitagawa, Hidenori. 1965. Indo Koten Ronrigaku no Kenkyū. Tokyo: Suzuki Gakujutsu Zaidan. Lasic, Horst. 2009. “Dignāga’s Description of a Logical Mark in Pramāṇasamuccaya 2.5cd.” International Journal of Buddhist Thought & Culture 13: 7–22. Ogawa, Hideyo. 2014. “Pānini Bunpōgaku Kotoba no Ryōikigai Fushiyō no Gensoku ni tsuite.” Indo Ronrigaku Kenkyū 7: 53–78. ———. 2017 “The Qualifier-Qualificand Relation and Coreferentiality.” In Reading Bhaṭṭa Jayanta on Buddhist Nominalism, edited by Patrick McAllister, 83–151. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Ono, Motoi. 2010. “Sōiketsujō (viruddhāvyabhicārin) o megutte.” Indo Ronrigaku Kenkyū 1: 125–43. Pind, Ole Holten. 2015. Dignāga’s Philosophy of Language: Pramāṇasamuccayavṛtti V on Anyāpoha. 2 parts. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Radich, Michael, and Chen-kuo Lin. 2014. “Introduction.” In A Distant Mirror, edited by Chen-kuo Lin and Michael Radich, 15–31. Hamburg: Hamburg University Press. Schmithausen, Lambert. 1965. Maṇḍanamiśra’s Vibhramavivekaḥ. Wien: Hermann Böhlaus Nachf. Steinkellner, Ernst. 1997. “Kumārila, Īśvarasena, and Dharmakīrti in Dialogue. A  New Interpretation of Pramāṇavārttika I 33.” In Bauddhavidyāsudhākaraḥ: Studies in Honour of Heinz Bechert on the Occasion of His 65th Birthday, edited by Petra Kieffer-Pülz and Jens-Uwe Hartmann, 625–46. Swisttal-Odendorf: Indica et Tibetica Verlag. ———. 2005. Dignāga’s Pramāṇasamuccaya, Chapter  1. A  Hypothetical Reconstruction of the Sanskrit Text with the Help of the Two Tibetan Translations on the Basis of the Hitherto Known Sanskrit Fragments and the Linguistic Materials Gained from Jinendrabuddhi’s Ṭīkā. https://www.oeaw.ac.at/ fileadmin/Institute/IKGA/PDF/forschung/buddhismuskunde/dignaga_PS_1.pdf Taber, John. 2005. A Hindu Critique of Buddhist Epistemology. London and New York: Routledge. Taber, John, and Kei Kataoka. 2017. “Coreference and Qualification: Dignāga Debated by Kumārila and Dharmakīrti.” In The Oxford Handbook of Indian Philosophy, edited by Jonardon Ganeri, 255–71. New York: Oxford University Press. Tillemans, Tom. 2004. “Inductiveness, Deductiveness and Examples in Buddhist Logic.” In The Role of the Example (dṛṣṭānta) in Classical Indian Logic, edited by Shoryu Katsura and Ernst Steinkellner, 251–75. Vienna: Wiener Studien zur Tibetologie und Buddhismuskunde. Tosaki, Hiromasa. 1985. Bukkyō Ninshikiron no Kenkyū (ge). Tokyo: Daitō Shuppansha. Tsukamoto, Keishō, Yūkei Matsunaga, and Hirofumi Isoda, eds. 1990. Bongo Butten no Kenkyū III Ronshohen. Kyoto: Heirakuji shoten. Williams, Paul. 1998. The Reflexive Nature of Awareness. A Tibetan Madhyamaka Defense. Richmond: Curzon Press. Yao, Zhihua. 2005. The Buddhist Theory of Self-Cognition. London and New York: Routledge.

302

18 DHARMAKĪRTI Philosopher and Defender of the Faith John Taber

Dharmakīrti is considered one of the co-founders of the so-called logico-epistemological school of Buddhist philosophy, together with his predecessor Dignāga. His contributions, however, are not limited to epistemology and logic. He was, in fact, one of the great systematizers of Buddhist doctrine at a time in Indian history when Buddhism was coming under increasing attack from competing traditions (Eltschinger 2014, 35–92). His works, especially his magnum opus, Commentary on the Means of Knowledge, represent collectively a systematic, rational defense of the main teachings of the Buddha such as they were understood by Dharmakīrti in his intellectual milieu, though at first glance they often appear concerned with problems of logic, language, ontology, and psychology that have little or nothing to do with soteriology. It is the underlying apologetic concern of his thought that I shall emphasize in this chapter by focusing on his Commentary.1

Life and Works Dharmakīrti’s dates are uncertain. Erich Frauwallner proposed 600–660, but other scholars have pushed his dates back by as much as seventy years (Krasser 2012, 580–87). The only reliable information we have about his life is that he was a famous teacher at the Buddhist university of Nālandā, but this fame may have been posthumous. Dharmakīrti himself laments in the final verse of his Commentary that his achievements were not duly appreciated by his contemporaries, and this is confirmed by reports of Chinese pilgrims who visited Nālandā in the seventh century. The delay in his renown may have had something to do with the difficulty of his works. He is considered one of the most challenging authors of the Sanskrit canon. His verses are extremely terse, practically like puzzles; his prose style is convoluted and his diction idiosyncratic. The explanations of commentators can vary considerably, though there is a consistent line of interpretation from Devendrabuddhi through Śākyabuddhi, Karṇakagomin, and Manorathanandin. Tibetan scholars felt freer to see problems in his theories and to suggest improvements. Modern Dharmakīrti scholarship has proceeded carefully, with much effort still devoted to recovering and editing his works. At this time, only portions of the Commentary have been translated into European languages, and much of it remains poorly understood. The summary offered here must therefore be considered provisional, subject to change as our knowledge of Dharmakīrti advances. DOI: 10.4324/9781351030908-27

303

John Taber

The chronological sequence of Dharmakīrti’s works proposed by Frauwallner (1954) has been generally accepted. They are: the Commentary on the Means of Knowledge (Pramāṇavārttika, PV, consisting of verses), his prose autocommentary on the first chapter of the same work (Pramāṇavārttikasvavṛtti, PVSV), the Ascertainment of the Means of Knowledge (Pramāṇaviniścaya, PVin), The Drop of Logic (Nyāyabindu), The Drop of Logical Reasons (Hetubindu), and The Rules of Debate (Vādanyāya). Two other works, The Examination of Relations (Sambandhaparīkṣā) and Proof of the Existence of Other Minds (Santānāntarasiddhi), belong to the same period as the Vādanyāya. Although probably his earliest work, the Commentary is his most ambitious. It covers the entire range of his interests and reveals how they are interrelated and arise out of his basic concern to defend the central teachings of Buddhism as rationally sound. In composing the Commentary, Dharmakīrti naturally built on the work of predecessors, as well as recent developments in Indian Buddhism. Besides Dignāga, he is clearly indebted to Vasubandhu, especially the system of Abhidharma that author works out in his Treasury of Abhidharma (Abhidharmakośa). Vasubandhu, indeed, was the giant on whose shoulders both Dignāga and Dharmakīrti stood. More broadly, Dharmakīrti was influenced by the movement of Indian Mahāyāna Buddhism known as Yogācāra, to which Vasubandhu also contributed,2 for he tends toward an idealist form of antirealism, which is distinctive of this “school.” Whether Dharmakīrti was really an adherent of a school called Yogācāra, however, is a matter of controversy, since much of the time he can simply be seen as defending the teachings of Buddhism in general, or as interpreted by more traditional, mainstream (Śrāvakayāna and Ābhidharmika) Buddhist teachers. In crafting his philosophical theories, he often set himself against established Brahmanical positions and criticisms of Buddhist doctrines, especially those of Nyāya-Vaiśeṣika and Mīmāṃsā. His archrival appears to have been the great Mīmāṃsā philosopher Kumārila – though he never mentions his opponents by name – who was probably an older contemporary. Dharmakīrti’s views were actively debated by his Buddhist and non-Buddhist successors in India up through the eleventh century. He is considered to have been one of India’s greatest philosophers. The Commentary, being ostensibly a commentary on Dignāga’s Compendium of the Means of Knowledge (Pramāṇasamuccaya), is organized in four chapters corresponding roughly to the contents of the Compendium. The question of precisely how Dharmakīrti intended it to be organized is confusing, however, due to the fact that different commentators sequence its chapters differently and none of the proposed sequences exactly matches the Compendium. The second chapter, according to what was probably the original order, treats just the initial dedicatory verse of the Compendium, which praises the Buddha using terms that express different aspects of his nature as a universal savior. The third chapter of the Commentary, on perception, corresponds to the first chapter of the Compendium, while the first chapter of the Commentary, on inference, roughly corresponds to the second of the Compendium and the fourth, on “inference-for-others” or proof, to the first eight verses of the third.3 Why the chapters of the Commentary were originally sequenced differently from the chapters of the Compendium, even though it is a commentary on it, whether in fact Table 18.1  Organization of the chapters of Dharmakīrti’s Commentary Compendium

Commentary

Dedicatory verse

Chapter 2, “The Establishment of the Means of Knowledge” Chapter 1, “Perception” Chapter 3, “Perception” Chapter 2, “Inference for Oneself ” Chapter 1, “Inference for Oneself ” Chapter 3, “Inference for Others,” vv. 1–8 Chapter 4, “Inference for Others”

304

Dharmakīrti

the above sequence is even the correct one, and why only the chapter corresponding to Dignāga’s chapter on “inference for oneself ” (the second chapter of the Compendium) is accompanied by Dharmakīrti’s own commentary – these are matters of speculation going back to classical times.

Buddha as a Means of Knowledge The first of five epithets that Dignāga applies to the Buddha in his dedicatory verse is “tantamount to a means of knowledge” ( pramāṇabhūta).4 The second chapter of Dharmakīrti’s Commentary, which analyzes Dignāga’s dedicatory verse, attempts to demonstrate this claim: the Buddha was fully awakened, knew the path to salvation, and taught it to his followers. We can be confident, therefore, that he was and, through the scriptures that preserve his teachings, continues to be a reliable source of knowledge with regards to not only how to escape suffering but also everything else he talked about. Dharmakīrti appropriately begins this chapter by defining “means of knowledge” ( pramāṇa), a concept that already had a long history but had not yet been given a general definition by a Buddhist author. We shall return to Dharmakīrti’s theory of pramāṇa in what follows. After dispensing with the idea that God (īśvara) could be a means of knowledge (2.10–28), he asserts that the “proof ” that the Buddha is a means of knowledge is his compassion, cultivated over many lifetimes (2.34a); for it was his desire to alleviate the suffering of others that drove him to practice “in many ways, without interruption, over a long time” the means of eliminating his own suffering, in particular, the perception of selflessness (2.135–36). Having become acquainted, through his own experience, with the various practices for eliminating suffering – that is, conditioned existence or saṃsāra – he was eminently qualified to teach the way to others (2.131cd–132ab). The very idea of cultivating compassion and practicing the means of eliminating suffering over many lifetimes, however – along with the notion that the Bodhisattva (the Buddha prior to enlightenment) was actually successful in completely and irreversibly eliminating the cause of suffering in his own case – raises certain philosophical questions. Does desire, as the Buddha claimed, really cause suffering? Is it possible to overcome desire and other moral defects such as aversion and delusion completely? Does the vision of selflessness really eliminate it? Indeed, is there really no self? And is there such a thing as rebirth (even though there is no self), so that it is possible for the Bodhisattva to have carried out his practice over many lifetimes? The second chapter is taken up with answering these and other questions, which concern some of the underlying presuppositions of the Four Noble Truths and the Buddhist path as a whole. Dharmakīrti suggested already in the first chapter of the Commentary that we can infer the truth of all that the Buddha taught by confirming the “main topic” of his teaching – that is, the Four Noble Truths (1.217). We shall consider this strategy for justifying belief in the Buddhist scriptures in what follows. The second chapter, then, can be seen as establishing for ourselves, using reasoning, the central teaching of the Buddha, from which we may infer the truth of all that he said as it has been transmitted to us in the Buddhist scriptures. To be sure, Dharmakīrti does not attempt to prove the Four Noble Truths directly; he does not attempt to prove, for instance, that everything conditioned is really suffering. But he does attempt to show that they cannot be overturned by considerations of psychology or metaphysics.

Critique of Physicalism From PV 2.34–79, Dharmakīrti considers doubts about rebirth based on materialist assumptions, arguing that it is more reasonable to believe that cognitions – not the body – are the cause of cognitions. In effect, he embarks on an ambitious critique of physicalism, which will 305

John Taber

continue to 2.175. At 2.34b’c, a materialist philosopher objects, “Since cognition depends on the body, repeated practice of compassion is not established for more than one lifetime.” Indian materialists, called Lokāyatikas or Cārvākas, believed that consciousness arises from the elements, like the power of intoxication arises from fermentation. Moreover, they believed that there is no world beyond this one; hence, “liberation is nothing but death.” When refuting materialism, Dharmakīrti proceeds primarily negatively – he does not offer any formal proof of rebirth of his own, but instead points out weaknesses in the materialist position. If consciousness could arise just from the body, then “there would be absurd consequences” (2.35– 36a’), by which he seems to mean that everything would be conscious. Life and consciousness indeed appear in a wide variety of kinds of matter. Of living beings, according to ancient Indian belief, there are the egg-born, the sprout-born, the womb-born, and the moisture-born (e.g., maggots coming forth from rotting food). It seems, then, that all matter is potentially sentient. But, then, why does sentience actually appear only in certain kinds of matter and not everywhere (2.37–38)? If one maintained that it is due to some transformation ( pariṇāma) of the matter in those bodies in which it appears, why does that transformation occur only in those forms of matter and not others? There is also a lack of correlation between the body and consciousness, Dharmakīrti argues. Cognitions are sequential; they arise successively, not all at once. But the body appears to be relatively static, and hence unsuitable as the cause of something constantly changing. If one were tempted to attribute change (one might say, plasticity) to the body as well in order to allow for the possibility of a causal relation, why do that when every cognition is always preceded by some other cognition (2.43–44)? What we observe, actually, is one stage of the body endowed with cognition and various faculties producing an immediately following stage endowed with cognition and faculties, which implies a causal correlation between the same types of entities (2.36, 41ab). “Let only what is seen to be the cause always be the cause!” Dharmakīrti urges (2.44cd). Nor could changes in the humors of the body – bile, phlegm, and wind – account for the required plasticity (2.54 ff.). Dharmakīrti also rejects the idea that the body could be the “support” (āśraya) of cognitions, analogous to the way a plate is the support of the berries resting on it, on the grounds that the very idea of a support is incoherent. “[The body] is not the support, because both existing and nonexisting [things] do not rest on [anything as their support]” (2.63ab’). A non-existing thing obviously could not have a support since it does not exist, but an existing thing would not need one as the cause of its continuity, either, since it already exists and its continuity is nothing different from its existence. Something can be called a “support” only insofar as it causes a series of momentary entities to occur in the same place, not insofar as it props up or assists the continued existence of an already existing entity (2.66cd–67; Franco 1997, 139–40). Rejection of the notion of a support has ramifications beyond the debate about cognition. Brahmin philosophers, who also were not materialists, believed the self (ātman) to be the support or substratum of mental states and, more generally, believed a substance to be the support of its properties. There could conceivably be a support of things like water in the sense of a “container” that prevents it from flowing away, but how could there be a container of motionless things like qualities, universals, and actions (2.68)? Rebirth is not the only thing necessary for the Buddha to have cultivated truly extraordinary compassion over many lifetimes, unadulterated by opposing tendencies. Especially if it is a matter of compassion toward all living beings, it must be able to increase indefinitely. Is that even possible? Most qualities can only increase up to a certain limit. No matter how much one practices jumping, one will never be able to jump a mile, and water can only be heated up to the boiling point before it begins to evaporate. Dharmakīrti argues, however, that mental 306

Dharmakīrti

qualities like compassion have no limit because as one cultivates them, they become part of one’s very nature and reproduce themselves without effort. Every additional effort, therefore, will contribute to their growth so that they can increase indefinitely. Jumping, on the other hand, requires effort at every stage just to sustain the level that has been reached; additional efforts yield diminishing returns, while the accumulation of heat in water is limited by the instability of its medium (2.120–31). The cause of suffering is desire or thirst, so the elimination of suffering depends on the elimination of desire. But what causes desire? One of the main themes of PV 2 is that desire is caused by the false belief in a self. One overcomes suffering, then, ultimately by abandoning the belief in a self (2.135, 140cd–141a, 185, 191, 199–201). This commits Dharmakīrti to refuting all other naturalistic theories of desire – in particular, that it is a property (dharma) of the humors (2.147cd–157) or of the elements (2.158–75). Thus, his critique of physicalism continues. Desire, aversion, and delusion are not properties of the humors “because of discrepancy” (2.147cd’). There is no correlation between dominance of a particular humor and the manifestation of some moral defect. And if desire were a property of all the humors collectively, then everyone would have the same desire. What actually causes desire to increase is pleasure (2.151c’d). Whatever physiological state one is in when one feels desire, one cannot at that time also experience aversion simply because desire and aversion are opposed, but “there is no restriction in regard to the humors” (2.156). Even when phlegm predominates, one can experience anger; even when bile predominates, one can experience desire. By similar reasoning, Dharmakīrti says, desire, etc., cannot be a property of the elements. Yet he introduces other arguments. “Since they have different appearances,” desire, etc., cannot be identical with or inseparable from the elements, as the power of intoxication is inseparable from liquor (2.160–61; cf. 169cd–171ab). One cannot even attribute the arising of desire to a particular object one perceives (for instance, someone of the opposite sex), since whether it arises or not has to do with how we conceive of the object (2.174cd).

Critique of Brahmanical Soteriology Thus, for Dharmakīrti, desire, the cause of suffering, arises from a cognition with a particular content, ultimately from the erroneous apprehension of a self and what belongs to a self. Hence, desire can be eliminated – the error can be corrected – and suffering really can come to an end (2.176–78)! In PV 2, however, Dharmakīrti does not attempt to prove that there is no self. In fact, nowhere in his corpus does Dharmakīrti work out a proof of the non-existence of a self, per se. What he offers instead, as we shall see, is a proof that all phenomena are momentary, which implies the non-existence of anything alleged to be permanent including, besides a self, God, words, and universals. Elsewhere in the Commentary and in other works, he also takes aim in passing at questionable inferences introduced by Brahmin philosophers as proving a self – for instance, the Nyāya argument from the fact that the body is breathing (see what follows) and the Sāṃkhya argument from the fact that the body is composite (since what is composite is “for the benefit of another”) (Eltschinger and Ratié 2013, 117–86). In PV 2, however, Dharmakīrti is not concerned with the question of whether the self exists, but rather with the question of whether any notion of a self is compatible with the elimination of desire. Brahmin philosophers argue that it is not the idea of a self that is responsible for rebirth, but attachment to one and especially to “what belongs to the self ” – the body, mind, senses, feelings, and cognitions – or else confusion of the self with primordial nature ( pradhāna) or matter ( prakṛti), specifically, the intellect (buddhi), which is a transformation of primordial nature. Liberation from the cycle of rebirth, which Brahmin philosophers also 307

John Taber

consider suffering, is gained by no longer mistaking body, mind, senses, and so on for the self or by discriminating (viveka) the self from everything material, including the intellect. In short, bondage is due not to belief in the existence of a self but to confusing self and not-self, which engenders attachment to the latter as one’s own self, whose precarious existence one seeks to preserve and enhance (Eltschinger and Ratié 2013, 188–99). Dharmakīrti challenges this view of liberation at length (2.220–56) on the grounds that attachment cannot be eliminated so long as one believes one has a self at all. One is attached to something because one sees good qualities (guṇas) in it. Now, existence is a good quality; therefore, so long as one believes that the self exists, and especially if one regards it as pure and free of blemishes, one will be attached to it (2.221–22). In that case, however, one will also be attached to the body and so forth because they are means of providing enjoyment for the self. More basically, though, one is attached to anything one believes is one’s own. We are not attached to the hair clippings that fall to the floor, but we love our hair when it is “ours,” still on our heads, simply for that reason. We are even attached to a limb bitten by a snake or a defective sense faculty so long as we consider it our own (2.226cd–228, 243–44). In short, we do not consider body, mind, and senses to be ours because we see good qualities in them, but we see them as having good qualities and are fond of them just because we consider them ours. Therefore, so long as there is belief in the existence of any kind of self, attachment to what belongs to the self cannot be broken – and suffering will continue. Dharmakīrti further criticizes Brahmanical accounts of liberation because they envision the self as completely dissociated from body, mind, and senses, and hence as unable to experience anything. A self that is not an experiencer, however, is not really a self. Thus, the self-theorists are inconsistent. They “absurdly defend the existence of a self while seeking to reach a state where the self as they define it ceases to exist” (Eltschinger and Ratié 2013, 282). Better just to accept that there is no self in the first place! Knowledge that there is no self brings about a complete “transformation” of one’s moral state, the destruction of all moral defects, including desire, together with the mental impressions that give rise to them (2.205ab). This transformation is irreversible, Dharmakīrti argues. An opponent – probably a Mīmāṃsaka, for Mīmāṃsā was highly skeptical of the claim that there can be morally perfect, omniscient beings such as the Buddha – suggests that even someone for whom the practice of compassion and right conduct has become natural could relapse and re-acquire moral defects such as desire, just as someone ruled by the defects can become compassionate and begin to practice the Path (2.205cd’). Dharmakīrti responds that the awareness of selflessness, which undermines desire and the other defects, is inherently stable. It is of the nature of cognition to know things as they are. The mind is “naturally luminous; impurities are adventitious” (2.208ab). One only stops knowing something if one’s belief is overturned by another cognition that reveals things to be otherwise. You cannot “unknow” something you know unless presented with evidence that contradicts it – and there is certainly no conclusive evidence that there is a self (Pecchia 2015, 219–37).

The Authority of Scripture In summary, the Buddha is a pramāṇa because, moved by compassion for all living beings, he diligently practiced the various methods of eliminating suffering over many lifetimes until he hit upon the right one, which finally enabled him to realize that there is no self. This in turn purified him of all moral defects, especially desire, all of which qualifies him as a knowledgeable, compassionate, and truthful guide for those seeking to eliminate suffering for themselves (2.280–82; Vetter 1984, 169–71). Some of the details of this story, however, derive from the 308

Dharmakīrti

Buddhist scriptures. How do we know that they are a reliable means of knowledge, a pramāṇa? If one answered “Because they are the word of the Buddha,” then one would obviously be reasoning in a circle. In the first chapter of the Commentary, Dharmakīrti in fact discusses different ways in which one can justifiably be confident in the truth of a scripture. Scripture is distinguished from other types of discourse in that it purports to tell us about things beyond experience; however, scripture also contains statements that we are able to verify for ourselves. Portions of the Veda, for instance, contain magical incantations (e.g., the Atharvaveda) and medical prescriptions (the Ayurveda). By verifying the truth of those portions of the Veda, by trying out the prescriptions and incantations to see if they work, we can be confident in the reliability of the author or authors of the Veda and from that infer that the entire Veda is true. This is actually the strategy for defending the authority of the Vedic scriptures that was adopted by Nyāya. Dharmakīrti modifies it for his own purposes in PV 1. There, he suggests that there are two main ways we can determine if a scripture is reliable. One is to confirm that it is right about empirical matters; specifically, that what it says is perceptible is in fact perceptible and what it says is not perceptible is in fact not perceptible; and that what it says is inferable independently of scripture (such as the Four Noble Truths) is indeed so and what it says is not inferable (God, the self) is, likewise, indeed not so. It should also be free of internal contradictions (1.215). Dharmakīrti, in the first chapter, argues that the Veda fails to meet these criteria. Another method is simply to confirm the reliability of a scripture in regard to “the main topic” of its teaching relating to “what is to be obtained and what is to be avoided and the means thereto,” that is to say, the path to liberation, from which one may infer its reliability in regard to all transcendent matters (1.217). As we have seen, PV 2 attempts to corroborate the Four Noble Truths not by proving them directly, but by showing that they – or at least the philosophical-psychological presuppositions on which they are based – withstand rational scrutiny. Therefore, we can be confident that everything in the Buddhist scriptures, even pronouncements about matters beyond our experience, is true, just because the Buddha is their author.5 When discussing how scripture can be reliable, however, Dharmakīrti makes a surprising concession. The apparent inference by which one concludes from the truth of one thing a person says the truth of everything he says is actually a fallacy; it is the same sort of fallacy one commits when one draws the conclusion that all of the rice in a pot is cooked from sampling just a couple of grains. Hence, strictly speaking, scripture is not a pramāṇa, Dharmakīrti says, or at least not a “real” one. People resort to it – justifiably – because they have no other way of finding out about important transcendent matters such as the workings of karma (PVSV 108,1–6; 109,19–22; 167,23–168,3; 173,22–174,2). This opinion may have led to tension between Dharmakīrti and his coreligionists – here he is supposed to be defending Buddhism, but now he says the Buddhist scriptures are not a bona fide means of knowledge! – and could have been one of the reasons why he chose, in his later writings, to focus on more technical logical, epistemological, and metaphysical questions. It is his contributions to these fields – for which he is most famous – to which we now turn.

Definition of Means of Knowledge and Levels of Reality Despite all the discussion and debate surrounding Dharmakīrti’s concept of pramāṇa in classical and modern times, he actually says very little about it in PV 2. There he offers two definitions, which are probably intended to be equivalent (Krasser 2001). Since ancient times, Indian scholars had discussed the kinds of evidence that should be cited in support of theses advanced in debate or scientific treatises. These were referred to as pramāṇas, literally, “means 309

John Taber

of knowledge”; the discussion initially focused on how many legitimate pramāṇas there are and their individual definitions. The influential Nyāya tradition, for instance, recognized four pramāṇas – perception, inference, scripture (more generally, verbal testimony), and comparison or analogy – and they offered precise definitions of each one. The Sāṃkhya school, on the other hand, recognized only perception, inference, and scripture, and gave different definitions of them. Dignāga, meanwhile, in his Compendium accepted only perception and inference – he regarded scripture as a form of inference. We have already seen how Dharmakīrti, following Dignāga, also considered the authority of scripture to be based on an inference, albeit a fallacious one, and again offered his own definitions while critiquing the definitions of the other schools. The nature of a pramāṇa in general, however, had been neglected. It seems that prior to Dharmakīrti, only the Mīmāṃsā philosopher Kumārila had ventured to offer a substantive general definition as opposed to a mere enumeration. The definition of pramāṇa became a matter of greater urgency now that Dharmakīrti was keen on defending Dignāga’s claim that the Buddha was “tantamount to a pramāṇa.” Dharmakīrti’s first definition of pramāṇa is “a nondeceptive cognition” (2.1ab’), whereby “nondeceptive” means reliable or not leading one astray. Dharmakīrti then immediately clarifies that non-deceptiveness consists in “the stability of the causal efficacy of the object” (2.1b’c’). That is to say, a cognition is a pramāṇa if, acting on it, one experiences the desired effect of the object it represents. For instance, a visual perception of a glass of water is non-deceptive if, by prompting one to go over, pick it up, and drink it, it leads one to experience the quenching of one’s thirst, the typical desired “causal efficacy” of water. The continuity or stability (sthiti) of the causal efficacy of the object may refer to the fact that, having initially produced a cognition of itself, the object of a pramāṇa (or, more precisely, a causal successor of it in the same series of momentarily existing objects) will later produce an experience of its causal efficacy.6 Some commentators, however, have specifically taken sthiti to mean the “capacity” ( yogyatā) of a cognition to be a pramāṇa. Even if one does not act on the cognition, it is a pramāṇa if it would lead one to experience the causal efficacy of its object if one did act on it. Thus, my visual perception of the jewels on the neck of the queen can be a pramāṇa, even though I will never get close enough to touch them. I know that my perception is a pramāṇa in this case – that it would result in an experience of the causal efficacy of its object – because my perception is similar in nature to others I have had in the past that have always been confirmed. In general, any perception of a familiar kind of object under normal circumstances will be known to be a pramāṇa (for instance, my perception of the glass of water on the table just in front of me) without actually having to experience the causal efficacy of its object (i.e., by picking up the glass and drinking from it). Thus, while the paradigm case of a means of knowledge for Dharmakīrti is a cognition which is confirmed “extrinsically” by acting on it and experiencing its object’s causal efficacy, we know many of our cognitions to be pramāṇas “intrinsically,” according to his commentators. Besides the “frequently repeated” type of cognition just mentioned, the cognition of the causal efficacy of an object itself does not require further confirmation in order to be (known to be) a pramāṇa; in fact, this is obviously required to avoid an infinite regress. For Dharmakīrti, all our knowledge is ultimately anchored in cognitions of the causal efficacy of objects whose status as means of knowledge is known intrinsically. In this way, he can be seen as a kind of foundationalist, though the dichotomy foundationalism – anti-foundationalism was not explicitly thematized in Indian philosophy: all empirical knowledge is rooted in cognitions which, though they may not be “infallible,” do not require justification by further cognitions in turn. Scholars have often depicted Dharmakīrti as holding a pragmatist theory of justification, however, since his view also suggests that a cognition is a pramāṇa if one may act on it successfully. Finally, inferential cognitions are also known to have the status of means of knowledge 310

Dharmakīrti

“intrinsically” (that is, without having to confirm them), since in order for an inferential cognition to arise at all, the knower must apprehend a necessary connection between reason and consequent, which guarantees its truth. An inferential cognition is indeed devoid of any doubt. Dharmakīrti’s second definition of pramāṇa is “the presentation of a previously unknown object” (2.5c). We can see this formulation as making explicit an essential feature of a pramāṇa that is only implicit in the first definition. In order for a cognition to be non-deceptive by virtue of the constancy of the causal efficacy of its object, it must be able to motivate one to try to obtain its object, at least potentially. It is, however, one’s initial apprehension of something that motivates one to obtain it, not the second or the third apprehension; hence, a pramāṇa will present a new object, one that is not already known. What Dharmakīrti is particularly keen to rule out as a pramāṇa here is the “perceptual judgment” (adhyavasāya) – the conceptual identification of an object – that arises immediately upon perceiving it. (In the case of the first definition, this had to be ruled out by an additional phrase, PV 1.3ab’; here, in the case of the second definition, its exclusion is implied by the definition itself.). Most Indian philosophers believed perceptual experience to be a two-stage process: first, there is a bare awareness of the object, a “non-conceptual perception,” then a judgment that identifies the object as a particular type of thing (a “cow”), having a certain property (“brown”) or motion (“walking”), and so forth. According to realist philosophers – Mīmāṃsakas and Naiyāyikas – this perceptual judgment is actually a “conceptual perception”; that is to say, it is itself a variety of perceptual experience that apprehends a real property of the object, hence a means of knowledge, since perception is acknowledged by everyone to be a pramāṇa. We shall see, however, that Dharmakīrti thinks that conceptual awareness pertains to the domain of thought, but not perception; conceptual awareness removes one from an immediate experience of the object as it is. Moreover, the properties attributed to things by thought are imagined, not real. A conceptual judgment, then, is not a pramāṇa by the second definition because it apprehends an object that has already been made known by perception;7 it is in fact a variety of memory. One might think that this second definition fails to capture the non-deceptiveness of a pramāṇa, since it does not explicitly say that a pramāṇa is a cognition that is confirmed by the experience of causal efficacy. But commentators suggest that this what is implied by the word “object”: an object in the fullest sense is real, hence something whose causal efficacy can be experienced. Dharmakīrti, following Dignāga, recognizes only two distinct pramāṇas: perception and inference. Most Indian philosophers before Dignāga and Dharmakīrti recognized at least scripture in addition to these two, and some identified several others (for example, comparison, abductive inference, and cognition of absence). Dharmakīrti justifies the reduction of the number of pramāṇas to only two at the beginning of the third chapter of his Commentary by declaring that there are only two kinds of object to be known: the “particular characteristic” (svalakṣaṇa) and the “universal characteristic” (sāmānyalakṣaṇa). A particular characteristic (or simply, particular) is that which is capable of causal efficacy and is what exists “ultimately” ( paramārthasat). In contrast, what lacks causal efficacy exists only in a dissimulated or “concealed” way (saṃvṛtisat), or as it usually rendered, “conventionally,” and the universal characteristic (or simply, universal) belongs to this domain (PV 3.3). While Dharmakīrti largely presupposes this ontology as he works out the details of his epistemology, we shall see that he eventually abandons it. Already, immediately after introducing the distinction between conventionally and ultimately existing, he appears ready to concede to an opponent that causal efficacy, the criterion of the ultimately existent, itself exists only conventionally (PV 3.4). Indeed, in his Ascertainment of the Means of Knowledge, he will say that the theory that defines a means of knowledge in terms of non-deceptiveness and distinguishes perception from inference pertains only to everyday practice (sāṃvyavahārika), not ultimate reality. 311

John Taber

The purpose of his epistemology, therefore – apart from clarifying the status of the Buddha as tantamount to a pramāṇa, and the nature of the means by which one properly reflects on his teachings – is to refute mistaken views of non-Buddhist thinkers whose epistemological theories promote a wholly different picture of reality from the one implied by Buddhist teachings and, hence, ultimately prevent the overcoming of suffering. At the heart of their conception of reality are: (1) belief in the existence of real universals, which we apprehend in true judgments as the properties of particulars; (2) belief in the existence of entities that endure through time – besides universals, selves and other substances; and (3), most fundamentally, belief in the existence of a world external to and independent of cognition. We shall see how Dharmakīrti’s epistemology renders these beliefs untenable in what follows. In general, Dharmakīrti’s epistemology has a dominant polemical tone, at least in the Commentary, rooted in his apologetic purpose (Eltschinger 2010, 405–6).

Universals, Exclusions, and Momentariness Although inference, according to Dharmakīrti, does apprehend universal characteristics insofar as it apprehends the properties of individuals (for instance, the fieriness of a mountain from the smoke seen rising from it), it is crucial to recognize that it only apprehends pseudouniversals  – that is to say, conceptually constructed universals, not real ones. Dharmakīrti develops a sustained, relentless critique of real universals throughout his Commentary. Two passages in particular, however, stand out: PV(SV) 1.143–84 and PV 3.11–50.8 Based principally on the latter, one may reconstruct his main argument as follows (cf. Franco and Notake 2014). A universal is not real, because it is the referent of a word. If it were the referent of a word and real, then the sense faculties would become superfluous, for if a universal were real, it would pertain to the unified, partless nature of the object that “has” it. In that case, having already apprehended the nature of the object by a verbal cognition arising from an utterance such as “A cow is standing in the pasture,” the senses could not tell us anything new about it; as a result, perception would not be a pramāṇa (3.11ab). If one maintained that the senses still could have a purpose because the word reveals just the universal by itself – say, cowness – while the senses comprehend the individual cow, then it must be pointed out that any awareness of a common aspect cannot really be an apprehension of a real universal, since a universal could be neither the same as nor different from the individual. It could not be different – this was the view of Nyāya-Vaiśeṣika; namely, a universal is distinct from the individuals in which it “inheres” – for, just insofar as it is distinct from those individuals, it could not be their nature and so could not really be the universal of them (PVSV 25,3–4). Nor could it be the (distinct) universal of those individuals insofar they are its support (āśraya) or substratum. We have already seen that Dharmakīrti rejects the notion of a support as incoherent. Moreover, something could only be called the support of something else if it caused it to occur continuously in the same location. A universal, however, is by hypothesis eternal; it cannot be “helped,” that is, caused to occur, by something else. Nor, for similar reasons, could the universal be “manifested” by the individuals characterized by it (3.19–26). If the universal, on the other hand, were not different from the individual – this was the Mīmāṃsā and Sāṃkhya view  – then all of the individuals falling under a universal would have the same nature; namely, that of the universal. Thus, they would not really be distinct individuals; but a universal by definition is a universal of many individuals (3.41ab; PVSV 24,24–25,3). And if an individual had both a common and a distinct nature, its nature would not be cohesive or united (3.41cd). Thus, neither the same nor different from the individual that “has” it, a universal is undefined; it is “without nature” (niḥsvabhāva, 3.27ab). Most 312

Dharmakīrti

significantly, Dharmakīrti insists, something having the form of a universal simply does not appear in experience (3.48ab; cf. PVSV 39,21–23). Indeed, it is difficult to see how a universal – being eternal – could cause a cognition of itself at all. Therefore, the universal characteristic that is known by inference is not a real universal but something else, an “exclusion” (apoha), which is a kind of pseudo-universal. An exclusion is a difference shared by many particulars, in themselves quite distinct from each other, from other particulars that lack certain distinctive effects. All cows, though individually distinct from each other, give milk; they also produce a certain cow-like cognition. The exclusion “cow” is the difference of cows from other things that do not have such effects – horses, humans, and so on. It is on the basis of such an exclusion that we consider cows to be the same kind of thing, though they do not share a common nature. Quite distinct things can have the same effects, Dharmakīrti points out, citing the example of different herbs that can reduce fever (1.74). The fact that several things have the same effect(s) has simply to do with their individual essences and nothing they have in common. Discerning an exclusion, however, one erroneously conceives a unique particular to be the same as others; one believes there is cowness inhering in, or not different from, all particular cows. Similarly for other properties, brownness, walking, and so forth. However many differences a thing is conceived to have from other entities, Dharmakīrti says, it is thought to have as many properties (1.40–42). The attribution of common properties to particulars, which obscures their uniqueness, is the essence of “concealment” (saṃvṛti) – that is to say, ignorance (avidyā). Concealment defines the “conventionally existent” (saṃvṛtisat), the world of everyday activity (PV[SV] 1.68–70). Conceptual thought has other distorting effects besides obscuring uniqueness. When one attributes multiple properties to a single particular one veils its inherently unified, partless nature, as one does when one distinguishes the particular as a property-bearer from its properties (PVSV 34,23–35,9; Eltschinger et  al. 2018, 63–65). The theory of apoha, then, can perhaps be seen as the first step Dharmakīrti takes in the direction of antirealism. It calls into question the world of objects – chairs and tables, dogs and cats, which other philosophers consider (enduring) substances of certain types possessing multiple properties – that we take for granted in everyday activity. It interprets them as, essentially, imagined. The theory of exclusion was developed by Dignāga in the fifth chapter of his Compendium as a Buddhist theory of the meaning of general terms designed to counter Brahmanical theories of words as referring to real universals. Dharmakīrti expands and modifies the theory to meet criticisms of Dignāga’s account in a lengthy passage of PV 1 (PV[SV] 1.40–185; Eltschinger et al. 2018). There, he analyzes exclusions both as the meanings of words and as objects of thought, and he explains in detail the psychological process through which exclusions are conceived on the basis of experiences of particulars (cf. Dunne 2004, 116–26). Hence, although an exclusion itself is not real – as a difference, it is essentially just a negation – it ultimately relates to real particulars. It is for this reason that an inferential cognition, e.g., a judgment such as “the mountain is on fire” derived from the observation that it is smoking, can be “nondeceptive,” hence a pramāṇa, even though it actually apprehends the mountain in a distorted way by ascribing to it a common nature which obscures its uniqueness; for to think that something is on fire is ultimately just to associate it with other (distinct) particulars that would produce the experience of certain effects under the right circumstances. In other words, it is just to predict that, if one approached it, it would cause a sensation of heat. From this, incidentally, we are able to see that the concept of being a means of knowledge or “epistemic instrumentality” ( prāmāṇya) does not precisely map onto the notion of truth. Dharmakīrti also insists repeatedly throughout his discussion of apoha that exclusions are assigned as the meanings of words by convention. There is no natural, eternal relation of a 313

John Taber

word with its meaning; rather, it is completely arbitrary. By contrast, Mīmāṃsā philosophers, who sought to ground the authority of the Veda on its authorlessness, were committed to the view that word, meaning, and the relation between them are eternal. A considerable portion of PV 1 (1.213–68; Eltschinger 2007) is devoted to refuting the Mīmāṃsā doctrine of an authorless scripture, along with the various linguistic theses that underpin it, while the final passage of the chapter (1.312–40) seeks to demonstrate that, even if the Veda were authorless, we could never know how to interpret it; we could never know what it really means (Eltschinger, Krasser, and Taber 2012). We have seen that Dharmakīrti asserts that particulars are “ultimately real” because they have causal efficacy. The capacity to causally influence things is an entirely plausible criterion of reality that other philosophers have embraced. Yet Dharmakīrti employs it to demonstrate a central thesis of the Buddhism of his day that is intuitively much less plausible, the momentariness of all phenomena. That is to say, things lack any temporal duration; they cease to exist as soon as they arise. Dharmakīrti actually offers two proofs for this thesis. One is based on the idea, accepted by all Buddhists, that things are impermanent or perishable, together with the claim that their perishing could not be caused by outside factors. If something ceased to exist due to outside factors, it could turn out to be eternal if those factors did not materialize. Moreover, ceasing to exist is commonly held to be a kind of “non-existence” (abhāva), and a non-existence cannot be caused. Things must, therefore, perish spontaneously, of themselves; transience is part of their very nature. In that case, they will perish as soon as they arise (PVSV 98,4–22; 100,8–19; Steinkellner 1968/1969). The other proof turns on the idea that phenomena are momentary due simply to the fact that they exist, which implies that they are causally efficacious. Dharmakīrti focuses on establishing the contrapositive of the statement, “All that exists is momentary”; namely, “All that is non-momentary is nonexistent.” He argues that something that is not momentary could not produce an effect either gradually or all at once. Something non-momentary that is capable of producing an effect would not delay in producing it; if it did not produce it as soon as it existed, insofar as its causal capacity is part of its nature, it would not do so later, since that would involve a change in its nature. Hence, it would not produce its effect gradually. But if it could produce its effect in a single moment, it would do so continuously, as long as it exists (which is never observed). Since, then, a non-momentary entity could not produce an effect in either of these ways, it would not exist (PVin 79,1–4; 80,1–6). The proof of momentariness is often seen as the cornerstone of Dharmakīrti’s metaphysics. It supplements his refutations of universals and the various eternalist linguistic theses that underlie the doctrine of an authorless Veda, and he may have felt that it made any explicit proof of the non-existence of a self unnecessary.

Perception In addition to a general definition of a means of knowledge, Dharmakīrti, of course, provides individual definitions of each of the two pramāṇas he recognized. In defining perception, he stays close to Dignāga’s definition, “a cognition free from conceptual construction,” adding only the qualifying expression “non-erroneous” to rule out cognitions arising from a sense faculty that is diseased or defective, but which are nevertheless non-conceptual: “Perception is a cognition that is free from conceptual construction and non-erroneous” (PVin 1.4ab’). The core idea is that a perception is a bare awareness of a particular as it is, which takes in all aspects of its nature – which is nonetheless “unified,” in that its various aspects are somehow inseparable from each other – without identifying it in any determinate way according to a genus, quality, motion, accessory (as in the determination “a man with a stick”), or proper name. 314

Dharmakīrti

While Dignāga understands “conceptual construction” (kalpanā) to be the association of a word with the object, Dharmakīrti takes it to be a cognition that represents something capable of being associated with a verbal designation (abhilāpasaṃsargayogya). As already mentioned, non-Buddhist philosophers believed an initial “non-conceptual” perception to be followed immediately by a “conceptual” one, which is caused by the former and merely articulates (typically, by the application of words) real features of the object that is non-conceptually apprehended. Obviously, Dharmakīrti cannot allow this second, conceptual awareness to be a perception. He has already attempted to rule it out, if only in an ad hoc way, by defining pramāṇa (explicitly in his second formulation, implicitly in the first) as a cognition that presents a previously unknown object. More fundamentally, however, we have seen previously that Dharmakīrti considers concepts to misrepresent or “conceal” the true nature of particulars. While it may be possible for inference to misrepresent its object by apprehending it conceptually, yet still be a pramāṇa insofar as it ultimately leads to an experience of its causal efficacy, there must be some mode of awareness that puts us directly and faithfully in touch with things (i.e., that reveals how they really are) for us to be confident that empirical knowledge connects us up with the world. In PV 3.123–191ab and PVin 1.5–14, however, Dharmakīrti raises additional considerations to show that the conceptual awareness or “perceptual judgment” that arises immediately upon the (non-conceptual) perception of a particular lacks the qualities we would normally associate with a perception. First, Dharmakīrti maintains that it is self-evident that perceptions are devoid of conceptual content. As we shall see, he believes that all cognitions are “selfaware”; they cognize themselves along with their objects. Thus, every cognition reveals its own character as a cognition as it occurs. A perceptual cognition immediately reveals itself as the awareness of some object-form devoid of conceptualization. This is confirmed by the experience of meditation. Even when thought has completely ceased, one can still have visual perceptions (presumably, meditating with the eyes open) (PV 3.124–25). Second, as we have seen, conceptual awareness involves the application of an expression to what is being cognized, at least potentially. Since a verbal expression cannot refer to a temporally and spatially restricted particular – for a linguistic convention could not be established in regard to such an ephemeral thing – conceptual awareness cannot have a particular as its object. Therefore, a perception, which apprehends a particular, cannot be conceptual (PV 3.126–28). If one held, however, that perception of a particular triggers a memory of the linguistic convention that relates a certain expression to particulars of that type, then the conceptual awareness would still be “removed” from the object. It would not be an immediate, direct awareness of it, but rather an awareness processed by the mind (PVin 1.5–7). Finally, and perhaps most fundamentally, perceptual experiences are vivid, while conceptual awarenesses – occupied not with particulars but the meanings of words – are not (PVin 1.15; PV 3.130). Dharmakīrti advances many other arguments in support of this thesis. Dharmakīrti identifies four types of perception. Besides sense perception there is mental perception, of somewhat mysterious status (it seems to be a sort of non-conceptual mental image of the next moment of the object apprehended by a sensory perception); the self-awareness of any kind of cognition (even a conceptual awareness), already mentioned; and yogic perception, the direct experience of an object, which is the culmination of meditation on it and which presents it in a vivid, non-conceptual way (3.285). The latter is of particular importance for Dharmakīrti. He notes that anything we think about intensely and long enough will eventually present itself as if it were immediately in front of us – even an “unreal” (abhūta) object, such as an imagined corpse one is meditating on in order to reduce desire for the body. Dharmakīrti held that the Buddha achieved such a direct perception of the real features of things – selflessness, momentariness, and suffering, as well as the Four Noble Truths – through 315

John Taber

constant meditation on them, having first known them through insight gained by understanding the meaning of scripture and rational reflection (Eltschinger 2010, 422–23).

Antirealism Dharmakīrti’s epistemology, as presented thus far, secures a world of causally efficient, hence momentary, particulars which are in themselves utterly unique, that is, completely distinct from each other, but having properties attributed to them on the basis of the observation of their causal powers, which define common “differences” among them. Such an account is consistent with the world-view of certain traditional (mainstream Buddhist) orientations that Dharmakīrti wished to provide room for in his system; it is, incidentally, also potentially consistent with certain modern versions of empiricism. In the latter half of the third chapter of his Commentary, however, he embraces a Mahāyāna perspective, rejecting the reality of a world of external particulars and advocating instead the distinctive Yogācāra teaching that everything is just consciousness (vijñaptimātra). There are three components to his demonstration of consciousness-only: (1) a deconstructive analysis of the object of perception; (2) proof of the thesis that the object of a cognition is not distinct from the cognition; and (3) development of the implications of the idea that all cognition is self-aware. (1) When discussing, at PV 3.197–219, the object of perception, which presents itself as something spatially extended  – a tree, a piece of fruit, and so forth  – Dharmakīrti initially entertains the idea that it is a collection of atoms which, though individually invisible, become visible when they arise in close proximity to each other. (For Dharmakīrti, following Abhidharma thought, atoms are not eternal, partless substances but momentary, autonomous properties – solidity, fluidity, and so forth.) In that case, the object is something variegated; it consists of many visible atoms, like a pile of sesame seeds. How, then, does it appear as one thing with a single, extended form – a tree or a cow? Rejecting the Nyāya proposal that the object is a single whole (avayavin) inhering in many parts, he also questions whether its oneness can be accounted for by the cognition that apprehends it. In order to cognize an object the cognition must somehow bear the image, or assume the form, of the object it represents. If the object is variegated, the cognition, too, must be variegated. But if the cognition itself is variegated, it cannot be one; nor can it really even be variegated (3.210ab) – here his precise argument is obscure, but his commentator Manorathanandin suggests that it has to do with the fact that its diverse aspects are not able to perceive each other, “like the cognitions of many [different] persons.” The cognition, then, can be neither one nor many; nor, by implication, can the object be either one or many. Alas, Dharmakīrti says, “What the wise say is forced upon us by the facts: the more we reflect on things, the more they dissolve!” (3.209). When we press the question, then, neither the object nor the cognition can have extension, yet they both appear to be extended. Hence, neither is real. Now, consciousness is structured by the opposition between an object and the cognition which apprehends it. If one of them is unreal, so is the other (and in this case, both are). Hence, the reality of cognition is “the emptiness of duality” of subject and object (3.211–13). The distinction of entities, however, which is fundamental to the empirical world as Dharmakīrti analyzes it, is grounded on the dichotomy of subject and object. If there are no objects standing over against cognition, there can be no distinct entities (3.214). Moreover, the fundamental constituents of reality identified by the Buddha, the “aggregates,” are either objective (matter) or mental (sensation, etc.) in nature. Entities devoid of the characteristics of being either an object or a cognition or aspect of cognition would be without essence (3.215). Thus, all the entities the Buddha talked about when 316

Dharmakīrti

teaching the Four Noble Truths, selflessness, and so on – namely, the aggregates (skandhas), the “bases” (āyatanas), and the “elements” (dhātus) – are in fact without essence, that is, ultimately unreal (3.216). Although himself cognizant of the true non-dual nature of cognition – hence, the non-existence of what we take to be external things – the Buddha spoke as if there were external objects for the benefit of those whose minds are afflicted by ignorance, “like an elephant closing one eye,” choosing just what is on one side of her (3.217–19) (Dunne 2004, 396–411). (2) We have seen that, according to the epistemological assumptions Dharmakīrti is working with, a cognition must somehow bear the form of the object it cognizes; it must take the object up into itself, so to speak. In a passage beginning with PV 3.320, Dharmakīrti asks whether this having the form of an object (tādrūpya) can establish that perception is the apprehension of external objects. In this passage, along with others later in PV 3, Dharmakīrti raises various problems with attempting to infer the existence of an external object from the apparent resemblance of a cognition with an object. On one interpretation, such reasoning would be inconclusive simply because we have so many cognitions which seem to be of external objects but which turn out not to be cognitions of external objects at all – dreams, for instance, or optical illusions (Kellner 2017, 110–11). But at PV 3.387–390ab, Dharmakīrti introduces what is allegedly a direct proof of the fact that one is aware only of an aspect of cognition itself, an “object-appearance” in cognition, whenever one seems to be aware of an external object. This proof, referred to as the “invariable co-apprehension” (sahopalambhaniyama) argument, can be stated very simply: cognition and object are not different, because they are invariably perceived together. As Dharmakīrti puts it in PVin 1.54ab, “Because they are invariably apprehended together, blue and the cognition of it are not different.” The major premise of the argument, as one might reconstruct it, is that two things that are invariably, or perhaps necessarily, found together are not different (consider: the commanderin-chief of the U.S. Armed Forces and the president of the United States). The minor premise of the argument is that one is not aware of an object without being aware of a cognition of the object, and vice versa. This premise expresses the principle that all cognition is self-aware. Both premises can be defended against prima facie objections; together, they entail the conclusion that cognition and object are not different. It is a corollary of the argument that everything exists in consciousness or that there are no objects external to consciousness. This was seen as a formidable argument for consciousness-only by non-Buddhist philosophers after Dharmakīrti and provoked intense debate (Taber 2020). (3) A third line of reasoning that denies the reality of the external world by transposing the experienced world within consciousness starts with the consideration of cognition as selfillumination. My own tentative reconstruction of this argument, first articulated in the Ascertainment, is as follows.9 Cognition must present something that only belongs to itself, because it is inherently self-aware. Every cognition is not just an awareness of some object-form, but also awareness of itself aware of that object-form. Thus, self-awareness is of the very essence of cognition, in which case the awareness of an object-form itself must also be a kind of selfawareness; that is to say, an awareness by the cognition of (an aspect of) itself. A cognition cannot be aware of anything outside itself. A fuller understanding of Dharmakīrti’s antirealism awaits closer study of the so-called “self-awareness section” of PV 3.425–539. Although he nowhere – to my knowledge – makes the Berkeleyan point that we cannot comprehend how a material thing could even impose an imprint of itself upon consciousness, we have seen a resistance in his thinking to any suggestion that matter and consciousness causally interact. An underlying theme of his critique of physicalism was that it is easier to explain the arising of cognitions from other cognitions than from the body. Here, he seems to be implying that the best explanation for why a cognition 317

John Taber

contains a certain object-form is that it assumes that form of itself, for this is consistent with its nature as a kind of self-awareness. “Cognition is by nature just an intransitive ‘appearing-in-acertain way,’ as opposed to a transitive apprehension of something else” (Kellner 2017, 113).

Theory of Inference Given Dharmakīrti’s concern to demonstrate the flaws of the epistemological and metaphysical theories of his non-Buddhist opponents, it is natural that he would be interested in patterns of reasoning as such. A separate entry would be required to do justice to his contributions to Indian logic. I must confine myself here to certain aspects of his treatment of inference that have particular bearing on his refutation of key Brahmanical teachings. The history of logic in India prior to Dharmakīrti was largely taken up with the quest for an adequate characterization of the relation between ground and consequent in a sound inference. By Dharmakīrti’s time, the most widely accepted view, to which Dignāga subscribed, was that a reason is a good one if it exhibits a certain pattern of occurrence. It must (1) be true of the subject of inference; (2) occur together with the property-to-be-proved in other cases; and (3) not occur where the property-to-be-proved is absent. Thus, in the inference, “The mountain is on fire, because it is smoking,” “smoking” is the reason, “mountain” the subject of inference, and “on fire” the property-to-be-proved of the subject. If the mountain really is smoking, the first condition is satisfied. The second condition is satisfied because other things that give off smoke are on fire, and the third is satisfied because nothing that is not on fire gives off smoke. Dignāga introduced refinements of this theory – in particular, he tightened the formulations of the second and third conditions so that they define a necessary connection between the reason and the property-to-be-proved – but stayed within its basic framework. Nyāya critics of Dignāga, however, pointed out certain problems with it. One question that was raised was whether in fact both the second and third conditions must be satisfied for a sound inference. It was generally understood that the statement “whatever is smoking is on fire,” which is captured by the first condition, is equivalent to “whatever is not on fire is not smoking,” expressed by the third condition. Why, then, for a reason to be a good one, do both conditions have to be satisfied, as Dignāga maintained? Why would it not be sufficient if we knew that just one of them is? In particular, the Nyāya author Uddyotakara pointed out that there seem to be inferences that are sound but have reasons that only satisfy the third condition. Take, for instance, the following inference for the existence of a self from breathing, mentioned previously: “the body must be possessed of a self because it is breathing.” One cannot establish a necessary connection between the reason “breathing” and the property-to-be-proved, “possessed of a self,” in a positive way by pointing to uncontroversial cases of things that breathe and have selves; a Buddhist, who denies the existence of a self, will naturally reject all such cases. Thus, the second condition cannot be satisfied. But all parties to the debate about the self will agree that things like rocks, furniture, and automobiles, which do not breathe, are without a self. Even the Buddhist would accept this. Hence the “negative concomitance” (vyatireka) between reason and property-to-be-proved – that is, the fact that there is nothing that lacks a self that breathes (or “everything that lacks a self is without breath”), which represents the third condition and is equivalent to the second condition – seems to be established, thereby rendering the argument sound. Thus an argument can be sound that does not satisfy both conditions (2) and (3), but only condition (3). It was partly because of such inferences that Dharmakīrti introduced a more stringent requirement for the relation between reason and property-to-be-proved than that the reason

318

Dharmakīrti

exhibit the pattern of occurrence reflected by conditions (1), (2), and (3). He proposed that a reason must, more fundamentally, be related to the property-to-be-proved by a “natural connection” (svabhāvapratibandha), which explains the satisfaction of conditions (2) and (3). Dharmakīrti recognizes two kinds of natural connection: causation and essential identity. The non-occurrence of the reason in the domain where the property-to-be-proved is absent is significant only if it is the result of one of two deeper conditions: either the reason is the effect of the property-to-be-proved, or else it is an essential property (svabhāva) of the subject of inference that is identical with another essential property of it.10 Otherwise, if neither of these conditions is met, the non-occurrence is not significant. Smoke does not occur in things that are not on fire because fire causes smoke. Simply never observing something that is not on fire to be smoking is not by itself sufficient to establish that everywhere there is smoke, there is fire. Similarly, never observing that nothing that is without a self breathes does not by itself establish that everything that breathes will have a self. One must determine that there is a causal connection between breathing and having a self, and a causal connection between X and Y, in general, can be established only by observing that, given X, Y arises and, in the absence of X, Y does not arise. Since a self is imperceptible, however, this is not something one could ever observe for a self and breathing. While it would be wrong, then, to say that Dharmakīrti revised the logical theory of his day with the express intent of refuting certain non-Buddhist arguments, it seems plausible that he arrived at logical insights as a result of reflecting on certain Brahmanical arguments, in particular, and becoming convinced that the theory he inherited failed to expose their flaws. Another argument Dharmakīrti criticizes as fallacious is one that was current among Mīmāṃsakas that impugned the authority of the Buddha, which, as we have seen, is Dharmakīrti’s central preoccupation in PV 2. The argument in question is: “The Buddha must have been possessed of desire because he spoke.” As we have seen, Dharmakīrti argues in PV 2 that the Buddha eliminated desire by eliminating its cause, the false view of a self, as a result of practicing the Path over many lifetimes. The Buddha’s freedom from desire is not just an indication of his enlightenment but also what partly justifies him as a trustworthy guide for others seeking to overcome suffering. Now, the Mīmāṃsā tradition, which was skeptical of any claims of superhuman attainment made by followers of non-Vedic religious traditions on behalf of their founders, questioned the Buddha’s special status on the grounds (among other reasons) that he spoke – which of course he had to do in order to teach. One only speaks out of some desire, the Mīmāṃsakas plausibly claimed. The Buddha spoke; therefore, he had to have been afflicted by desire – just like the rest of us. Dharmakīrti mentions this argument when developing another key point of his theory of inference at PV 1.12: a negative concomitance  – the absence of the reason in the domain where the property-to-be-proved is absent – cannot be established merely by non-observation. Dignāga, along with Dharmakīrti’s teacher Īśvarasena, held that simply the non-observation of any counterexample to the “pervasion” of the reason by the property-to-be-proved is all that is needed to establish that the reason is never found where the property-to-be-proved is missing, i.e., condition (3). Dharmakīrti, however, was aware that just because one never encounters a counter-example does not mean that one does not exist – and that, in fact, is the case when it comes to speaking and having desire. One may never have observed someone speak who did not have some desire, but that does not mean that there could not ever be such a person. Indeed, the Buddha was such a person. We must not draw the conclusion that something is never the case from not having, so far, observed it to be the case. Negative concomitance cannot be established simply by non-observation.

319

John Taber

Once again, for there to be a natural connection between reason and property-to-be-proved, the reason cannot occur without the latter. This is founded on there being either a causal relation or an identity between the two properties. In the case of a causal relation, the cause must be a necessary and sufficient condition of the effect.11 While desire, to be sure, is one of the causes of speaking, it is not the only cause. Compassion could also be the cause of teaching others the way to overcome suffering, as it allegedly was in the Buddha’s case. The cause which underlies both the positive and negative concomitance between the reason and property-to-be-proved of a sound inference must be a necessary condition for the arising of its effect.

Notes 1 In taking this approach, I  have been influenced by the groundbreaking research of Vincent Eltschinger. 2 This is assuming there was only one Vasubandhu, not two (see Chapter 16 of this volume). 3 The correspondences between the chapters of Dharmakīrti’s Commentary and Dignāga’s Compendium, as proposed by his earliest commentators, Devendrabuddhi (630–690) and Śākyabuddhi (660– 720), are presented in Table 18.1.   From Prajñākaragupta (ca. 800) on, commentators began to consider the Commentary as beginning with the “The Establishment of the Means of Knowledge,” often referred to as “The Pramāṇasiddhi chapter,” followed by the chapter on perception, the chapter on inference for oneself, and the chapter on inference for others, in that order. This brings the Commentary more in line with the organization of the Compendium (even though one of Prajñākaragupta’s commentators, Jayanta, denied that it was intended as a commentary on the Compendium). The Tibetan translation of the Commentary has the original sequence; that is to say, it begins with the chapter on inference for oneself. 4 Hattori (1968, 23, slightly emended): Saluting Him, who is tantamount to a means of cognition, who seeks the benefit of all living beings, who is the teacher, fully accomplished, the protector, I shall for the purpose of establishing the means of knowledge compose the Compendium, by uniting here my theories scattered in many treatises. 5 Alternatively, one may infer the Buddha’s knowledge from the truth of the main part of his teaching, the Four Noble Truths, independently established. From that, one may infer his compassion. From his knowledge and compassion taken together, his authority and trustworthiness are inferred, on the basis of which one may be confident in the truth of the non-essential parts of the scriptures authored by him (Franco 1997, 68). 6 However, sthiti is also interpreted simply as “cognition.” 7 Dharmakīrti clarifies this point with PV 2.6. 8 See also PV 3.145–73. 9 See Kellner (2017, 112–13) for a much more nuanced interpretation. 10 For instance, “being produced” is identical with “being impermanent,” and hence can serve as evidence for the latter, because both are essential properties of a thing that arise from the same causal complex. The property-to-be-proved, in an inference based on identity, “follows upon the mere existence of the reason,” without need of any further causal factors. 11 A cause is a sufficient condition of the effect, given that all auxiliary factors are present and their operation is unobstructed. Since, however, the auxiliary factors are not always in place and obstructions can always occur, one cannot reliably infer an effect from its cause, but only the cause from the effect, as its necessary condition.

References Dunne, John. 2004. Foundations of Dharmakīrti’s Philosophy. Boston: Wisdom Publications. Eltschinger, Vincent. 2007. Penser l’autorité des Écritures. Vienna: Verlag der Österreichischen Akademie der Wissenschaften.

320

Dharmakīrti ———. 2010. “Dharmakīrti.” Revue Internationale de Philosophie 64 (3): 397–440. ———. 2014. Buddhist Epistemology as Apologetics. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Eltschinger, Vincent, Helmut Krasser, and John Taber. 2012. Can the Veda Speak? Dharmakīrti Against Mīmāṃsā Exegetics and Vedic Authority. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Eltschinger, Vincent, and Isabelle Ratié. 2013. Self, No-Self, and Salvation: Dharmakīrti’s Critique of the Notions of Self and Person. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Eltschinger, Vincent, John Taber, Michael Torsten Much, and Isabelle Ratié. 2018. Dharmakīrti’s Theory of Exclusion (apoha), Part I: On Concealing. Studia Philologica Buddhica 36. Tokyo: International Institute for Buddhist Studies. Franco, Eli. 1997. Dharmakīrti on Compassion and Rebirth. Wiener Studien zur Tibetologie und Buddhismuskunde 38. Vienna: Universität Wien. Franco, Eli, and Miyako Notake. 2014. Dharmakīrti on the Duality of the Object. Leipziger Studien zu Kultur und Geschichte Süd- und Zentralasiens 5. Leipzig: LIT Verlag. Frauwallner, Erich. 1954. “Die Reihenfolge und Entstehung der Werke Dharmakīrti’s.” In Asiatica. Festschrift Friedrich Weller, edited by Johannes Schubert and Ulrich Schneider, 142–54. Leipzig: Otto Harrassowitz. Hattori, Masaaki. 1968. Dignāga, On Perception. Harvard Oriental Series 47. Cambridge, MA: Harvard University Press. Kellner, Birgit. 2017. “Proofs of Idealism in Buddhist Epistemology.” In Indian Epistemology and Metaphysics, edited by Joerg Tuske, 103–28. London: Bloomsbury. Krasser, Helmut. 2001. “On Dharmakīrti’s Understanding of pramāṇabhūta and His Definition of pramāṇa.” Wiener Zeitschrift für die Kunde Südasiens 45: 173–99. ———. 2012. “Bhāviveka, Dharmakīrti and Kumārila.” In Devadattīyam. Johannes Bronkhorst Felicitation Volume, edited by François Voegeli et al., 535–94. Bern: Peter Lang. Pecchia, Cristina. 2015. Dharmakīrti on the Cessation of Suffering. Brill’s Indological Library 47. ­Leiden: Brill. PV. Dharmakīrti’s Pramāṇavārttika with Commentary by Manorathanandin, edited by Rāhula Sāṅkṛtyāyana. Journal of the Bhandarkar Oriental Research Institute 24–26 (1938–1940), Appendix. PVin. 2007. Dharmakīrti’s Pramāṇaviniścaya. Chapters 1 and 2. Edited by Ernst Steinkellner. Vienna: Austrian Academy of Sciences Press. PVSV. 1960. The Pramāṇavārttikam of Dharmakīrti. The First Chapter with the Autocommentary. Edited by Raniero Gnoli. Rome: Istituto Italiano per il Medio ed Estremo Oriente. PV(SV). PV together with PVSV. Steinkellner, Ernst. 1968/1969. “Die Entwicklung des Kṣaṇikatvānumānam bei Dharmakīrti.” Wiener Zeitschrift für die Kunde Südasiens 12–13: 361–77. Taber, John. 2020. “Philosophical Reflections on the Sahopalambhaniyama Argument.” In Reverberations of Dharmakīrti’s Philosophy. Proceedings of the Fifth International Dharmakīrti Conference, edited by Birgit Kellner et al., 441–62. Vienna: Austrian Academy of Sciences Press. Vetter, Tilmann. 1984. Der Buddha und Seine Lehre in Dharmakīrtis Pramāṇavārttika. Wiener Studien zur Tibetologie und Buddhismuskunde 12. Vienna: Universität Wien.

321

PART 5

Early-Period Commentators (Fifth–Seventh Century)

EARLY-PERIOD COMMENTATORS Introduction to Part 5

Indian Buddhist philosophers regarded as the founders of philosophical schools – such as the figures in the preceding part of this volume – often wrote texts that are terse and sometimes so obscure as to be incomprehensible without a commentary. Because the root texts were composed with the expectation of oral or written commentary to explain them, commentaries were not simply a supplement to an otherwise sufficient work; they were essential to the root text. This is why some Indian Buddhist philosophers wrote auto-commentaries on their own texts. In many instances, it would be reasonable to consider the root text and its commentary as together constituting a kind of hybrid text, and this is one way in which authorship is rendered complicated in Indian Buddhism. Yet, it was through these hybrid texts that different schools were constructed as distinct and coherent. It was through the commentarial literature that these schools developed and became more sophisticated as commentators drew on the resources of their traditions and the indeterminate meaning of the texts they were explaining to introduce novel interpretations and ideas in response to new philosophical challenges. By the sixth century, developments in logic and epistemology contributed to the rise of debates across Indian philosophical traditions. Commentators often understood their role to include not just defending their own view but also demonstrating the deficiency of other Buddhist and non-Buddhist philosophical positions. As with other South Asian traditions, then, even as many Indian Buddhist philosophers were creative and original thinkers, they often presented their thought as explication and practiced philosophy though commentary on earlier works. Buddhaghosa’s commentaries on Buddhist scriptures and his formulation of a Buddhist life in the Path of Purification (Visuddhimagga) played a significant role in the formation of what today is called the Theravāda tradition. At the heart of Chakravarthi Ram-Prasad’s chapter is what he terms Buddhaghosa’s “hermeneutic of contextualization,” which is especially attentive to the context in which the Buddha offers his teaching. Applying the hermeneutic of contextualization to Buddhaghosa’s own texts, Ram-Prasad argues, suggests that he is not ­primarily concerned with the ontological analysis of objects of experience. Rather, Buddha­ ghosa practices a phenomenological method aimed at transforming human experience. Drawing on Buddhaghosa’s hermeneutics and phenomenological method, Ram-Prasad shows how his account of understanding in the Path of Purification is a form of phenomenological practice that prepares the practitioner for systematic, methodical, and morally informed meditation.

DOI: 10.4324/9781351030908-29

325

Early-Period Commentators (Fifth – Seventh Century)

In the next chapter, Karl-Stéphan Bouthillette shows how Bhāviveka (sixth century) drew on recent developments in epistemology and argumentation to interpret Nāgārjuna’s philosophy of the Middle Way. Bhāviveka’s interpretation – using the three-part inference (anumāna) developed by Dignāga and the epistemological framework of valid means of cognition or knowing ( pramāṇa) – became the dominant view of Nāgārjuna in Indian Buddhist scholarship. Interpreting and defending Nāgārjuna’s philosophy with the fifth- and sixth-century methods of critical reasoning and debate firmly established it as a system that could compete with other Indian philosophical schools. Indeed, Bhāviveka devoted considerable resources to critiquing the views of other Indian philosophical traditions, setting up a doxographical hierarchy that became a model of much Buddhist philosophy. In this hierarchy, philosophy is paramount and wrong ideas need to be refuted because they are obstacles to liberation. Bhāviveka’s work, then, emphasizes the importance of rational argument, and he himself was an original and rigorous thinker who was able to synthesize different philosophical approaches into a coherent Madhyamaka system. Like Bhāviveka, Dharmapāla (sixth century) employed Dignāga’s logic and epistemology to interpret and defend views first articulated before the fifth- and sixth-century developments in argumentation and reasoning and to critique other Buddhist and non-Buddhist schools. In his chapter on Dharmapāla, Ching Keng suggests that one of his opponents was in fact Bhāviveka and that the debate between the two thinkers is the beginning of the debates between Madhyamaka and Yogācāra views that became important for many later Buddhist philosophers. While Dharmapāla’s commentaries are conservative, hewing closely to the root texts, works attributed to him also include original thinking on a variety of important Yogācāra themes, especially meta-cognition. Keng suggests that perhaps this tension can be resolved by thinking of “Dharmapāla” as including the Indian scholar and his students and followers and that it was their combined work that had such a significant influence on Buddhist philosophy in India and especially in East Asia. Authorship, in this context, works more like the label on a “workshop” rather than a designation of a single individual. As Roy Tzohar and Jowita Kramer show in the next chapter, Sthiramati (ca. sixth century) used his commentaries on Yogācāra texts to make novel philosophical contributions. In the first half of the chapter, Kramer provides an overview of current research on Sthiramati, including the philosophical themes he explored in his work. In the second half, Tzohar gives a close reading of Sthiramati’s commentary on the beginning of Vasubandhu’s Treatise in Thirty Verses (Triṃśikā), emphasizing Sthiramati’s innovative understanding of meaning, metaphor, and language. Sthiramati’s linguistic figurative theory of meaning complements the Yogācāra account of the causal activity of consciousness resulting in a more systematic philosophy. Indeed, it was through his commentaries that Sthiramati synthesized a variety of texts into one coherent Yogācāra system. Alexander Yiannopoulos, in the following chapter, focuses on Devendrabuddhi (seventh century) and Śākyabuddhi (seventh century), the two earliest commentators on Dharmakīrti’s Commentary on the Means of Knowledge (Pramāṇavārttika). Yiannopoulos reflects on how these commentators interpret Dharmakīrti and explores particular distinctions and themes that exemplify their approach to the root text, such as Devendrabuddhi’s account of “unmediated instrumental effect” and Śākyabuddhi’s analysis of reflexive awareness. Because Dharmakīrti’s Commentary on the Means of Knowledge is so terse, it requires additional commentary to make the arguments explicit. Indeed, Dharmakīrti’s text is not just philosophically obscure; the commentaries are often necessary even to make sense of the syntax and grammar. Given how thoroughly interdependent this particular root text and its earliest commentaries are, Yiannopoulos suggests that we should consider them as together constituting one “hybrid text.”

326

Introduction to Part 5

Candrakīrti (seventh century) is widely regarded as one of the most significant Madhyamaka philosophers. This is especially true in Tibet, where his commentaries on Nāgārjuna’s texts are often taken to be authoritative on Madhyamaka philosophy. Focusing on Candrakīrti’s understanding of the two truths, the importance of his approach to language, and his overall philosophy as a form of anti-essentialism, Mattia Salvini gives an overview of Candrakīrti’s writings and an analysis of his main ideas. In addition, Salvini emphasizes three aspects of Candrakīrti’s work that are often neglected in English language scholarship. The first is how Candrakīrti’s philosophical practice is grounded in traditions of Sanskrit grammar, which is key to understanding his anti-essentialist philosophy of language. Second, Salvini shows how Candrakīrti is attentive to the ways in which Nāgārjuna himself was working in an abhidharma philosophical context and how Candrakīrti employs abhidharma analysis in his own works. The third aspect is Candrakīrti’s inclusive hermeneutics, as he finds arguments for emptiness in both Mahāyāna and non-Mahāyāna sources and presents his own ideas without relying exclusively on the authority of any one circumscribed set of Buddhist sūtras.

327

19 BUDDHAGHOSA Phenomenology, Hermeneutics, and Understanding Chakravarthi Ram-Prasad

Introduction Buddhaghosa (fifth century) is the most prominent commentator on Theravāda Buddhist scriptures. He is also the author of what is widely regarded as the definitive Theravāda formulation of a Buddhist life, the Path of Purification (Visuddhimagga), which is grounded in his exegesis. Together, the commentaries and the Path of Purification constitute perhaps the most systematic Theravāda presentation of the Buddha’s teachings after the early Buddhist texts, and they played an influential role in the very formation of what today we think of as Theravāda Buddhist traditions. (The term “Theravāda” meant something very different in the fifth century, when Buddhaghosa would have identified with the Mahāvihāra, or Great Monastery, tradition.)1 According to legend, Buddhaghosa was born in India and ordained as a Buddhist monk so he could study Abhidhamma. His prodigious intellectual powers were quickly recognized in the monastery. Eventually, Buddhaghosa was called to Sri Lanka to work on the massive project of editing and translating the ancient Sinhala commentaries on the Pāli Buddhist scriptures into the more widespread language of Pāli. He composed the Path of Purification, it is said, to demonstrate his scholarly skills and prove his worthiness for the project. While he could not have written all the works attributed to him, it would seem that some major commentaries on canonical literature, such as The Expositor (Atthasālinī) and The Dispeller of Delusion (Sammohavinodinī), so closely resemble the ideas and content of what is undeniably his magnum opus, The Path to Purification, that we can talk about them as his works even if they were written under his close supervision.2 In addition to determining the final form of the Pāli Canon, this work also reinforced the importance of Pāli in Theravāda traditions. This chapter draws on a way of thinking about Buddhaghosa that has been argued for by Maria Heim (2014; 2018) and which I have worked on myself (Ram-Prasad 2018, chapter 3), and also jointly with her (Heim and Ram-Prasad 2018). Our basic contention is that Buddhaghosa is primarily concerned with a phenomenological approach that he regards as conducive to transforming experience in contrast to the interpretation of Buddhaghosa as primarily interested in metaphysics and the ontological tabulation of the objects of experience (Heim and Ram-Prasad 2018, 1087). We are at pains to point out that this is not at all to deny that he has doctrinal commitments to how things truly are, as conveyed in the Buddha’s teachings. 328

DOI: 10.4324/9781351030908-30

Buddhaghosa

Doctrinal commitments do not automatically imply metaphysical views, and such is the case with Buddhaghosa as Buddhist. Moreover, we should be prepared to treat Buddhaghosa as a philosopher for reasons other than whether he does or does not deploy a metaphysics and defend an ontology in his texts; we may look at his contributions in ways that, in comparativist perspective, locate him in other areas of philosophy. We will explore these points through a reading of some of his work in this chapter. Drawing on Maria Heim’s work (2018) on Buddhaghosa’s practices of reading text, I will suggest that his interpretive theory gives us a hermeneutics with a quite general significance, one that can be called a hermeneutic of contextualization. Drawing out this hermeneutics will be the dominant task of this chapter. I will, however, then turn to that area of his work that most naturally has been assimilated into a larger history of metaphysics, specifically, atomistic ontology, as is held to be found in the teachings on the abhidhamma.3 I will suggest that, given his hermeneutic of contextualization, if we properly contextualize his study of the supposed ontological constituents of the person, the aggregates (khandhas), then an ontological reading of the aggregates becomes inconsistent with his purposes. Buddhist philosophers share the same beginning and end: the teachings of the Buddha and the cessation of suffering, respectively. But what goes on in between is not always argument about the structures of reality and the pursuit of an epistemic vindication of one’s argument. In Buddhaghosa, we have a figure who is original in many ways within Buddhist thought but should also be taken seriously for his contribution to global philosophy.

Hermeneutics: Buddhaghosa’s Contextualization Strategies Buddhaghosa’s way of reading text matters philosophically in two ways. One is that his methods can enrich hermeneutics as a cross-cultural philosophical discipline so that we are not tempted to see his practices as limited to a distant Buddhist past, of interest only to a history of ideas. The other is that taking his hermeneutics seriously has consequences for how we locate his treatment of abhidhamma categories as something other than an atomistic (and, from the hostile perspective of Madhyamaka opponents, a conveniently simplistic) ontology. If there is one word that sums up Buddhaghosa’s approach to Buddhist texts, it is “contextualization.” As we will see, this is a paradoxical approach, one in which attention to detail does not limit but rather opens up inquiry – a fact routinely missed in treatments of Buddhaghosa as someone bogged down in endless disquisitions on the minutiae of monks’ conduct.4 One specific aspect of Buddhaghosa’s approach to texts makes it particularly apt when thinking of modern hermeneutics as a discipline framed by its response to Christian theology: the “canonical” nature of the texts with which he was dealing. This can sometimes be forgotten in a pan-Buddhist approach to hermeneutics. In a path-breaking work, Robert Thurman makes the point that Buddhist hermeneutics must concern itself with an incredibly vast array of texts and teachings and yet treat them as somehow uniformly authentic (1978, 22). But developing this thought from a Mahāyāna perspective, Thurman almost casually sets aside the “simplistic approach” of the “Hīnayāna Abhidharma.”5 He then focuses, for its vastness, on various linguistic traditions of the Mahāyāna. Referring to Thurman, Lopez points to “the sheer bulk of the canon,” explaining it by saying “the Buddhists seemed to have simply written a new sūtra or interpolated new material into an existing text” (1998, 2). Two different points are elided here. The quantitative issue is one thing, but Lopez expands on it by talking about something else, Mahāyāna Buddhism’s qualitatively different way of treating sacred text by generating new, authoritative ones. But the vastness of the “Pāli canon” does not equate to the method of generating new material; Buddhaghosa is very concerned with defending the anomalous 329

Chakravarthi Ram-Prasad

character of the Abhidhamma, for example. His strategies are comparable in this specific sense with Biblical hermeneutics, in having to deal with types of reading and consequently find meaning in a fixed set of materials.6 It is contextualization of a fixed textual range of teachings that offers depth to his hermeneutic practice, but if we are to suggest what that practice offers to a cross-cultural discourse on hermeneutics, we must start with acknowledging his location as a Buddhist. Yet, as we will see, the theoretical originality he would bring to such a discourse cuts across the divide between theological exegesis and critical interpretation that defines the (Western) history and discipline of hermeneutics.

Framing Buddhist Truth: Discovered Omniscience Let us begin with an outline of Buddhaghosa’s hermeneutics, both its framework and contextualizing practice, following Heim’s detailed study (2018). Obviously, we have the straightforward consideration that he is committed to faith in the Buddha, as any commentator within a religious community would be to sacred texts. This requires acceptance of doctrines as ultimately holding true, as in other religious traditions. But their holding true is due to a striking combination in Buddhism: these doctrines articulate what was always true but came to be taught only because insight into them was attained by the Buddha. This stands in contrast to doctrines being true due to revelation by an always already divine source – a view that was only just gaining some traction in Hindu theisms in the time of Buddhaghosa. The ambiguities in the canonical material about whether the Buddha himself claimed omniscience are decisively settled by the later consensus that he was, although interpreted in very different ways by Mahāyāna systems (Griffiths 1989; Perrett 1989). It is important to note that Buddhaghosa’s hermeneutic is driven by his conception of the Buddha’s omniscience (Heim 2018, 44–49). It is a dynamic conception and draws attention to a continuous quality of being enlightened. He comments on a canonical discussion of the term “Buddha” (Heim 2018, 48): Here, just as it is said in the world that someone has “understood” because of “understanding,” this is the case too for one who is Awakened (“Buddha”), who has discovered the Truths. . . . Awakened because of knowing all things: “awakened” is said because of a wisdom capable of discovering [literally, “attaining knowledge of,” bujjhana] all things. Awakened because of seeing all things: “awakened” is said because of a wisdom capable of knowing [as also teaching, bodhana] all things. Awakened because of being unguided by another: “awakened” is said because of being awakened only by himself, not woken up by another. Awakened because of expanding: “awakened” is said in the sense of expanding like a lotus blossoming with many qualities. (Paṭisambhidāmagga-aṭṭhakathā, ii.485, Pāli Text Society) What we should keep in mind here is that the buddhahood of the Buddha consists in his attaining omniscience and being “capable” of or “able” to make such discoveries and teach them. There are two consequences that are relevant to our study, one regarding hermeneutics, the other pertaining to phenomenological method. First, in as much as the Buddha discovered “the Truths,” a Buddhist hermeneutics makes the same commitment to ultimacy as any Christian hermeneutics. But because the Buddha discovered “the Truths,” his omniscience stands in a different relationship to the Buddhist hermeneuticist (and thus to us) than does the status of truths conveyed by divine revelation. This is important because, in the West, hermeneutics as a discipline grew out of the dialectic between Christian commitment to a particular understanding of ultimate truth 330

Buddhaghosa

and the secular rejection of it in favour of context. Buddhaghosa’s framing of the Buddha’s authority as consisting in his attainment of omniscience encloses a conception of hermeneutics as the constant contextualization of our understanding of such knowledge. He therefore offers us a way of re-working the relationship between truth and context, as we will see shortly. Relatedly, Buddhaghosa’s understanding of the Buddha’s omniscience is dynamic and ongoing: it is the activity of awakening, it is a potency lying in his ever-ready capacity to transmit his teachings, and it expands like a lotus, an ever-unfolding capacity (Heim 2018, 48–49). According to Buddhaghosa, this is demonstrated by, amongst other things, the Buddha’s methods (naya) of teaching how to read experience endlessly. It is Buddhaghosa’s reading of the nature and content of the Buddha’s omniscience and the Buddha’s means of teaching such knowledge that together show how Buddhaghosa’s hermeneutics points to his phenomenological method, free from ontological concerns. This understanding of the teachings is expressed through Buddhaghosa’s elaboration of the idea that the teachings (as contained in the three piṭakas) are an “ocean of methods” (nayasāgara) in The Expositor. Of the Abhidhamma, he says: And endless joy and happiness arise for those monks who are Abhidhammikas [that is, specialists] reflecting on the Abhidhamma scripture thus: “the Teacher taught us by analyzing the name-and-form dhammas, making this or that grouping, this or that division, as though arranging the stars in the sky, which is an abstruse and subtle teaching that has classified into what is form and what is formless the manifold aggregates, bases, elements, faculties, powers, limbs of awakening, kamma and its results.” (Heim 2018, 241) Characteristically, he says this should be understood through an incident in which an Elder, contemplating the ocean on a voyage, asked which had the greater strength, the ocean or the methods, and concluded joyfully that, whereas the ocean was eventually bound by land, there could be no circumscription to the method. The entire classificatory method of the Abhi­ dhamma here is given an organization of various groupings, like discerning constellations as patterns in the sky. This is the practice that has been taught, and its enactment is phenomenological. So we see again the folding in of how the Buddha taught with how we should learn what he taught: the infinitude of methods by which he taught keys us to approach those teachings through these infinite methods. Note, too, that the experiencing of the practice of endless classificatory analysis is joyful, as we see with what the Elder undergoes. So the joy within the teaching becomes the joy of those who learn.

Contextual Teaching as Guide to Contextual Reading In order to understand the significance of Buddhaghosa’s hermeneutic of contextualization, we must first notice how salient a contrast to certain Western conceptions of sacred authority it is to say that the Buddha’s authority came through his discovery of truth. Jussi Backman delineates the nature of classical Western and Christian hermeneutics thus (2016, 29): For the ontotheological ethics of Aristotle, firmly inscribed within the ancient and medieval ethical tradition reaching from Parmenides to Thomas Aquinas, the culmination of human activity and intellectual being is thus the escape of thinking from the situatedness of “merely” human existence to the autarkic beholding of the divine and the absolute. Philosophical hermeneutics, by contrast, is precisely the philosophical embrace of this irreducible mortal situatedness of humanly accessible meaning, and 331

Chakravarthi Ram-Prasad

in this respect it indeed marks an entirely “new philosophical epoch” with regard to antiquity. For Buddhaghosa, the Buddha is indeed “autarkic” (for he is “unguided by another”), and we can think of his omniscience as “absolute” in the specific sense of being unlimited relative to what needs knowing. Yet, at the same time, omniscience consists precisely in the Buddha’s specific orientation to the situatedness of existence. Buddhaghosa’s insight is that because the Buddha taught according to context, we must see the Buddha’s teaching through attending to context. “Situatedness,” in other words, is itself the vector of the “absolute” truths. Let us see an example of how this is so. A key concept to consider here is the nidāna: It is “a context, origin, source, occasion, introduction, or causal condition. When used to name part of a sutta or other text it is the introduction or narrative context in which the text occurred in the Buddha’s teaching career” (Heim 2018, 116). The nidāna is a device that can be used to frame teachings at the canonical level but also be used by the commentators. The first point is important because it establishes the nidāna as integral to the very existence of the body of teachings and not a mere convenience for the commentators. But the second is relevant because it shows how Buddhaghosa was able to utilize the form of the nidāna ingeniously. Heim points out that modern scholarship has thought of nidāna as primarily serving to show the authenticity of the teachings, and Buddhaghosa’s sub-commentator Dhammapāla sees it exactly as performing this solemn task (Heim 2018, 118, 121): For a teaching connected to a time, place, teacher, story, and the recipient of the Dhamma is a long-lasting, unconfused doctrine, and credible, like a legal contract provided with notations of place, date, maker, and witnesses. (Dīgha ṭīkā i.70) Heim contrasts this concise presentation with Buddhaghosa’s approach, which is lavish in describing the purpose of nidāna and which Dhammapāla queries for its “poeticity” (kabyaracanā). But I want here to focus on the specific issue of contextualization and Buddhaghosa’s reading of it hermeneutically. It is not that he deviates from the standard understanding that a nidāna is about context and involves situating a given teaching in a moment and time and in relation to a specific audience (see Heim 2018, 117). Indeed, Buddhaghosa is perfectly consistent with the tradition’s reading of the nidāna as demonstrating that the Buddha taught in a very contextual way. And in this, Buddhism markedly contrasts with the commitment to the ontotheological universalism that marks Biblical teaching.7 Yet this is only one half of the principle of contextualization. Buddhaghosa’s achievement, I think it can be argued, is that he turns the fact that the Buddha taught according to context into a hermeneutic principle. Heim points out the aesthetic role a nidāna plays for Buddhaghosa “in preparing the imagination for entering the sutta perspective to the beauty and pleasurable delight” of faith in the Buddha (Heim 2018, 121). And this is evident in a passage she quotes in extenso. Let us look at the first part of this passage. Thus far, the nidāna, adorned with time, place, teacher, narrative, assembly, and region, is spoken by Venerable Ᾱnanda, and the commentary on its meaning is complete. It is spoken for the sake of the ease or pleasure of entering this sutta, which is perfect in meaning and phrasing and indicates the power of the Buddha’s qualities. (Sv.i.50; Ps.i.15; Pv-a.iii.536) 332

Buddhaghosa

This nidāna, Buddhaghosa says, is like a crossing place (tittha), a staircase (sopāna), and a great door (mahādvāra), and he describes each metaphor in exuberantly, lavishly poetic terms, evoking pearls and blue water lilies, golden creepers and delicately carved ivory, sweet laughter and the tinkling of bracelets. It is clear that he wants to suggest that the nidāna is meant to delight and prepare the reader. But I also think that it is crucial to notice that the nidāna is meant for entering into (avagāhati) the sutta. In other words, it is a method for interpretation. If we are to understand the content of the teaching that is to come, we are to get to it through context. This explains the densely phenomenological approach he takes to the nidāna, in contrast to the precise epistemological reading of its function that Dhammapāla gives (and which, in a different form, is taken to be its aim in modern scholarship). The nidāna is important because the context it provides through time and place and the rest is integral to how we are to receive the teaching we will encounter. It prepares us to read the teachings. It delights but also informs. This information is not only about the nature of the teaching itself (which, of course, he takes for granted) but about the way to learn the teaching. My suggestion is that Buddhaghosa uses the tradition’s invocation of the Buddha’s contextualization of teaching to himself develop a hermeneutic of contextualization. Let me draw this out with an example. Heim offers a detailed reading of Buddhaghosa’s commentary on the Discourse on the Root of All Things (Mūlapariyāya Sutta) of the Middle Length Discourses (Majjhima Nikāya) (Heim 2018, 136–41). The sutta itself has no elaborate nidāna but ends with the enigmatic note that, after the Buddha delivered a sermon, “the monks did not delight” in what he had said. Buddhaghosa goes to great lengths to explain the context: the monks were former brahmins who were conceited due to their knowledge of both Vedic and Buddhist texts. The Buddha’s teaching was to “break down the conceit” (mānabhañjanatthaṃ) of these monks: it shows the subtle ways in which conceptuality (sañña) distorts experience and generates “craving, conceit and wrong views.” Buddhaghosa reads the dismay of the monks as being due to their overgrown learning – which actually reinforced their conceit – impeding their recognition of the truth of the teachings. Strikingly, Buddhaghosa increases the contextualization. When his followers ask why those monks were dismayed, the Buddha says that they were brahmins in a past birth who were similarly stymied by the Buddha in a previous life and that in the future they will become realized after hearing another of his teachings. Heim explores this commentary in terms of the Buddha’s omniscience, as well as to make the point that Buddhaghosa narrates the enactment of the sutta’s teaching so that the “frame story performs the sermon” (Heim 2018, 139). I want to extend the analysis to say that this is a striking demonstration of the double-aspect contextualization which I suggest is at the heart of Buddhaghosa’s hermeneutics. The double aspect is that context is both the intrinsic feature of the events (name, place, occasion, etc.) surrounding the truthful teaching that is now being read, but it is also the technique by which the truth is read. His hermeneutic is a form of narrative reasoning, arguing through the story by which doctrine is conveyed, and, if the story is not in the canon, developing an explanatory version himself. If we disregard this strategy as over-imaginative piety so that we can isolate the “philosophy” of conceptualization, then we find Buddhaghosa both prolix where it does not seem to matter and didactic when he should be making an argument. From this perspective, Buddhaghosa asserts the perpetual truth that conceited conceptualization of selfhood impedes felicitous freedom not through arguing for the epistemic and moral incoherence that follows from the conceit but through the narrative of why and at what point(s) and to whom the Buddha taught. The Buddha’s teaching does not need expatiation; it is our reckoning of its truth that does. This double-aspect contextuality makes Buddhaghosa’s hermeneutic attitude to the text radically different from how modern hermeneutics has understood religious reading. Felix Ó 333

Chakravarthi Ram-Prasad

Murchadha cites Ricoeur’s assumption with regard to the challenge which a critical hermeneutic faces in its relationship to earlier religious hermeneutics: [H]ermeneutics bears the mark of this interpretive effort of understanding enigmatic, apparently inscrutable, divine communications. . . . [I]t has a text‐like quality in the sense of being preserved as meaningful beyond the situation of its utterance. Crucial here is the relation between tradition and experience, mediating between a past event of divine communication expressed in written form and the present experience of those reading and applying that text. (Ó Murchadha 2016, 77, referring to Ricoeur 1995) The significance of the nidāna, by contrast, is that the teachings require exegesis not because they are inscrutable or enigmatic but rather because their meaning is encountered and understood in every context that could occur (for the Buddha is omniscient about the contexts to which his teachings should be directed). Jussi Backman (2016, 26) broadly defines hermeneutics as “discursive contextuality,” which works well to define Buddhaghosa’s approach to canon. But for Backman, this means that philosophical hermeneutics “abandons the prospect of attaining an ultimate, pre- or supracontextual point of reference.” Buddhaghosa cannot exactly subscribe to this understanding. The Buddha’s teachings are indeed an “ultimate” point of reference; they are not “ultimate” in the sense of a higher-order ontology beyond the objects of ordinary experience but in the sense of being that beyond which there is no further source for Buddhist doctrine, since they are rooted in the omniscient Buddha’s authority. But this point of reference is precisely not “precontextual,” as we have just seen. So when it comes to a Buddhaghosan hermeneutic, we must tease apart some of the assumptions which, in contemporary Western theories of hermeneutics, are bundled together. As for “supra-contextual” points of reference, even the abstract teachings of the Buddha are not really supra-contextual. It is that insight that will lead us from the hermeneutics to a non-metaphysical phenomenology of abhidhamma. But we are getting ahead of ourselves. Let me quickly outline some of Buddhaghosa’s contextualizing techniques; these are dealt with in detail in Heim (2018). Broadly, contextual hermeneutics in Buddhaghosa proceeds through offering paired distinctions, the choice between each member of the pair being keyed by the manner in which the Buddha presented the teachings, with that choice then opening up each teaching according to our understanding of that manner. We have space only for a few of these double-aspect contextualization techniques. One pair is “meaning” (attha) and “phrasing” (byañjana), each containing a “toolkit of hermeneutical devices” (Heim 2018, 74). It would take a different enterprise to demonstrate how exactly these tools could contribute to a cross-cultural global hermeneutics. But I draw attention to the potential for precision of analysis offered by the listing under “meaning” of “syllables, words, phrasing, modes, language, descriptions” (akkharaṃ padaṃ byañjanaṃ ākāro nirutti niddeso; in Nettiprakaraṇa 9, which Buddhaghosa draws on), and “showing, making known, opening up, distinguishing, making clear, and denoting” (saṅkāsanā pakāsanā vivaraṇā vibhajanā uttānīkammaṃ paññatti) under “phrasing.” Another pair is “definitive” and “interpretable,” which we encountered earlier. A  consequential pair that plays an important role in seeing that Buddhaghosa utilizes language common to him and Buddhist metaphysicians, but in a different way, is that between “conventional” (sammuti or vohāra) and “further sense” ( paramattha) teachings. The term paramattha here is not “ultimate truth” in the sense of being what holds in a higher-order ontology behind ordinary objects of experience or as a proposition that is unsublatably or ineliminably true in contrast to those 334

Buddhaghosa

that are only provisionally true. Rather, paramattha is “ultimate” in the doctrinal sense of what is of highest value, as when nibbāna is called the highest truth (sacca), or the final authority, as in the Buddha’s insight. Heim’s exegetical demonstration that “paramattha” should be treated as “further sense” (Heim 2018, 85–94) when it comes to the Buddha’s teachings (desanā) is consistent with the latter but does not directly pertain to it. She demonstrates that Buddhaghosa explains at length that the distinction between “conventional” and “further” is about the context of who needs teaching what. There are contexts that require teaching about persons, gods, and so on and others that require language utilizing concepts of suffering, no-self, aggregates, elements, and similar categories. The former are not lesser in relation to a higher but a type of discursive analysis pertaining to what is ordinarily presented, namely persons and the rest, as opposed to analysis that goes further to what requires more sustained types of contemplative inquiry. The equal legitimacy of conventional language – both in terms of spiritual perfection and of exegetical analysis – is manifestly evident in The Wish Fulfiller (Manorathapūraṇi), the commentary (aṭṭakathā) of the Numerical Discourses (Aṅguttaranikāya) (Heim 2018, 86): If someone, having heard a conventional teaching, is able to attain distinction, having abandoned delusion and comprehended the meaning, then the Lord teaches a conventional teaching. (i.94) And the same is said of the further teaching. This critical fact should enable us to see that the Buddhaghosan distinction pertains to hermeneutic context and not to metaphysical levels. Heim drives this home through analyzing several passages, a particularly extraordinary one being a comparison of the contextual application of one or the other discursive strategy by the Buddha with pedagogical code-switching: The Buddha is like a skilful teacher of the Vedas who gives commentary on them according to the language of the students, be it the Tamil language or the Andhra language (Heim 2018, 88). If the Buddha taught according to context, then our reading of his teaching should likewise make the contextual distinction. Again, how the Buddha taught and how we read his teachings are folded together; hermeneutic context follows pedagogical context. I believe that this drawing out of techniques from what a text itself enacts is a critical insight that can be applied outside the Buddhist context: the text yields methods even as it performs its tasks.

The Continuous Horizon Now, the idea at the very heart of modern hermeneutics is the temporal connection between textual context and the reader’s context that Gadamer famously called “the fusion of horizons.” As has been pointed out, Gadamer’s hermeneutics is deeply involved with the texts of the past, since his programme assumes that the ancients have something to teach us (Weinsheimer 1985, 133). He constantly warns against the spurious separation from sacred text by which a supposed “objective knowledge” is gained (Weinsheimer 1985, 183). Scripture is not meant merely for its own time and place, but it should also apply to situations in time to come since it is revelation. We can straightforwardly locate Buddhaghosa within this approach. But then Gadamer says that the consequence of the sacred text’s applicability for the future is that the circumstances in which its truths may be applied are unforeseen and unknowable (2004, 309). The text must continue to apply, but in these circumstances, we need a fusion of horizons: “the horizon is the range of vision that includes everything that can be seen from a particular vantage point” (Gadamer 2004, 302). The text’s continued applicability beyond the original temporal boundaries 335

Chakravarthi Ram-Prasad

comes through its having an alterity in relation to the reader’s current perspective (“prejudices,” to use the Gadamerian term) but yet, at the same time, becoming part of that current perspective and thereby altering it. Understanding therefore consists in the fusion of horizons, in hermeneutically bridging the gaps between then and now. However, Gadamer rejects the notion that in order to “understand” another historical period, one needs to enter into it, to transpose oneself, as it were, to become one with the mindset of another time and place. Rather, fusion implies a common understanding at a level of a “higher universality.” True (traditionary) understanding in Gadamer’s view consists in the bridging or the “fusing” of horizons; in particular, the present horizon, which one maintains, and the past horizon, out of which the present horizon develops (Knotts 2014, 239). As Gadamer clearly indicates, this is particularly and powerfully the case when the text stands in the relationship of (in)forming the tradition of the reader (2004, 306). I want to locate the Buddhaghosan hermeneutic here in conversation with Gadamer. While wrestling with the same issue of the temporal distance and the transtemporal applicability of sacred text, Buddhaghosa’s framework of contextualization offers a different mode of treating the horizon of meaning. For him, contextualization is not the objectification of the past that Gadamer rightly criticizes. Rather, it is the very means through which the text is present, as also at the same time the means by which the reader makes it present. How, then, does Buddhaghosan contextualization change the terms of the hermeneutic relationship between text and commentator or between text and reader when they are separated by time (including, in the case of the last, the cultural rupture of modernity)? If we can discern the outlines of an answer to that, then we can have an idea of how such a hermeneutic can make a philosophical contribution to hermeneutics. Contextualizing to situations was an integral part of how the Buddha taught in the suttas. So, it seems, he anticipated the experience of Buddhaghosa’s reading. While, in turn, Buddhaghosa did not anticipate our experience and context as such (for he claims no universality for himself), his method does anticipate our experience in the sense of always calling for contextualization. We get an idea of Buddhaghosa’s understanding of the temporal relationship to the truthful content of the Buddha’s insightful teaching – the Dhamma – in his section on the contemplative practice called the recollection (anussati) of the Dhamma. The Dhamma, it is taught by the Buddha, is both “visible here and now” (sandiṭṭhika), and it is “timeless” (akālika) (VII.76–80 of Ñāṇamoli; Heim 2018, 226). He explains that the Dhamma is “visible here and now” in the sense that the qualified person can see it for himself when free of lust and of injurious intent, and free of mental suffering and unhappiness. Furthermore, the nine-fold world-transcending Dhamma is visible here and now for, in whatever manner it is attained, it is to be regarded as that reflective understanding that is to be approached through benefiting from faith in another. (VII.77) For the world-transcending Dhamma is the seeing that leads away from the fear of the cycle of births, through the comprehension from cultivation and the comprehension from realization. Just as the one who is clothable is worthy of being clothed, so too the “visible here and now” is what is worthy of being seen here and now. (VII.79) There is no time between it and the bearing of its fruit, so it is “timeless.” The “timeless” takes no time. It does not give of its fruit after a gap of, say, five days or seven

336

Buddhaghosa

days, for it means that it bears fruit immediately after its own occurrence. On the other hand, what takes time is that which needs some distant time before it bears fruit. What is that? The good Dhamma of this world. (VII.80) While the Dhamma holds beyond the world (lokuttara, “world-transcending”) and its understanding is dependent on faith (saddha), it is also the reflective understanding ( paccavek­ khaṇañāṇa), attained through attention or consideration, of the recipient of its textual form. In the final section of this chapter, we will look more carefully at Buddhaghosa’s use of the term “understanding” ( paññā) and its implications for our interpretation of his philosophical approach. But for now, it is sufficient to point out that this understanding is not primarily about epistemic grasp of propositional content; as we saw earlier, reflection ( paccavekkhaṇa) is said to bring the Abhidhamma monks endless joy and happiness (ananta pītisomanassa). Now (as it were), VII.79 has it that “here and now” is a quality of transtemporal worth, not a temporal index. What the Buddha said was here and now at the time of his teaching (the Dhamma) is the here and now of Buddhaghosa’s commentarial attention and the here and now of our current reader. This idea, I think, is not so strictly limited in concept (even if it is in source) to the teaching of Dhamma. Rather, we may think of a sacred text’s ultimacy as consisting not in its temporal horizon needing to fuse with ours but in its constantly being what is present here and now. But we must note that what is transtemporal is not atemporal, for if it were, we would lose traction on the very discipline of hermeneutics, for we are creatures of time such that a text beyond time is uninterpretable. Buddhaghosa’s reading of ‘timeless” (akāla) at VII.80 shows this. If we translated akāla literally as “without-time,” we would indeed be in that quandary. But his reading is that the Dhamma is akāla in the sense that the attainment of meaning is an instant occurrence. Adroitly, he points to “timeless” being the obverse of “here and now”; obverse not in the sense of “opposite” but as in the other side of the same coin. The here and now of readability is also the instant of realizability. He distinguishes this temporal punctuation from the longer drawn-out span of the consequences in this world of having understanding of the Dhamma. This, again, is of quite general applicability. If the continuity of horizon were an explanatory metaphorical alternative to the fusion of horizons, we may ask how we are to account for the understanding that occurs at the point of fusion in the Gadamerian paradigm. The Buddhaghosan account points to an answer – the understanding is the temporal point of attaining reflective knowledge. It does the work that fusion does in the Gadamerian account.

Hermeneutics and/to Phenomenology I have argued for seeing Buddhaghosa as a hermeneuticist, primarily to bring out his crosscultural significance for this contemporary discipline of “philosophy” rather than metaphysics and epistemology, which have for decades been seen as appropriate areas within which to bring Buddhist thought into conversation with the West. Furthermore, this sketch enables me to shift to a different angle of inquiry, to offer the interpretation that Heim and I have defended, that Buddhaghosa has a phenomenology that is methodic all the way through and not directed towards the identification of ontological irreducibles. What is this angle? It is that we must use his contextualizing hermeneutics to approach the non-ontological nature of his phenomenology. Here again, a Gadamerian formulation offers striking comparison when it comes to the

337

Chakravarthi Ram-Prasad

relationship between the contextualized reading of an abhidhamma text and the contemplative practice of discerning its teachings in Buddhaghosa. Gadamer himself, in the foreword to the second edition of Truth and Method, explains that the work is “phenomenological in its method”. . . . By referring to his “method,” Gadamer means, not a set of procedural rules, but rather the discipline of attending to things. By calling his method “phenomenological,” he does not thereby subscribe to Husserl’s account of the phenomenological method but rather indicates that the task of the enterprise is descriptive – in this case, descriptive of the human experience of understanding, i.e., of hermeneutical experience. (Dostal 2002, 251) As I  have indicated, Maria Heim and I  have argued at length that the phenomenology of Buddhaghosa is a descriptive methodology: rigorously structured (“Not any list goes”; Heim 2018, 153) but with potentially endless, ramifying practices of attentively analyzing experience so that their appropriate classification under the different valences as given in the canonical material – the good (kusala), the bad (akusala), and the indeterminate (abyākata) – purifies the life of the practitioner on the path to perfection (see especially Heim 2018, 163–80). The determination of the various dhamma classifications is a continuous practice of phenomenological analysis that does not call for a reading of the dhammas as elements of an ontology. Here, rather than reiterating these findings, I want to ask: what does being sensitive to the hermeneutics of Buddhaghosa demonstrate about the incoherence of then going on to treat abhidhamma categories ontologically? I suggest that this may thereby demonstrate in a different way the nature of his phenomenological method. The conclusion is that if we recognize Buddhaghosa as an hermeneuticist of contextuality, we will have a better understanding of the way he treats phenomenological analysis as an endless exploration and classification of experience leading to perfection. Let me show this through a key passage in The Path of Purification, at the start of the lengthy exposition on the aggregates (Chapter XIV in Ñāṇamoli), which is the first chapter of the third module of The Path to Purification, namely understanding ( paññā).8 Buddhaghosa has already taken the monastic reader through virtue (sīla) and contemplation (samādhi), although the three are always continuously interrelated in the life of the practitioner along the path. Buddhaghosa’s application of understanding will start with the aggregates and proceed to various Buddhist doctrines, such as suffering, dependent origination, and name-and-form, all of which follow the programmatic framing of the concept of understanding that we are considering. The aggregates – form (rūpa), cognition (viññāṇa), feeling (vedanā), conception (saññā), and syntheses (saṅkhāra) – are the five focal points of the analysis of the human being, meant to demonstrate that they together generate the conception of a unitary self (attā) where there is none. I have previously argued that Buddhaghosa’s treatment of the aggregates there, especially and crucially concerning the first aggregate, “form” (rūpa), is a methodical analysis and clarification of phenomenological features and not the determination of irreducible ontological entities (Ram-Prasad 2018, 125–39). The focus is on breaking down the sense of self into constituent phenomenological dimensions; it is not an argument for an atomistic ontology to replace the metaphysics of selfhood. Now, I want to look more closely at Buddhaghosa’s framing of his description of the aggregates by inquiring into the kinds of understanding ( paññā) that must be clarified before one embarks on the study of the aggregates. To lay my cards on the table, if we look at the context of his analysis of the aggregates, it is clear that what Buddhaghosa undertakes is not an epistemological enterprise by which to 338

Buddhaghosa

determine the means of knowledge that are required to elucidate the irreducible features of the world. Rather, it is a project of purifying the phenomenology of the practitioner by understanding the valence of experiential features (as taught by the Buddha) in order to transform them. By attending to how Buddhaghosa frames the reading of that core abhidhamma category, the aggregates, we can see the purpose of that reading. In short, a hermeneutic sensitivity to context, which I hope to have shown is the defining feature of Buddhaghosa’s philosophical approach, brings out the undertaking as one about phenomenological method and not ontological determination. If we take hermeneutics as one result of the recognition in high modernity that specificity (primarily historical, secondarily cultural) informs the interpretation of what is given as knowledge, then we are not surprised that Jürgen Habermas thought of interpretive understanding as “postmetaphysical thinking” (1992, 34) in the context of a condition under which religion (by which of course he means European Christianity) has lost its tethering as an authoritative account of reality but which yet continues to offer spiritual sustenance. After metaphysics, the nonobjective whole of a concrete lifeworld, which is now present only as horizon and background, evades the grasp of theoretical objectification . . . what has, following the disintegration of metaphysical and religious worldviews, been divided up on the level of cultural systems under various aspects of validity, can now be put together – and also put right – only in the experiential context of lifeworld practices. (Habermas 1992, 50–51) Of course, from Habermas’s perspective, this is because metaphysics as the understanding of reality as a whole is required for the authority of religion, an authority that cannot be granted through any expert epistemology in modernity. But I am arguing that we should not saddle Buddhaghosa’s doctrinal commitments with metaphysics, and therefore I am in agreement with Habermas – albeit in a way he would not have anticipated – when he says that, Viewed from without, religion, which has largely been deprived of its worldview functions, is still indispensable in ordinary life for normalizing intercourse with the extraordinary.  .  .  . [R]eligious language is the bearer of a semantic content that is inspiring and even indispensable. (Habermas 1992, 51) In effect, what we may claim for a Buddhaghosan hermeneutic is that it bypasses the historical need to rediscover experiential meaning without metaphysics, for it already functions as a purely phenomenological method for the exploration of meaning. Clearly, the relationship between metaphysics and hermeneutics in Buddhaghosa is quite different than the historical shift that Habermas delineates in the West. Nevertheless, the argument for postmetaphysical thinking can be read as being about an ametaphysical thinking, such as we find in Buddhaghosa. By this I mean that Buddhaghosa’s rigorous adherence to the doctrines taught by the Buddha structures the way he frames various practices of attention, whether textual or contemplative (or, in fact, both simultaneously). I propose to explore how he talks about understanding before he proceeds to his analysis of the aggregates (and thereafter, many other key doctrinal features of the Buddha’s teachings). The location of this description, before the analysis begins, is clearly meant to prime the reader/practitioner in preparation for the analysis of key doctrines. Given the scrupulous contextualization that marks his approach 339

Chakravarthi Ram-Prasad

to the study of doctrines, what he has to say about understanding will teach us what he means to do with his subsequent analysis. What is understanding? Understanding is of many sorts and has many aspects. An answer that attempted to explain it all would accomplish neither its intention nor its purpose, and would, besides, lead to distraction; so we shall confine ourselves to the type intended here, which is understanding consisting in the knowledge from introspective meditation associated with good awareness. (XIV. 2) Is this perhaps an oblique reference to – and dismissal of – the flourishing epistemological traditions in Sanskrit that he is said to have left behind in India? In any case, this is an emphatic signalling of a tightly focused and pragmatic programme of explication that will take the transformation of phenomenology and the cultivation of valence as the objective of understanding. Only if we ignore this explicit orientation would we begin looking for the epistemological pursuit of the constituents of reality. In what sense is it understanding? It is understanding (  paññā) in the sense of being an act of understanding ( pajānana). What is this act of understanding? It is cognizance (  jānana) in a pre-eminent mode, distinct from conceptualizing (sañjānana) and cognizing (vijānana). For although conception and cognition and understanding are equally of the nature of cognizance ( jānana-bhāva), nevertheless conception is only the mere conceiving of a phenomenal object (ārammaṇa) as, say, “blue or yellow”; it cannot bring about the penetration into its characteristics as “impermanent, painful, and not-self.” Cognition has cognizance of objects as “blue or yellow,” and attains penetration into its characteristics, but it cannot bring about, by ascetic practice, the appearance of the path. Understanding has cognizance of objects in the way already stated, it brings about the penetration of the characteristics, and it also brings about, by ascetic practice, the manifestation of the path. (XIV.3) Understanding is a phenomenologically dynamic treading of the path. Notably, too, Buddha­ ghosa is not in the least dismissive of ordinary epistemic functions: conception and cognition determine our apprehension of the objects of experience. But he moves swiftly on from that to understanding as an ascetic teleology, the disciplined mental activity of taking what is experienced and directing it towards Buddhist perfection. However, it is not always to be found where conception and cognition are. But when it is, it is not separate from those states. And because it is not possible to separate it thus: “This is conception, this is cognition, this is understanding,” its distinction is consequently subtle and difficult to see. Hence the venerable Nāgasena said: “A difficult thing, O King, has been done by the Blessed One.” – “What, venerable Nāgasena, is the difficult thing that has been done by the Blessed One?” “The difficult thing, O King, done by the Blessed One was the defining of the formless states of awareness and its concomitants, which occur with a single object, and which he declared thus: ‘This is contact, this is feeling, this is perception, this is intention, this is awareness.’ ” (XIV. 6) (Milindapañha 2.7.16) 340

Buddhaghosa

This is a place where we see why it is best to think of “ārammaṇa” not merely as “object” but as “phenomenal object,” that which occurs in experience. What is relevant here are the phenomenal features, that is, the subjective states that constitute the experience of an object. For what the Buddha did was take a phenomenal occurrence and tease apart its subtle analytic components: the Teacher as First Phenomenologist. We must pay attention to what Buddhaghosa presents as the types of understanding, because it is an elaborate and ramifying list, and what is  – and what is not  – in it should tell us something about how to work with his interpretive exposition (niddesa) of that core abhidhamma category, the aggregates. He lists twelve types, one that is unitary, five that are two-fold, four that are three-fold, and two that are four-fold. We cannot go through them all (one needs to read the whole of The Path of Purification to know what Buddhaghosa is doing, such is his rhizomatic presentation of that path). But let us look at a few of them and draw out their implications for the phenomenological method. Some are self-evidently about spiritual stages, ways of being on the path: one dyad is that which is subject to “cankers” (sāsava) and free from it, the cankers being the persistence of ideas deleterious to progress (XIV.3). Another is understanding of the realm (bhūmi) of seeing (dassana) and development (bhāvana) (XIV.13). We may well begin with the expectation that what is said about understanding (or indeed, as happens midway through, about “knowledge” [ñāṇa]) is going to be epistemological, but it turns out to be descriptions of analytic reflexivity, a fine-grained and methodical paying of attention to phenomenal states. The reflexivity is not only cognitive; it is also affective: the fourth dyad has to do with understanding in the mode of good awareness (kusalacitta) involving meditative attainments (the jhānas) “accompanied by joy” (somanassasahaghata) and likewise good awareness “accompanied by equanimity” (upekkhāsahagata) (XIV.5). An important classification of understanding that drives home the phenomenological method has to do with understanding attained purely by thought (cintāmayā).9 Here, perhaps, in the middle of this elaborate location of understanding as an integrated way of being, is an epistemological kernel that might carry some of the flavour of the famous Sanskritic pramāṇa system of the means of valid, true cognition. He quotes the Analysis (ibhaṅga 768). Then it is said, what is this understanding only by thought? In organized application in the sphere of action, in organized application in the sphere of the crafts, in the organized application of the eight wisdoms, adaptation, forbearance, perspective, inclination, opinion, wish, pondering intently on things, with regard to the ownership of action, or which is in conformity with the truth, or is such a type as conforms with the teachings, “form is impermanent, or feeling . . . conception . . . syntheses, or cognition is impermanent,” which one attains without hearing it from others – that is called understanding only by thought. (XIV.14) Buddhaghosa goes on to say that understanding attained through others is understanding by only hearing, and finally, understanding purely through development (bhāvanāmayā) is all the understanding of one who has attained what is to be attained (XIV.14). Pure reasoning for Buddhaghosa is pragmatic, dynamic, reflexive on the teachings. There is no structured statement on how to know objects, let alone irreducible particulars such as the factors (dhammas), as is supposed to be the case in the metaphysical reading of abhidhamma. Buddhaghosa also gives a triadic definition of understanding that concerns skill in improvement (āya), detriment (apāya), and means (upāya). His description of the understanding that 341

Chakravarthi Ram-Prasad

consists of improvement – which is “twofold, the elimination of the harmful and the arousing of the good” (duvidhā anatthahānito atthuppattito ca) (XIV. 16) – clearly shows that his analysis is going to be oriented to moral cultivation, the transformation of the valence of experience. He quotes the Analysis. Here, what is skill in improvement? When one brings factors to mind, and bad factors that have not yet arisen do not occur, and the bad factors that have arisen are abandoned by him; or when he brings factors to mind, and good factors that have not arisen do occur. And good factors that have arisen increase, become abundant, are perfected in him. This is what is understanding, the act of understanding. (XIV.16; Analysis, 771) This is an important consideration when coming to grips with the hermeneutic of Buddhaghosa’s analyses of various core Buddhist doctrinal principles such as the aggregates: He treats understanding as being about valence. If the salience of the understanding of dhammas lies in the disciplining of thought so as to abandon bad ones and cultivate the good, it becomes incoherent to think that the dhammas are, for Buddhaghosa, the entitative foundation of reality. There are two reasons for this. One is that, if they are the building blocks of reality, then it would make no sense to think they can be stopped from occurring, whereas it makes perfect sense that they can be disciplined into not occurring if we thought of them as experiential states whose arising can be controlled through meditative skill. The other is that, to put it in the classic terms of modern Western philosophy, an ontological reading of Buddhaghosa here would mysteriously treat what “is” in terms of what “ought” to be the case. In other words, there would be an abyss between what Buddhaghosa says understanding of the dhammas is for and what the dhammas are held to be on the ontological view. And yet again, to see them as the factors that structure phenomenology is immediately to make sense of their role as valences to be abandoned or cultivated on the path to perfection. In the fourth triad, understanding through introspective meditation that is initiated by apprehending one’s own aggregates is the interpretation of the subjective. That which is initiated by apprehension of another’s aggregates outside or forms not bound up with sense-faculties is the interpretation of the external. (XIV.19) Buddhaghosa is not interested in a metaphysical distinction between subject and object; he simply takes for granted the features of experience by which the subjective and the external are presented and builds contemplative strategies upon them. This is manifest in his requirement that understanding here is through vipassana or introspective meditation on what he already takes for granted is presented in certain ways. Once more, if we are looking for someone hunting for an ontology, it makes no sense that the subject-object distinction is made without any further comment on their metaphysical status. Furthermore, the means of attending to the aggregates is meditative rather than epistemological, for no mode of validating the determination of ontological constituents is given, whereas the rigorous phenomenological method of vipassana meditation is given as the mode of interpreting how experience occurs. This is, in its own striking way, a hermeneutic phenomenology. Proceeding with his discursive contextualization of the analyses to come, Buddhaghosa also lists two tetradic types of what he now shifts to calling “knowledge” (ñāṇa): these turn out to be entirely about suffering (dukkha), its origin, its cessation, and the way leading to it. 342

Buddhaghosa

He talks of the fourfold modes of discrimination ( paṭisambhida): discrimination of meaning (attha), the ordered teachings (Dhamma), linguistic analysis (nirutti), and utter clarity ( paṭibhāna) (XIV. 21), and goes on to state that the Abhidhamma analysis maps each of the Noble Truths onto the four discriminations. (XIV.24). In all this, there are five aspects ( pañcahākārā) in common with the understanding of the one who is still learning and the one who has already learnt perfectly (sekha and asekha): Here, what is called attainment is the reaching of arahantship. What is called proficiency is proficiency in the word of the Buddha. What is called hearing is hearing the Dhamma carefully and finding its value. What is called questioning is the investigation of knotty and explanatory passages in the canonical text, commentaries, etc. What is called prior discipline is application to the doctrines of previous Buddhas right up to the stages of conformity and change-of-lineage, by one who has cultivated going and coming back with the meditation subject. (XIV.28) By now we should be well attuned to seeing Buddhaghosa’s exploration of understanding prior to analysis of categories such as the aggregates as the hermeneutic priming of the practitioner for disciplined, systematic, and morally configured meditation. Understanding is the organic term for the entire practice of this phenomenological method. We must pay attention to the centrality of his hermeneutic contextualization of all the teaching of the Buddha, as the examination of his techniques for reading show. And once we pay such attention to the context within which he locates the analysis of dhammas, we notice that he has given us no scope to derive an epistemological system through which to determine the dhammas as ontological entities. It would be perfectly possible to read Buddhaghosa’s techniques in isolation and apply them to our contemporary, cross-cultural philosophical interests, but then we could not claim to be representing what he was actually doing with Buddhist categories. While I do not deny the legitimacy of such a comparativist undertaking as the application of Buddhaghosa’s ideas to concerns far removed from his, I think it is important to see how he can contribute to a cross-cultural philosophy more directly, from closer to his own interests, through following his own words across his oeuvre. In taking this path between philosophy and Indology, I have tried to show that Buddhaghosa’s great contributions to the emerging global philosophy would be in what we can discern as a highly original hermeneutic phenomenology.

Notes 1 For more on the complex history and meaning of the term Theravāda, see Skilling et al. 2012. 2 For the historical discussion of what is and what is not the work of the fabled fifth-century figure Buddhaghosa, see works by Lance Cousins (2015), Tosiichi Endo (2013), and Oskar von Hinüber (2008). A relatively recent outline of Buddhaghosa’s oeuvre is given by Ole Pind (1992). 3 See Heim in this volume for the argument that the canonical materials themselves, as found in the Dhammasaṅgaṇī and Vibhaṅga, should not be interpreted as offering an ontology of irreducible entities. 4 Damien Keown, for example, disparages what he takes to be “dry and disconnected classifications” on virtue (sīla) (1996, 61). 5 He mentions its concern with the distinction between “definitive meaning” (nītārtha in Sanskrit) and “interpretable meaning” (neyārtha) and passes over it as being “merely” about explicit and implicit meaning (Thurman 1978, 25–26). We may contrast this with Heim’s remark about the Pāli tradition that, while the Buddha sometimes spoke in nītattha and at other times in a neyyattha manner, “often the first interpretive task is to discern which is which” (2018, 7), an undertaking which is surely not a simple matter, let alone a simplistic one.

343

Chakravarthi Ram-Prasad 6 The Hindu traditions combine these through the distinction between a fixed set of “heard” (śruti) and an expansive category of “remembered” (smṛti) texts that call for different exegetical strategies. 7 For a spirited defence of ontotheology as a sustainable approach to Christian thought, see Adams 2014. 8 I use the Chapter.Paragraph style of reference given in Ñāṇamoli’s translation. Although I have benefited from this major work, translation in the rest of this chapter is mine. 9 The Pali Text Society’s Dictionary offers the strict meaning of cintāmayā as “consisting of pure thought” but then, in a leap of lexical faith, adds as an alternative “metaphysical.”

Bibliography Adams, Marilyn. 2014. “What Is Wrong with the Ontotheological Error?” Journal of Analytic Theology 2: 1–12. Backman, Jussi. 2016. “Hermeneutics and the Ancient Philosophical Legacy. Hermēneia and Phronēsis.” In The Blackwell Companion to Hermeneutics, edited by Niall Keane and Chris Lawn, 22–33. Oxford: Blackwell. Cousins, Lance. 2015. “The Case of the Abhidhamma Commentary.” Journal of the Association of Buddhist Studies 36/37: 389–422. Dostal, Robert J. 2002. “Gadamer’s Relation to Heidegger and Phenomenology.” In The Cambridge Companion to Gadamer, edited by Robert J. Dostal, 247–66. Cambridge: Cambridge University Press. Endo, Toshiichi. 2013. Studies in Pāli Commentarial Literature: Sources, Controversies, and Insights. Hong Kong: Centre of Buddhist Studies. Gadamer, Hans-Georg. 2004. Truth and Method. Translated by Joel Weinsheimer and Donald Marshall. 2nd ed. London: Continuum. Gold, Jonathan. 2015. Paving the Great Way: Vasubandhu’s Unifying Buddhist Philosophy. New York: Columbia University Press. Griffiths, Paul J. 1989. “Buddha and God: A  Contrastive Study in Ideas About Maximal Greatness.” Journal of Religion 69 (4): 502–29. Habermas, Jürgen. 1992. Postmetaphysical Thinking: Philosophical Essays. Translated by William Mark Hohengarten. Cambridge, MA: MIT Press. Heidegger, Martin. 1962. Being and Time. Translated by John Macquarrie and Edward Robinson. New York: Harper & Row. Heim, Maria. 2014. The Forerunner of All Things: Buddhaghosa on Mind, Intention, and Agency. New York: Oxford University Press. ———. 2018. Voice of the Buddha: Buddhaghosa on the Immeasurable Words. New York: Oxford University Press. Heim, Maria, and Chakravarthi Ram-Prasad. 2018. “In a Double Way: Nāma-Rūpa in Buddhaghosa’s Phenomenology.” Philosophy East and West 68 (4): 1085–115. Jaini, Padmanabh S. 2001. “On the Sarvajñatva (Omniscience) of Mahāvīra and the Buddha.” In Collected Papers on Buddhist Studies, edited by Padmanabh S. Jaini, 97–122. New Delhi: Motilal Banarsidass. Keown, Damian. 1996. “Morality in the Visuddhimagga.” Journal of the International Association for Buddhist Studies 6 (1): 61–75. Knotts, Matthew W. 2014. “Readers, Texts, and the Fusion of Horizons: Theology and Gadamer’s Hermeneutics.” Acta Universitatis Carolinae Theologica 4 (2): 233–46. Lopez, Donald S. 1998. Elaborations on Emptiness: Uses of the Heart Sutra. Princeton: Princeton University Press. Matilal, Bimal. 1986. Perception: An Essay on Classical Indian Theories of Knowledge. Oxford: Clarendon Press. Mills, Ethan. 2018. Three Pillars of Skepticism in Classical India: Nāgārjuna, Jayarāśi Bhatta and Śrī Harṣa. Lanham: Lexington Books. Ñāṇamoli, Bhikkhu, trans. 1975. The Path of Purification (Visuddhimagga). Kandy: Buddhist Publication Society. ———. 1987. The Dispeller of Delusion. Translation of Sammohavinodanī. Revised by L. S. Cousin, Nyanaponika Mahāthera, and C. M. M. Shaw. London: Routledge and Kegan Paul. Ó Murchadha, Felix. 2016. “Religion.” In The Blackwell Companion to Hermeneutics, edited by Niall Keane and Chris Lawn, 77–85. Oxford: Blackwell.

344

Buddhaghosa Perrett, Roy W. 1989. “Omniscience in Indian Philosophy of Religion.” In Indian Philosophy of Religion, edited by Roy W. Perrett, 125–42. Dordrecht: Kluwer Academic Publishers. Pind, Ole Holten. 1992. “Buddhaghosa: His Works and Scholarly Background.” Bukkyō kenkyū 21: 135–56. Ram-Prasad, Chakravarthi. 2001. Knowledge and Liberation in Classical Indian Thought. Basingstoke: Palgrave Macmillan. ———. 2018. Human Being, Bodily Being: Phenomenology from Classical India. Oxford: Oxford University Press. Ricoeur, Paul. 1995. “Manifestation and Proclamation.” In Figuring the Sacred: Religion, Narrative, and Imagination, edited by Mark I. Wallace, 48–67. Minneapolis, MN: Fortress Press. Skilling, Peter, Jason A. Carbine, Claudio Cicuzza, and Santi Pakdeekham, eds. 2012. How Theravāda is Theravāda? Exploring Buddhist Identities. Chiang Mai: Silkworm Books. Thurman, Robert A. F. 1978. “Buddhist Hermeneutics.” Journal of the American Academy of Religion XLVI (1): 19–39. Von Hinüber, Oskar. 2008. A Handbook of Pali Literature. New Delhi: Munshiram Manoharlal. Weinsheimer, Joel C. 1985. Gadamer’s Hermeneutics: A Reading of Truth and Method. New Haven, CT: Yale University Press.

345

20 BHĀVIVEKA Madhyamaka Dialectic, Doxography, and Soteriology Karl-Stéphan Bouthillette

The Author and His Context The name “Bhāvaviveka” was first introduced to a Western audience by Eugene Burnouf (1801– 1852), designating an early Buddhist Madhyamaka philosopher whose life and seminal work generated substantial scholarly interest (Burnouf 1844, 499).1 The exact recording of the name, however, varies depending on sources.2 Tibetan sources mainly gave the name Bhavya, though it is not found in any known Sanskrit text. Chinese translations consistently use a set of characters matching the name Bhāviveka, from the Sanskrit bhā, for ‘light,’ ‘brightness,’ and so on, and viveka, for ‘discrimination’ and ‘discernment.’ Hence, what looks like a laudatory title refers to a man of ‘brilliant intellect’; it is a suitable name for the sharp dialectician who was the first Indian philosopher to systematically compile, synthesize, and criticize all the prominent philosophical views of his time, including those alien to his Buddhist tradition, within a coherent dialectical compendium, known in English as the Verses on the Essence of the Middle Way, the Madhyamakahṛdayakārikā (MHK). This doxographical3 text is the central focus of the present discussion. Although seven other titles are attributed to the famous dialectician, only two more are generally acknowledged to be from his hand: the Lamp of Wisdom (Prajñāpradīpa, PP)4 and the Jewel in the Hand Treatise (Karatalaratna, KR) (Iida 1980, 12–19; Watanabe 1994, 9–15). The first is Bhāviveka’s commentary on Nāgārjuna’s Fundamental Verses on the Middle (Mūlamadhyamakakārikā, MMK), which was commented upon by Avalokitavrata and is now available only in Tibetan and Chinese. The second is Bhāviveka’s final and shortest of the three works mentioned and appears to be a concise and more accessible version of the MHK, summarizing Bhāviveka’s arguments against the Sāṃkhya, Vaiśeṣika, and Yogācāra. It is available only in Chinese (Dàchéng zhǎng zhēn lùn/大乘掌珍論).5 Thus, the MHK alone is extant in Sanskrit, while its commentary, the Blaze of Reasoning (Tarkajvālā, TJ), is preserved in Tibetan only. A shorter text is sometimes added to the list of Bhāviveka’s accepted oeuvre, as in Eckel and Eltschinger (2019), A Summary of the Meaning of the Madhyamaka (Madhyamakārthasaṃgraha, MAS), which focuses on Bhāviveka’s understanding of the two levels of truth in Buddhism: conventional truth (saṃvṛtisatya) and ultimate truth ( paramārthasatya). Compared with some other Indian philosophical figures, putting a date on the life of Bhāviveka is a relatively easy task because his works engage many well-known Buddhist 346

DOI: 10.4324/9781351030908-31

Bhāviveka

thinkers such as Dharmapāla, Buddhapālita, Dignāga, and Sthiramati. Today, scholars believe Bhāviveka lived around 490–570 or 500–570 CE (Kajiyama, 1968–1969). Pinpointing a precise location for the author’s life is a less certain enterprise. Chien Y. Hsu concluded, from a review of the many sources at hand, that “Bhāviveka was born into a royal family in Dhānakataka, which is presently a western neighbor of Amarāvatī” (Hsu 2013, 14). By the second century CE, Amarāvatī was one of the greatest Buddhist centers of the eastern Deccan (Verardi 2011, 74). By the sixth century, when Bhāviveka was active, it was one of the three major centers of Buddhist studies in India, along with Nālandā in the east and Valabhī in the west. Bhāviveka was thus active in the agonistic context of sixth-century India. By the second half of the fifth century CE already, we notice the beginning of a sustained philosophical confrontation between competing groups in the form of a systematic approach to debate (vāda) and critical examination ( parīkṣā), making use of new developments in the field of argumentative reasoning (hetuvidyā), which were to develop into a fully epistemological framework ( pramāṇa), used and debated throughout the Indian cultural sphere. It is within this context that Bhāviveka began his intellectual work. But why would philosophers suddenly feel obliged to strengthen their position and to engage with the views of others, and not only with those circulating within their own religious tradition? Although the trend was already set in the second to third centuries by the works of Nāgārjuna and Āryadeva, for example, it reaches a new systematic height in Bhāviveka’s oeuvre. Vincent Eltschinger suggests that these innovations, though introduced by new developments in thought, were also influenced by a given sociohistorical context: “a dramatic increase of hostility towards Buddhism as the Gupta dynasty starts to crumble” (Eltschinger 2014, 72). If doctrinal issues could fuel such animosity, one can safely assume that, as elsewhere, Indian religions and politics became intimately intertwined in an explosive mix. Eltschinger convincingly argues that the social dynamics of the sixth century onwards brought about two major innovations in Indian Buddhism: Buddhist esotericism6 (tantra) and epistemology ( pramāṇa). He notices how both “legitimate themselves in a structurally homologous way, and in quite martial terms, as that which, by defeating the outsiders, removes the obstacles to the path towards liberation” (Eltschinger 2014, 174). The work of Bhāviveka gains in depth and historical significance when framed within that context: the emergence of a new apologetic strategy of Buddhism, making use of epistemology to defend itself and to attack the positions of others. Debate, indeed, preoccupied the sixth-century dialectician. However, assessing Bhāviveka’s MHK, one notices how this dialectical endeavor was not a purely worldly pursuit after fame and scholarly career, nor a mere apologetic strategy, but it was conceived as an essential dimension of the spiritual practice leading to the highest realization of religious life. Few attempts have been made to reflect on Bhāviveka’s work in the light of the turbulent sociohistorical context of the sixth century CE. There has been ample discussion of the reasons that led Tibetan doxographers to classify Bhāviveka as a proponent of the trend of Madhyamaka which favors independent (svatantra) inference (anumāna) in debate, a school known as Svātantrika-Madhyamaka, in opposition to the school championed by Candrakīrti, known as Prāsaṅgika-Madhyamaka, which rather favors reducing its opponent’s assertions to their absurd ‘consequences’ ( prasaṅga). More precisely, since Bhāviveka’s methodology also seems to favor direct references to the texts (sūtras) attributed to the Buddha rather than resorting to the later scholastic materials of the Abhidharma literature, some Tibetan doxographers classified him as a Sautrāntika-Svāntantrika-Madhyamaka thinker. They thus considered him a Mādhyamika commentator who relies on the sūtras and on independent inference (svatantraanumāna) in support of his arguments. This classification stems from Bhāviveka’s logical method deployed to elucidate Nāgārjuna’s Fundamental Verses on the Middle Way in his Lamp 347

Karl-Stéphan Bouthillette

of Wisdom. Critical of the previous Buddhapālita’s conventional method of argumentation on the same verses, relying strictly on the exposition of absurd consequences ( prasaṅga), Bhāviveka rather promotes a rendering of Nāgārjuna’s arguments into the form of a standard three-part inference inherited from the famous Buddhist logician Dignāga (480–540 CE). Candrakīrti eventually came to the defense of Buddhapālita, whence the Tibetans saw in this opposition the birth of two separate lines of interpretation: the Svātantrika branch of Madhyamaka inaugurated by Bhāviveka and the Prāsaṅgika branch defended by Buddhapālita and Candrakīrti. While much attention has been given to the philosophical implications of both positions, little consideration has been given to the sociohistorical context Bhāviveka was facing, such as the rising antagonism between Buddhism and Brahmanism, an ideological confrontation which likely contributed to his stance on inference (anumāna). As Giovanni Verardi rightly stressed, “Indian Buddhism should not be studied per se, but in counterpoint with Brahmanical theorisations” (2011, 106).7 Taking into consideration the pan-Indian context of philosophical disputation (vāda) where one is required to assert his own position prior to engaging in the fray of debate, a prerequisite already stated in the Brahmanical Nyāya-Sūtra-s of Gautama, and knowing how crucial debates had become for the financial and physical security8 of the debater and of his following, one might better appreciate the desire of Bhāviveka to elaborate on and defend a solid dialectical position of his own in the name of the Middle Way (Madhyamaka).

Status of Bhāviveka’s Work in Buddhist Philosophy For the sake of the overall coherence of the present Handbook, Bhāviveka’s chapter has been placed within the section on ‘philosophical commentators.’ As attested by his earliest work, the Lamp of Wisdom, Bhāviveka was indeed concerned with the interpretation of Nāgārjuna’s thought, which laid the foundations of Madhyamaka Buddhism. Yet categorizing him as a philosophical commentator should not obscure the fact that Bhāviveka was an original thinker in his own right, as well as an acute synthesizer and staunch apologist. These salient features of his thought are observable in his magnum opus, the Verses on the Essence of the Middle Way. Bhāviveka’s interpretation of his revered Madhyamaka master, Nāgārjuna, quickly became authoritative. But, like most commentators, he was not satisfied with repeating verbatim the stances of his predecessors. For example, as shall be discussed more in depth, his logical method of argumentation required him to finely distinguish the two levels of truth, bringing forth a more nuanced application of this distinction than that articulated by Nāgārjuna. Bhāviveka also extended the scope of Nāgārjuna’s dialectical method to encompass the philosophical developments of his days, such as the three-part inference (anumāna) inherited from Dignāga and the epistemological framework of the standard means of knowing ( pramāṇa). In India, according to reports of Chinese pilgrims, Bhāviveka’s influence was so important that he came to be considered the legitimate successor of Nāgārjuna. Nāgārjuna and his pupil Āryadeva (second–third century CE) did not seem to use the term “Mādhyamika” to designate their philosophy. It was mainly due to Bhāviveka that it acquired its name and became an established philosophical system, equipped with a rigorous logical and dialectical methodology to engage with opposition – something which was missing when the Mādhyamikas resorted to prasaṅga alone. Modern scholars have argued that the rigor and the norm of the Mādhyamika system owes a great debt to Bhāviveka and that most later Indian Mādhyamika commentators followed in his footsteps (Iida 1980, 19–23). Prior to Bhāviveka, as Mahāyāna philosophy developed based on a new set of sūtras, notably the ‘Perfection of (Transcendent) Wisdom’ literature (the Prajñāpāramitā sūtras, composed somewhere between the first century BCE and the sixth century CE), proponents of the new 348

Bhāviveka

movement, including Mādhyamika and Yogācāra scholars, made a particular effort to distinguish their commentaries from those of the Ābhidharmikas. Such treatises dialectically engaged with the views of Buddhist and non-Buddhist thinkers.9 Bhāviveka’s MHK is an excellent example of this trend, which emphasizes debate and logical arguments. Āryadeva, an earlier proponent of Nāgārjuna’s thought, in his Catuḥśataka and Śataśāstra, had already critically engaged the views of non-Buddhist schools such as Sāṃkhya, Vaiśeṣika, Jaina, Lokāyata, and even Vaiṣṇava and Śaiva (Warder 1970, 368–69). Bhāviveka, however, was the first Mādhyamika to adopt the method of the ‘logical science’ (hetuvidyā) developed by Dignāga. He established his own strategy of presenting independent (svatantra) logical demonstrations – inference (anumāna) – in defense of the Madhyamaka doctrine, in contrast to prior Madhyamaka methodology, which had employed reductio ad absurdum arguments ( prasaṅga)10 against the opponent’s thesis without presenting any counter-thesis. Bhāviveka’s major inspiration for the composition of a doxographical treatise (the MHK), according to Olle Qvarnström, may have been Dignāga’s Compendium of Valid Cognition (Pramāṇasamuccaya, PS), which addresses Nyāya, Vaiśeṣika, Sāṃkhya, and Mīmāṃsā, drawing on the methods of the epistemological ( pramāṇa) tradition. Bhāviveka’s further innovation would thus have been to supplement Dignāga by criticizing from the Mādhyamika perspective the various schools known to him which discuss epistemological and logical claims ( pramāṇa), adding the Vedānta system, for example, to the ones found in the Pramāṇasamuccaya (Qvarnström 1988, 6). Moreover, I have argued that Bhāviveka’s presentation of each system of thought within the MHK follows a systematic dialectical teleology meant to establish his philosophical position as the sole valid soteriological method, a rhetorical device which inspired subsequent Indian doxographies (Bouthillette 2020).

The Doxographical11 Method and Its Function Within the MHK The Sanskrit manuscript of the Verses on the Essence of the Middle Way was first discovered by the adventurer Pandit Rāhula Sāṅkṛtyāna in August 1936 in the archive of Shalu (zhwa lu) monastery. Iida informs us that the manuscript, ascribed to Bhavavadviveka, was written in Rañjanā script.12 The MHK consists of roughly 927 anuṣṭubh verses (of two lines, each with 16 syllables) and is divided into 11 chapters. It is one of the first comprehensive and systematic study of the doctrines (siddhānta-s) of all major schools of Indian philosophy of the time and was composed when these philosophies were still in the process of formation. Hence, it is an invaluable historical document, even if its purpose is dialectical and not historiographical. At the same time, or rather in the first place, as an independent work, it presents an original articulation of Bhāviveka’s own Madhyamaka system. In that sense, it is also a genuine philosophical reflection, with numerous witty metaphors and poetic allusions to a wide array of pan-Indian literary and cultural lieu communs. If we locate the MHK as the first example of a historically extended and broad literary genre of systematic philosophical writings, generically referred to as ‘Indian doxography,’ then the overall structure of the text becomes a key element to interpret its philosophical method and the way its worldview is articulated in relation to competing claims. Hence, it is significant to notice that the structure of the whole text develops out of a primary section presenting the author’s own position elaborated through an emptying out of Abhidharmic categories. This is to say that the introductory section of the MHK, Chapter Three, discusses core topics in scholastic Abhidharma, highlighting how every constituent (dharma) of any mental and physical event listed within the Abhidharma account of sentient experience is devoid of ‘intrinsic nature’ (svabhāva) and is thus empty (śūnya). This initial treatment of Abhidharma categories allows Bhāviveka to establish the position of his own Madhyamaka system without 349

Karl-Stéphan Bouthillette

actually enunciating an independent set of propositions in its name. This method stands in sharp contrast to the doxographies composed by philosophers defending other systems. Wilhelm Halbfass observes, for example, that the structure of Advaita Vedānta doxographies, such as the ones doubtfully attributed to Śaṅkara and Mādhava, tends to present a hierarchical classification at whose apex stands the Vedānta (Halbfass 1988, 351). In the case of the MHK, every view is refuted from the beginning based on the no-view principle of the Madh­ yamaka, expressing the perspective of the ultimate truth, which itself barely admits being a view. Indeed, the Madhyamaka subscribes to an attitude found early on in the discourses of the Buddha, in the Collection of Discourses (Sutta Nipāta), for example, to the effect that the speculative and religious views (dṛṣṭi) of non-Buddhists are tainted mental states obscuring the immediate perception of reality. For this reason, the Madhyamaka presents its dialectical methodology as an antidote devised to alleviate the affliction of those who grasp onto views. Hence, it does not elaborate a view of its own but rather negates the views of others, as medicine removes diseases to restore health, in order to open their eyes on the true nature of things, the ultimate ‘emptiness’ (śūnyatā) or ‘dependent co-arising’ ( pratītyasamutpāda) of all phenomena. Medical metaphors alluding to the healing function of the Madhyamaka pervade the MHK. For example, in Chapter Nine, Bhāviveka argues that Then, it should be granted that, since it is the antidote to that ignorance, liberation comes from knowledge, just as freedom from sickness comes about through medicine, because it is the antidote to illness. (MHK 9.21, Bouthillette 2020, 37) Accordingly, Madhyamaka is not a mere theoretical view but a remedy to the various forms of mental grasping caused by speculative philosophies. This is what Bhāviveka expresses in the following passages from the MHK, translated by Chikafumi Watanabe, where the eye is clearly identified as the locus of application for the Madhyamaka’s remedy: When the eye-disease passes away, one whose eye becomes completely clean and pure does not see the hair (keśa), mosquito (maśaka), two moons (dvicandra) and the eye in a peacock’s plumage (śikhicandraka). Likewise, when the eye-disease of darkness which envelopes what is to be known and defilements passes away, a wise man whose eye has become pure by means of proper knowledge does not see anything. (MHK 3.251–2, Watanabe 1994, 85) Not seeing anything (superimposed over reality) metaphorically captures the highest wisdom gained through the Mādhyamika negational analysis of reality. Thus, the Madhyamaka is said to be a non-view. It does not add a view to one’s mind; it removes existing ones. As Jan Westerhoff notes, the general position of the Madhyamaka is not foundational, nor does it develop out of the “theory-free or unsophisticated conventionalities of the ordinary man on the street” (2014, 123). Rather it works out of the intricate analytical framework of the Abhidharma, which it accepts as a somewhat reliable depiction of the conventional world, once properly deconstructed to be made svabhāva-free, free of any notion of ‘intrinsic nature.’ In order to point out the ultimate, then, Mādhyamikas like Bhāviveka do not elaborate their own systematization of conventional categories but analyze and refute existing ones. In brief, they deconstruct the conventions of others to assert the ultimate truth, the no-view of emptiness. Bearing this method in mind, doxography, within the MHK, becomes an ingenuous 350

Bhāviveka

soteriological and propaedeutic device where ‘poison’ is used as ‘medicine.’ It indicates that all doctrines presented within the MHK are taken as more or less problematic depictions of the conventional world to be analyzed and ultimately relinquished by those seeking ultimate truth. The sequence of presentation of each view within the MHK suggests that the further away a system stands from the initial exposition of the author’s own stance, the more misguided it is and the more it stands in contrast to the Madhyamaka.13 This overall scheme confirms Westerhoff’s observations to the effect that: [W]hen the Mādhyamika deliberates between different systems of categories at the conventional level he is considering two main criteria: (a) its general philosophical usefulness and (b) the extent to which the system can be made ‘svabhāva [intrinsic nature]-free.’ This allows him to rank such systems and provides criteria to choose between systems, none of which have any claim to truth at the absolute level. (2014, 131) This might explain the particular virulence of Bhāviveka’s criticism towards the Mīmāṃsakas in the last chapter, a group which he discourteously introduces as the ‘shameless’ (anapatrapa) who value ritual acts above meditation and knowledge. The MHK opens on a brief discussion about safeguarding the ‘aspiration towards awakening’ (bodhicitta-aparityāgaḥ) in Chapter One, followed by a reflection on the significance of monastic vows (muni-vrata-samāśrayaḥ) in Chapter Two and a revision of foundational Abhidharma doctrinal expositions examined in a way which reveals the ‘no-view’ approach of the Madhyamaka propounded by Bhāviveka in Chapter Three. Thus, the third chapter, titled the Quest for the Cognition of Reality (tattva-jñāna-eṣaṇā), employs the analytical method initiated by Nāgārjuna, submitting familiar Abhidharmic ontological categories to an analysis which shows that none of them has intrinsic nature (svabhāva), a hermeneutical strategy found in the text In Praise of the Supramundane One (Lokātītastava), the Verses on the Heart of Dependent Origination (Pratītyasamutpādahṛdayakārikā), and similar texts. V. V. Gokhale suggests that these first three chapters were originally composed by Bhāviveka, under the title Penetration into the Essence of Ultimate Reality (Tattvāmṛtāvatāra), to form the core of the MHK, while the remaining chapters were added later (1985, 78). However, since the whole MHK is accepted as coming from the hand of the same author, Gokhale’s remarks need to be moderated. There are numerous allusions to non-Buddhist tropes within these three initial sections, which suggests that the author already had in mind engaging with non-Buddhist materials. In these first three chapters, it is argued that reality (tattva) is to be discerned through the eye of wisdom ( prajñā), a higher intelligence essential for liberation, which Bhāviveka understands as that which is in conformity with the highest truth ( paramārtha), the nectar (amṛta) of the Madhyamaka. It is that which wise men seek. The introductory verses of Chapter Three, translated by Iida Shotaro, explain that Whoever has the eye consisting of knowledge, has the eye; the other [eye] is not [it]. Therefore the wise man should be dedicated to the quest for knowledge of ultimate reality (tattva-jñāna). (MHK 3.1) One bereft of intelligence is as if eyeless even with a thousand eyes. Because he does not see the real path to heaven and release. (MHK 3.3, Iida 1980, 56–57) 351

Karl-Stéphan Bouthillette

Ultimately, perceiving reality (tattva) amounts to seeing the buddha through nonconceptual cognition. In MHK 4.2, for example, Bhāviveka explains, that, “[t]he Teacher’s body is not the locus of nonconceptual cognition, because it is a body, like the body of a cowherd.”14 Thus, as explained in the commentary (TJ), to have nonconceptual cognition of the Buddha amounts to, for example, apprehending the four noble truths and so on. To ‘truly’ see the Buddha means to nonconceptually cognize the truth of the teachings, not merely to see the teacher’s body. In brief, the expression ‘seeing the Buddha’ is but a mere linguistic expedient. Enjoying paradoxes and apophatic statements, Bhāviveka describes how: Seeing the Buddha is something which belongs to lower truth (saṃvṛti). [It is said to come] from removing the dust of afore-mentioned deficiency [and] from lighting the torch of true dharma (saddharmolkāśana). But this is a conventional [expression] as an assistance towards activity (kriyopakārarūpa). Because what is similar to illusion never takes place, is not truth, from the standpoint of [the highest] truth, that (seeing) is not really seeing. It [really seeing the Buddha]15 is beyond inference (apratarkya), is indiscernible (avijñeya), cannot be demonstrated (anirūpya), does not have an illustrative example (anidarśana), is without any mark, is devoid of any image (nirābhāsa), is beyond mental constructions (nirvikalpa), is beyond words (nirakṣara) and is to be awakened by the mind of the viewer ( paśyato buddhiboddhavya) and is an act of seeing from the standpoint of non-vision. (MHK 3.244–46, Watanabe 1994, 82–83) In paradoxical words, the philosophy of the Madhyamaka is a non-philosophy in the sense that it is not a doctrinal ‘view’ detailed and supported by an independent system of reasoning. Its highest truth is not established through mere syllogisms and arguments, nor through mere scriptures. Madhyamaka rather makes use of such means to get rid of views altogether, since they prevent one from ‘seeing’ reality as it is. Logical devices are skillful expedients on the path, and Bhāviveka makes full use of them every time he analyzes down a view presented by an opponent or, again, when it comes to analyzing the Buddhist teachings. But, by themselves, these expedients are not establishing some sort of highest truth. They are removing obstacles to seeing, like medicine removes diseases. In the remaining chapters of the MHK, after having presented his own position, Bhāviveka launches his doxographical conquest. To highlight the dramatic potential of the text, I read Bhāviveka’s philosophical venture like the conquest of a ‘World Monarch’ (cakravārtin) of philosophy, as he begins his dialectical campaign on his own turf by directing his attacks at the doctrines of previous Buddhists, which he regroups under the label of the ‘hearers’ of the Buddha’s teachings (the śrāvakas) (MHK chapter 4: śrāvaka-tattva-viniścayāvatāra) and of the proponents of Yogācāra or the proponents of consciousness (vijñānavādins; MHK chapter 5: yogācāra-tattvaviniścayāvatāra). Through his corrective critique, Bhāviveka’s MHK strategically unifies the realm of Buddhism under his banner before marching into Brahmanical lands. His criticism is meant to stress that not only are the teachings of the Great Vehicle (mahāyāna) – in opposition to those of the hearers – authoritative but that their presentation articulated within his Madhyamaka conveys their ultimate meaning – a reply to the exclusivity claim of the Yogācāras (Eckel 2008, 62–81). Malcolm David Eckel summarizes Bhāviveka’s critic of the two groups of Buddhist opponents as follow: In the case of the Yogācāras, the error takes two forms: they fall into the extreme of “improper denial” (apavāda) by denying the reality of imagined identity, and they

352

Bhāviveka

fall into the extreme of “improper reification” (samāropa) by affirming the reality of dependent and absolute identity. In the case of the Śrāvakas, the mistake has to do principally with an act of reification (they take the eightfold path as if it were a real thing). (2008, 80–81) Bhāviveka invites these competing schools of Buddhism to abandon their extreme views and to join ranks and adopt what he considers the best approach to interpreting Buddhist scriptures, the Madhyamaka. Following these intra-Buddhist skirmishes, Bhāviveka expands his criticism to the nonBuddhist school of the Sāṃkhya (MHK chapter 6: sāṃkhya-tattvāvatāra), portrayed as entertaining two different theories concerning the process of knowledge, which, according to Olle Qvarnström, are “reminiscent of the doctrine of two truths (dvisatya)” found in Madhyamaka Buddhism and in Advaita-vedānta, among others (2012, 398). In the following chapter (MHK chapter 7: vaiśeṣika-tattvāvatāra), the theory of liberation of the Vaiśeṣika becomes the main target.16 Then, Bhāviveka becomes the first-ever Buddhist critic of Vedānta (MHK chapter 8: vedānta-tattvaviniścaya) as such, particularly criticizing its theory of the self as having an intrinsic nature and suggesting that all that is well spoken in Vedānta was taken from Buddhism.17 Finally, the author kept his utmost wrath for the Mīmāṃsakas (MHK chapter  9: mīmāṃsā-tattvanirṇayāvatāra). This last chapter, however, presents a hodgepodge of Brahmanical views going beyond what came to be traditionally designated as Mīmāṃsā, ridiculing the belief in such deities as Śiva and Kṛṣṇa, for example. In the end, after having trampled the Brahmins, Bhāviveka finally turns his gaze towards the notion of omniscience (sarvajñasiddhinirdeśa), pivotal in Jainism, Yoga, and Śaivism, on which he will spend a few verses, as in a postscript, before setting the final boundary of his doctrinal conquest. This overall dialectical strategy is meant to conquer the entire realm of philosophy that preoccupied Bhāviveka. It sets the tone for later Indian philosophical compendia. The eloquence of the MHK, gilded with subtle allusions and humor, at times incredibly scathing, also makes it a literary masterpiece. On several occasions, Bhāviveka is able to skillfully appropriate non-Buddhist tropes and repurpose them in a Buddhist context. For example, the concluding verses of the second chapter present the purpose of the MHK, its method, and its target: Bhāviveka calls upon his brethren and his opponents to investigate their own beliefs through sound reasoning ( pratisaṃkhyāna) in order to realize the truth which lies therein, the resplendent Sāvitrī of dependent origination. This reference to a famous Brahmanical devotional prayer would immediately be recognized by Bhāviveka’s audience, identifying Sāvitrī with the solar deity found in the Ṛg-veda and made famous by the celebrated verses RV. iii, 62.10, also called gāyatrī, particularly used during the initiations of the “twice-born” or brahmins. Here, Bhāviveka suggest that the real light of wisdom (remember Bhāviveka’s earliest work, the Lamp of Wisdom or Prajñāpradīpa) is the one shining in emptiness (śūnyatā). This was the impetus of the first Indian systematic philosophical doxography. Thus, the MHK is not a mere historiographical account of sixth-century Indian philosophy but the exposition of a dialectical training aimed at converting the views of those who hold erroneous ones; at eradicating doubts in minds confused by competing claims; and at consolidating the insight of the others who, having given birth to knowledge already, seek to see ‘no-thing.’18 Within philosophy, Bhāviveka is chiefly concerned with soteriology. The fundamental problem in each of the ideas he refutes is that they are impediments to liberation. The rest is method.

353

Karl-Stéphan Bouthillette

Nature of the Path and Purpose of the MHK In the eyes of Bhāviveka, wisdom ( prajñā) is gained through the analytical process of negation, which leads one to see through ‘not seeing’ (adarśana), ultimately leaving one with no view. This is made clear by his definition of ultimate wisdom: Ultimate wisdom negates the entire network of concepts, and it moves without moving into the clear sky of reality, which is peaceful, directly known, nonconceptual, nonverbal, and free from unity and diversity. (MHK 3.10–11, Eckel 2008, 41–42) As already elucidated by Nāgārjuna, ultimate reality ( paramārtha), in Bhāviveka’s system, relies on a corresponding relative (saṃvṛti) truth.19 In terms of Madhyamaka practice, the crux of the matter is to know and explain how to perceive relative views and similarly attain ultimate reality, the non-view, ‘liberation’, in other words, how to move from a produced and thus impermanent conventional knowledge to a permanent nonconceptual one. For that purpose, Bhāviveka innovated by integrating formal logical demonstrations, inherited from Dignāga, into his therapeutic methodology. To better understand how this process is supposed to work, one must become familiar with the three possible ways Bhāviveka understood the meaning of ‘ultimate reality.’ Conveniently, Hsu has summarized how, for Bhāviveka, ultimate reality contains three connotations: 1) from an ontological perspective, “ultimate reality” means the ultimate object which is understood as the true nature of things; 2) from an epistemological perspective, ultimate reality is nonconceptual wisdom, which refers to how the Buddhist sages view reality; 3) the teachings in accord with non-arising is ultimate reality.20 This compact summary can be traced back to the TJ commentary breaking down Bhāviveka’s first syllogism within the MHK. Watanabe provides the translation of the syllogism and of Bhāviveka’s own commentary: (proposition) from the standpoint of the highest truth ( paramārthataḥ) the earth, etc. do not have the gross elements as their own-beings (intrinsic nature), (reason 1) because they are things which are made, or (reason 2) because they are things which have cause, etc., (instance) just like knowledge (  jñāna). (MHK 3.26, Watanabe 1994, 18) Commentary: In [the word] paramārtha, artha is what is to be obtained ( pratipattavya) and what is to be understood (adhigantavya) because artha is the object to be known (  jnātavya). Paramārtha means21 “the most excellent.” The compound parama-artha [can be interpreted in three ways]. (1) It means “the most excellent object” because it is the object and the most excellent (karmadhārya compound) (2) Or, it is the object of the most excellent. That is, because it is the object of the most excellent knowledge that is beyond discrimination (nirvikalpa-jñāna), it means the object of the most excellent (tatpuruṣa compound) 354

Bhāviveka

(3) Or, it is “comformable to paramārtha” ( paramārthānukūla). That is, since there is that paramārtha in wisdom ( prajñā) which is conformable to obtaining of paramārtha, it means “one which is in conformity with paramārthaḥ” (bahuvrīhi compound). Paramārthataḥ [in syllogisms] is taken as the third type of compound (the bahuvrīhi compound), i.e., paramārthataḥ, in the ultimate reality. (TJ; Watanabe 1994, 18–19) These kinds of subtle definitions demonstrate what is meant when scholars say that Bhāviveka brought forth a much more nuanced application of the distinction between the two truths than that articulated by Nāgārjuna. Here, Bhāviveka begins his syllogism by stating the perspective from which it is to be understood. This perspective is that of the highest truth ( paramārthataḥ), in opposition to that of the relative (saṃvṛti) truth. The commentary explains that the highest truth can be understood in three cumulative ways, which correspond to the summary given by Hsu previously. The first two definitions refer to truth itself, which lies beyond concepts and thus cannot be understood through any conventional means. However, when it comes to the highest truth mentioned in a syllogism, it is to be understood as the third interpretation: that which is in conformity with the highest truth, which is the wisdom ( prajñā) found in Buddhist teachings, such as those of Bhāviveka. Thus, the third definition involves concepts and language, used in such a way as to approach truth itself. It allows Bhāviveka to formulate a logical position on topics of a conventional nature (the earth, etc.) by delimiting its perspective (that of a specific understanding of the absolute) to avoid contradictions (the absolute, merely understood as being nonconceptual, cannot be referred to in the analysis of concepts). Watanabe fleshes out the syllogism further, but, for our purpose, it suffices to notice that it has the characteristic of a threepart inference (anumāna) with (1) a proposition (the earth, etc. do not have the gross elements as their intrinsic nature), (2) a reason (reasons 1 and 2), and (3) an illustration (like knowledge). This last part, the illustration, can help us shed more light on Bhāviveka’s understanding of the function of logic in matters of liberation. To one who might ask in what manner knowledge is (  jñāna) a produced thing, like the earth and so on, Bhāviveka’s auto-commentary replies: “it is accomplished by logic (tarka) which aims at heaven (svarga) and nirvana (apavarga)” (Watanabe 1994, 24). Clearly, Bhāviveka gives a soteriological function to logic (tarka). That function must be explained in relation to Bhāviveka’s understanding of ultimate reality. By introducing these three nuances within his definition of the highest reality, Bhāviveka allows for a twofold nature of ultimate reality, one devoid of volitional concepts (anabhisaṃskārapravṛtti) and another which involves them (sābhisaṃskārapravṛtti) (Watanabe 1994, 23). The third definition of the ultimate truth allows conventional means such as logic (tarka) within its scope, whereas the other two definitions forbid such means of access to truth. This volitional aspect of wisdom ( prajñā) can now act as a lever to pass from the conceptual to the nonconceptual, through the door of the Buddhist teachings. To understand how this works, this volitional wisdom can be further divided into three stages. The first is obtained from hearing teachings and is hence called the ‘wisdom acquired from hearing’ (śrutamayīprajñā). At this level, conceptuality and language remain throughout. Then, through a critical analysis of scriptures, which have been ‘heard,’ a second wisdom is produced, intellectual in nature, called the ‘wisdom acquired through reflecting’ (cintāmayīprajñā) on the teachings, which could be described as an internal and critical dialogue with the teachings, an analytical contemplation of their meaning by means of logic. At this point, conceptuality remains but becomes more abstract. This second and subtler 355

Karl-Stéphan Bouthillette

conceptual activity is meant to support a third and last layer of wisdom, a nonconceptual one this time, a direct experience of contemplative insight called the ‘wisdom born out of cultivation’ (bhāvanamayīprajñā), often rendered as meditative wisdom. Thus, the divide between the conceptual and the nonconceptual is said to be progressively crossed over by a bridge whose entry is rooted in firm concepts (heard – śruta – from scriptures) and that ends in pure nonconceptual perception, tied at its center by the ever-subtler junction ( yoga) of reasoning and analysis. This theoretical framework, outlining the pivotal role of reasoning and hermeneutics in Mahāyāna religious praxis, was already enunciated in the well-known third-century Treatise on the Foundation for Yoga Practitioners (Yogācārabhūmiśāstra).22 Bhāviveka’s account of ultimate truth ( paramārtha) is informed by this framework, which strengthens the soteriological resonance of his Dignāga-inspired practice of inference. This gradual threefold mode of insight into the nature of reality has also been asserted by other prominent Buddhist thinkers, such as Dignāga, Dharmakīrti,23 and Kamalaśīla. In order to cultivate the non-view, they argued that worldly conventional views had to be removed, like cataracts from the eye, with the appropriate critical antidote. This is precisely the task of the MHK. In other words, the existential disease and its remedy are found within the conventional realm. It is through conceptual analysis, directed by logic (tarka), that one can finally move beyond conceptuality altogether. The conventional alone – through language, logic, and so on – allows the meaning of scriptures to indicate the ultimate, which, in and by itself, has no indicator and nothing to indicate. Scriptures are composed according to conventions and must thus be understood through conventional means (inference). The wisdom born from reasoning is thus a crucial step on the stairway to the conceptual palace of liberation, a metaphor dear to Bhāviveka: The nectar of wisdom, bestowing satisfaction, is a lamp of unobstructed light, the stairway to the palace of liberation, the oblation-fire which burns afflictions. (MHK 3.6, Bouthillette 2020, 37)

Notes 1 Although not exhaustive, the following list covers the most important scholarly contributions to date, in relation to Bhāviveka and the MHK: Ames 2019; Bahulkar 1994; Eckel 1992, 2008; Ejima 1980b, 1980a; Eltschinger 1998; Gokhale 1958, 1972; Ham 2016; He 2014; Heitmann 1995, 2004; Hirabayashi and Iida 1978; Honda 1967; Hoornaert 1999, 2001, 2003, 2004; Hsu 2011, 2013; Iida 1966, 1980; Jones 2011; Kawasaki 1973, 1976; Krasser 2011, 2012; La Vallée Poussin 1932–1933, 1933; Lindtner 1982, 1986, 1995, 2001a, 2001b; Miyasaka 1954a, 1954b, 1958; Nakada 1972; Nakamura 1950, 1955, 1958, 1968; Potter 2003; Qvarnström 1989, 1999, 2012, 2015; Saitō 2005; Ruegg 1990; Warder 1970; Watanabe 1994, 1998, 2012; Yamaguchi 1941 [1964]. 2 One can find: Bhavya, Bhavyaka, Bhavyakāra, Bhavyaviveka, Bhavyavivikta, Bhavyaviveka, Bhāvivika, Bhāvivikta, Bhagavadviveka, and Bhavavadviveka. Shotaro Iida presents a table associating the various names with their textual reference Iida 1980, 5. On Bhāviveka’s name, see Ejima 1980a; Hsu 2013, 10–12. 3 By ‘doxography,’ I  understand a text recognizable through the conjunction of six main features. A doxographical text is (1) either a whole text, or a part of a text; (2) where competing views of philosophers or schools are presented following a division of topics organized into sets and sub-sets with specific differences to which a name-label is attached in most cases; (3) where the original argument supporting such views may or may not be given; (4) where the author’s own view and arguments may or may not be criticized; (5) where the philosophical content consists either in literal or in non-literal renderings of sources; and (6) where the overall concern is primarily systematic, dialectic, and with little or no historiographical character. See Bouthillette 2020, 13–14. 4 For translations and commentaries on the PP, see Ames 1986, 1993, 1999, 2000, 2019. 5 For more information on the two other texts of Bhāviveka, see Hsu 2013, 40–43. 6 Ronald Davidson (2002) was the first to fully theorize such ideas in relation to Buddhist esotericism.

356

Bhāviveka 7 Jonardon Ganeri also made a similar argument in the context of the scholastic tradition; see Ganeri 2001. 8 Verardi’s work on the downfall of Buddhism in India gives ample illustrations of the many kinds of physical tortures and death penalties allegedly befalling those losing public debates in medieval India. 9 For a general discussion of the issue, see Hsu 2013, 9. See also Eltschinger 2014 on the engagement of Buddhist scholars with non-Buddhists. 10 Shoryu Katsura argues that Nāgārjuna succeeded in formulating the formal proof of reductio ad absurdum ( prasaṅga) probably for the first time in India (Katsura 2007, 70). The prasaṅga argumentative method basically rests on pointing out the ‘consequence’ ( prasaṅga) that results from any position or thesis operating with the idea of the real existence of an entity (bhāva). A general introduction to the topic can be found in Ruegg 1981, 36. 11 I discuss the topic of doxography, Indian and Classical, at length in Bouthillette 2020. 12 For more details on the MHK manuscript, see Iida 1980, 52–53 and, more recently, Watanabe 1994, 14–15. Concerning the Rañjanā script, see Hartmann 1998. 13 Thus, I disagree with Nicholson’s remarks to the effect that, within Bhāviveka’s MHK, “there is no apparent order in the sequence of doctrines he refutes” (Nicholson 2010, 152). 14 The translation is from Eckel 2008, 105. 15 The brackets here are my own insertion to facilitate the reading. 16 For more information on this chapter, including a critical evaluation, see He 2011. 17 On the vedāntatattvaviniścaya chapter of the MHK, see Qvarnström 2015. 18 Similarly, Eltschinger observed that “according to Sthiramati, the yukti/hetu-vidyā enterprise was expected to shape arguments in order (1) to found and support key Buddhist doctrines, (2) to defeat the non-Buddhist intellectuals’ hostility towards Buddhism; (3) to convert them to the Buddhist saddharma or śāsana, (4) strengthen the coreligionists’ adherence to Buddhism” (Eltschinger 2010, 562). 19 In reference to Nāgārjuna’s MMK, XXIV.9–10: “Those who do not understand the distinction between these two realities do not realize the profound truth embodied in the Buddha’s doctrine. Without relying on conventional [reality], an ultimate [reality] cannot be taught. Without understanding ultimate reality, nirvāṇa is not realized.” The translation is from La Vallée Poussin 1903–13, 494. 20 See Hsu 2013, ii–iii, 61, 143–45. 21 I am correcting here what appears to be a syntactic mistake in Watanabe’s translation. His original line reads as follow: “Paramārtha which means ‘the most excellent’.” Since the reading is odd, I removed the “which.” 22 Its first appearances are noticed as early as the Saṅgītisutta (Dīghanikāya III 219) and the Vibhaṅga (324). 23 On cintāmayīprajñā in Dharmakīrti, see Eltschinger 2010.

Bibliography Ames, William Longstreet. 1986. “Bhāvaviveka’s Prajñāpradīpa: Six Chapters.” PhD diss., Department of Asian Languages and Literature, University of Washington. ———. 1993. “Bhāvaviveka’s Prajñāpradīpa. A Translation of Chapter One: ‘Examination of Causal Conditions’ (Pratyaya).” Journal of Indian Philosophy 21: 209–59. ———. 1999. “Bhāvaviveka’s Prajñāpradīpa: A Translation of Chapters Three, Four, and Five, Examining the āyatanas, Aggregates, and Elements.” Journal of Buddhist Literature 1: 1–119. ———. 2000. “Bhāvaviveka’s Prajñāpradīpa: A Translation of Chapter Six, Examination of Desire and the One Who Desires, and Chapter Seven, Examination of Origination, Duration, and Cessation.” Journal of Buddhist Literature 2: 1–91. ———. 2019. The Lamp of Discernment: A Translation of Chapters 1–12 of Bhāvaviveka’s Prajñāpradīpa. Moraga: Numata Center for Buddhist Translation & Research. Bahulkar, S. S. 1994. “The Madhyamaka-Hṛdaya-Kārikā of Bhāvaviveka: A Photographic Reproduction of Prof. V. V. Gokhale’s Copy.” Nagoya Studies in Indian Culture and Buddhism Saṃbhāṣā 15 (1–4): 1–49. Bouthillette, Karl-Stéphan. 2020. Dialogue and Doxography in Indian Philosophy: Points of View in Buddhist, Jaina, and Advaita Vedānta Traditions. Dialogues in South Asian Traditions: Religion, Philosophy, Literature and History. London and New York: Routledge. Burnouf, Eugène. 1844. Introduction à l’histoire du Bouddhisme Indien. 1876 Reprint ed. Paris: Imprimerie Royale. Davidson, Ronald M. 2002. Indian Esoteric Buddhism: A Social History of the Tantric Movement. New York: Columbia University Press.

357

Karl-Stéphan Bouthillette Eckel, Malcolm David. 1992. To See the Buddha: A Philosopher’s Quest for the Meaning of Emptiness. Princeton: Princeton University Press. ———. 2008. Bhāviveka and His Buddhist Opponents. Harvard Oriental Series. Cambridge, MA: Harvard University Press. Eckel, Malcolm David, and Vincent Eltschinger. 2019. “Bhāviveka.” In Brill’s Encyclopedia of Buddhism: Volume Two, Lives, edited by Jonathan A. Silk, Richard Bowring, Vincent Eltschinger, and Michael Radich, 81–84. Leiden: Brill. Ejima, Yasunori 江島恵教. 1980a. “Bhāvaviveka/Bhavya/Bhāviveka [in Japanese].” Indogaku Bukkyōgaku Kenkyū 印度學佛教學研究 38: 98–106. ———. 1980b. Chūgan shisō no tenkai – Bhāvaviveka kenkyū 中観思想の展開： Bhāvaviveka 研究 [An Exposition of Madhyamaka Thought. A Study of Bhāvaviveka]. Tokyo: Shunjūsha 春秋社. Eltschinger, Vincent. 1998. “Bhāvaviveka et Dharmakīrti sur āgama et contre la Mīmāṁsā.” Asiatische Studien/Études Asiatiques 52: 57–84. ———. 2010. “Studies in Dharmakīrti’s Religious Philosophy: The Cintā-mayī-prajñā.” In Logic and Belief in Indian Philosophy, edited by Piotr Balcerowicz, 553–91. New Delhi: Motilal Banarsidass. ———. 2014. Buddhist Epistemology as Apologetics: Studies on the History, Self-Understanding and Dogmatic Foundations of Late Indian Buddhist Philosophy. Beiträge Zur Kultur Und Geistesgeschichte Asiens. Wien: Ôsterreichisch Akademie Der Wissenschaften. Ganeri, Jonardon. 2001. Philosophy in Classical India: Proper Work of Reason. London and New York: Routledge. Gokhale, V. V. 1958. “The Vedānta-Philosophy Described by Bhavya in the Madhyamakahṛdaya.” IndoIranian Journal 2: 165–80. ———. 1972. “The Second Chapter of Bhavya’s Madhyamakahṛdaya.” Indo-Iranian Journal of Indian Philosophy 14: 40–45. Gokhale, V. V., and S. S. Bahulkar. 1985. “Madhyamakahṛdayakārikā Tarkajvālā, Chapter I.” In Miscellanea Buddhica, edited by Chr. Lindtner, 76–108. Copenhagen: Akademisk Forlag. Halbfass, Wilhelm. 1988. India and Europe: An Essay in Understanding. New York: State University of New York Press. Ham, Hyoung Seok. 2016. “Buddhist Critiques of the Veda and Vedic Sacrifice: A Study of Bhāviveka’s Mīmāṃsā Chapter of the Madhyamakahṛdayakārikā and Tarkajvālā.” PhD diss., Asian Languages and Cultures, University of Michigan. Hartmann, Jens-Uwe. 1998. “The Ranjana Script, Translated from German by Prof. Dr. Ursula Heise.” The Fifth Seal Calligraphic-Icons/Kalligraphikons, The Radheshyam Saraf Art Collection at Hotel Yak & Yeti, Kathmandu, Nepal. He, Huanhuan. 2011. “Bhavya’s Critique of the Vaiśeṣika Theory of Liberation in the Tarkajvālā.” Studies in Indian Philosophy and Buddhism (インド哲学仏教学研究) 18: 23–38. ———. 2014. “Further Notes on Bhāviveka’s Principal Oeuvre.” Indo-Iranian Journal 57: 299–352. Heitmann, Annette L. 1995. “Momentane (ekakṣaṇa) Gnosis ( jñāna) im Sinne der Madhyamakahṛdaya­ kārikā und der in tibetischer Übersetzung vorliegenden Tarkajvālā I – IV.” Asiatische Studien/Études Asiatiques 49: 391–427. ———. 2004. Nektar der Erkenntnis: Buddhistische Philosophie des 6. Jahrhunderts: Bhavyas Tarkajvala I-III. 26. Aachen: Shaker Verlag. Hirabayashi, Jay, and Shotaro Iida. 1978. “Another Look at the Mādhyamika vs. Yogācāra Controversy Concerning Existence and Non-Existence.” Edward Conze Festschrift (Berkeley Buddhist Studies Series) 1: 341–60. Honda, Megumu. 1967. “Sāṃkhya Philosophy Described by His Opponent Bhavya.” Journal of Indian and Buddhist Studies 16 (1): 33–38. Hoornaert, Paul. 1999. “An Annotated Translation of Madhyamakahṛdayakārikā/Tarkajvālā V.1–7.” Kanazawa Daigaku Bungakubu Ronshū Kōdō Kagaku. Tetsugaku hen 金沢大学文学部論集行動科 学・哲学篇 沢大学文学部論集行動科学・哲学篇 [Bulletin of the Faculty of Letters of Kanazawa University, Behavioral Sciences and Philosophy] 19: 127–59. ———. 2001. “An Annotated Translation of Madhyamakahṛdayakārikā/Tarka-jvālā V.27–54.” Kanazawa Daigaku Bungakubu Ronshū Kōdō Kagaku. Tetsugaku hen 金沢大学文学部論集行動科学・ 哲学篇 21: 149–90. ———. 2003. “An Annotated Translation of Madhyamakahṛdayakārikā/Tarka-jvālā V.85–114.” Kanazawa Daigaku Bungakubu Ronshū Kōdō Kagaku. Tetsugaku hen 金沢大学文学部論集行動科学・ 哲学篇 23: 139–70.

358

Bhāviveka ———. 2004. “The Dharmapāla-Bhāvaviveka Debate as Presented in Dharmapāla’s Commentary to Catuḥśataka XVI.23.” Kanazawa Daigaku Bungakubu Ronshū Kōdō Kagaku. Tetsugaku hen 金沢大 学文学部論集行動科学・哲学篇 24: 119–49. Hsu, Chien Yuan. 2011. “Bhāvaviveka’s Syllogism as an Initial Step to Enlightenment (正觀第五 十九 期/二Ο一 一年十二月二十五日).” Satyābhisamaya: A Buddhist Studies Quarterly 59: 57–92. ———. 2013. “Bhāviveka’s Jewel in the Hand Treatise: Elucidating a Path to Awakening Utilizing Formal Inference.” PhD diss., Department of Religious Studies, University of Calgary. Iida, Shotaro. 1966. “Āgama (Scripture) and Yukti (Reason) in Bhāvaviveka.” In Kanakura Kinenronbunshû (Kanakura Festschrift), edited by Kanakura Hakushi Koki Kinen Ronbunshū Kankōkai, Enshō Kanakura, and Yukio Sakamoto, 77–96. Kyoto: Heirakuji Shoten. ———. 1980. Reason and Emptiness: A Study in Logic and Mysticism. Tokyo: Hokuseido Press. Jones, Richard H. 2011. Indian Madhyamaka Buddhist Philosophy After Nagarjuna, Volume I: Plain English Translations and Summaries of the Essential works of Aryadeva, Rahulabhadra, Buddhapalita, and Bhaviveka. New York: Jackson Square Books. Kajiyama, Yuichi. 1968–1969. “Bhāviveka, Sthiramati, and Dharmapāla,” WZKSO 12–13: 193–203. Katsura, Shoryu. 2007. “How Did the Buddhists Prove Something? The Nature of Buddhist Logic.” Pacific World 3 (9): 63–84. Kawasaki, Shinjo. 1973. “Bhavyano tsutaeru Mīmāṃsā shisô (The Mīmāṃsā thought transmitted by Bhavya).” In Indo Shisô to Bukkyô (Nakamura Festschrift), 71–86. Tokyo: Shunjûsha. ———. 1976. “The Mīmāṃsā Chapter of Bhavya’s MHK – Text and Translation (1) Pūrva-pakṣa.” Studies 1976, Institute of Philosophy, The University of Tsukuba: 1–15. Krasser, K. 2011. “How to Teach a Buddhist Monk to Refute the Outsiders – Text Critical Remarks on Some Works by Bhāviveka.” Dhīḥ 51: 49–76. ———. 2012. “Bhāviveka, Dharmakīrti and Kumārila.” In Devadattīyam. Johannes Bronkhorst Felicitation Volume, edited by F. Voegeli et al., 535–94. Bern: Peter Lang. La Vallée Poussin, Louis de, trans. 1903–1913. Mūlamadhyamakakarikas de Nagarjuna avec la Prasannapadā Commentaire de Candrakīrti. Bibliotheca Buddhica, vol. 4. St. Petersburg: l’Académie Impériale des Sciences. ———. 1932–1933. “Madhyamaka, L’auteur du Joyau dans la main; 3, traduction du Joyau dans la main.” MCR 2: 60–138. ———. 1933. “Madhyamaka.” Mélanges Chinois et Bouddhiques II. Lindtner, Chr. 1982. “Adversaria Buddhica.” Wiener Zeitschrift für die Kunde Südasiens 26: 167–94. ———. 1986. “Bhavya, the Logician.” Adyar Library Bulletin 50 (Golden Jubilee Volume): 58–84. ———. 1995. “Bhavya’s Madhyamakahṛdaya (Pariccheda Five) Yogācāratattvaviniścayāvatāra.” Adyar Library Bulletin 59 (C. KunhanRaja Birth Centenary Volume): 37–65. ———. 2001a. Bhavya on Mīmāṃsā, Mīmāṃsātattvanirṇayāvatāraḥ, with English Translation by Christian Lindtner. Adyar: The Adyar Library and Research Centre. ———. 2001b. Madhyamakahrdayam of Bhavya. Adyar: The Adyar Library and Research Center. Miyasaka, Yûshô. 1954a. “Bhavya to Vaiśeṣika gatusetsu (Bhavya and the Vaiśeṣika doctrines).” Shûkyô Kenkyû 133: 173–77. ———. 1954b. “Shôben inyô no Vaiśeṣika tetsugakusetsu (The Vaiśeṣika philosophy quoted by Bhavya).” Bunka 18 (3): 24–40. ———. 1958. “Ronri no honoho ni okeru Vaiśeṣika Tesugaku (Vaiśeṣika philosophy in the Tarkajvālā).” Kôyasan Daigaku Ronsô 1 (1): 1–37. Nakada, Naomichi. 1972. “The Sanskrit Text of the MHK and the Tibetan text of the MHK-vṛtti Tarkajvālā Ṣaṣṭaḥ-paricchedaḥ/Sāṃkhya-tattvāvatāraḥ // (Part 1. Pūrvapakṣa).” Tsurumi Jyoshidaigakubu Kiyô 6: 145–46. Nakamura, Hajime. 1950. “Shoki no Vedānta Tetsu-gaku (The Early Vedānta Philosophy).” Tôkyô: 238–332. ———. 1955. “Upaniṣadic Tradition and the Early School of Vedānta, as Noticed in Buddhist Scripture.” Harvard Journal of Asiatic Studies 18 (1–2): 74–104. ———. 1958. “The Tibetan Text of the Madhyamaka-hṛdaya-vṛtti-tarkajvālā.” Indo-Iranian Journal 2: 181–90. ———. 1968. “Shoki Vedānta-tetsugaku kenkyû no hatten (The Progress Report on the Study of the Early Vedānta.” Journal of Indian and Buddhist Studies 16 (2): 10–22. Nicholson, Andrew J. 2010. Unifying Hinduism: Philosophy and Identity in Indian Intellectual History. New York: Columbia University Press.

359

Karl-Stéphan Bouthillette Potter, Karl H., ed. 2003. Buddhist Philosophy from 350 to 600 A.D. Encyclopedia of Indian Philosophies, vol. 9. New Delhi: Motilal Banarsidas Publishers. Qvarnström, Olle. 1988. “Space and Substance. A Theme in Madhyamaka-Vedānta Polemics.” Studies in Central and East Asian Religions 1: 3–34. ———. 1989. Hindu Philosophy in Buddhist Perspective. The Vedantatattvavinishcaya Chapter of Bhavya’s Madhyamakah. 4 vols. Lund Studies in African and Asian Religions, vol. 4. Lund: Almqvist & Wiksell. ———. 1999. “Haribhadra and the Beginnings of Doxography in India.” In Approaches to Jaina Studies: Philosophy, Logic, Rituals and Symbols, edited by Narendra K. Wagle and Olle Qvarnström, 169–210. Toronto: Center for South Asian Studies. ———. 2012. “Sāṃkhya as Portrayed by Bhāviveka and Haribhadrasūri.” Journal of Indian Philosophy 40 (4): 395–409. ———. 2015. Bhāviveka on Sāṃkhya and Vedānta, The Sāṃkhya and Vedānta Chapters of the Madhyamakahṛdayakārikā and Tarkajvālā. Harvard Oriental Series, vol. 78. Cambridge, MA: Harvard University Press. Ruegg, David Seyfort. 1981. The Literature of the Madhyamaka School of Philosophy in India. Wiesbaden: Otto Harrassowitz. ———. 1990. “On the Authorship of Some Works Ascribed to Bhāvaviveka/Bhavya.” In Earliest Buddhism and Madhyamaka, edited by David Seyfort Ruegg and Lambert Schmithausen, 59–71. Leiden: E. J. Brill. Saitō, Akira 斎藤明. 2005. “Chūgan shinron’ no shōmei to sono seiritsu o meguru shōmondai 『中観心論』 の書名とその成立をめぐる諸問題 [Some Problems with the Title of the ‘Madhyamakahṛdayaśāstra’ and Its Formation].” Indogaku Bukkyōgaku Kenkyū 印度學佛教學研究 53: 832–38. Verardi, Giovanni. 2011. Hardships and Downfall of Buddhism in India. Singapore: ISEAS Publishing. Warder, Anthony Kennedy. 1970. Indian Buddhism. 3rd Revised ed. New Delhi: Motilal Banarsidas. Watanabe, Chikafumi. 1994. “Bhā-viveka (A.D. c. 490–570)’s Madhyamaka-hṛdaya-kārikā, Tattvajñānaiṣanā, verses 137–266, An English Translation and Explanation.” Master of Arts, Department of Asian Studies, University of British Columbia. ———. 1998. “A Translation of the Madhyamakahrdayakarikd with the Tarkajvala III. 137–146J.” Journal of the International Association of Buddhist Studies 21 (1): 125–55. ———. 2012. “Madhyamakahṛdayakārikā III. 147–158.” In Saṁskṛta-sādhutā: Goodness of Sanskrit. Studies in Honour of Professor Ashok N. Aklujkar, edited by Chikafumi Watanabe, Michele Desmarais and Yoshichika Honda, 545–51. New Delhi: D. K. Printworld. Westerhoff, Jan. 2014. “Madhyamaka: Conventional Categories in Madhyamaka Philosophy.” In Categorisation in Indian Philosophy: Thinking Inside the Box, edited by Jessica Frazier, 115–31. Surrey and Burlington: Ashgate. Yamaguchi, Susumu. 1941 [1964]. Bukkhyô ni okeru Mu to U no Tairon. (Buddhist controversies concerning existence and non-existence). Tokyo: Sankibô.

360

21 DHARMAPĀLA A Janus-Faced Interpreter of Yogācāra?1 Ching Keng

Dharmapāla (active in the sixth century)2 was an important Yogācāra scholar whose work exerted a significant influence on the development of Buddhist philosophy in India and in East Asia. Dharmapāla was active between two productive periods of Indian Buddhist thought: the time of Vasubandhu (c. 400–480) and Dignāga (c. 480–540) and the time of Dharmakīrti (c. 600–660). Thus, his works provide insight into Indian Buddhist philosophy during the transitional period of the sixth century. According to the Chinese tradition, there were quite a few interpreters of Yogācāra after Vasubandhu, but most of their works are no longer extant. Dharmapāla is an exception, as are Dignāga and Sthiramati (sixth century). Dharmapāla is also a major figure of Indian Buddhism given the lacuna of logico-epistemological ( pramāṇa) Buddhist texts from the period between Dignāga and Dharmakīrti. Dharmapāla was among the earliest Buddhist scholars to employ Dignāga’s logic (hetuvidyā) in his writings, including applying Dignāga’s logic to the interpretation of root texts as well as to debates with opponents. In addition, Dharmapāla is famous for disagreeing with a certain Mahāyāna master who held a very different interpretation of emptiness (śūnyatā). According to the Chinese interpretation, this opponent was the Madh­ yamaka thinker Bhāviveka (sixth century). In this regard, Dharmapāla and Bhāviveka were probably the first participants in the debates between Madhyamaka and Yogācāra, which have marked so profoundly the shape of Indian Buddhist philosophy and its interpretations in Central and East Asia. Finally, Dharmapāla left a significant impact on the Yogācāra school in East Asia, the so-called School of Dharma-characteristics, known as the Faxiang School (faxiang zong 法 相宗) in China and Hossō School (hossō shū 法相宗) in Japan, which is centered around the Yogācāra teachings of A Treatise for the Establishment of Consciousness-Only (Cheng weishi lun 成唯識論; T1585, henceforth abbreviated as CWSL), compiled by the Chinese pilgrim and erudite scholar monk Xuanzang (玄奘; 602? – 664). Despite his importance, Dharmapāla remains insufficiently studied. This is partly because all the works attributed directly or indirectly to him are preserved only in Chinese translations. Moreover, these texts are extremely challenging to interpret even for scholars with expertise in classical Chinese Buddhist texts. None of his works survive in Sanskrit or Tibetan. This casts doubts on whether he was as influential as claimed by his Chinese interpreters. Another hindrance to our study of Dharmapāla is that he seems to have two very different identities. In the eyes of the Faxiang School, Dharmapāla was an original and creative thinker. He holds distinctive opinions about almost every important issue, according to the commentary on the DOI: 10.4324/9781351030908-32

361

Ching Keng

CWSL by Master Ji (基法師, also known as Kuiji 窺基; 632–682), Xuanzang’s famous disciple. The significance of his positions is manifest by the fact that Master Ji always treats them as definitive and authoritative. As shown in the following, philosophically original and insightful doctrines were attributed to Dharmapāla, such as the strategy for resolving the tension regarding the issue of meta-cognition (svasaṃvitti or svasaṃvedana). This provided a template for defending Yogācāra idealism philosophically, an important aspect that is largely missing in post-Dignāga Buddhist philosophy and also largely neglected in contemporary philosophical discourse. However, judging from the works reliably attributed to him, Dharmapāla actually appears to be a rather conservative interpreter. He does not seem to offer particularly original interpretations or creative doctrines beyond what the root texts seem to already mean. The fact that Dharmapāla has two different identities casts doubt on whether the doctrines attributed to him actually originated with him, which then casts further doubt on the historical image of Dharmapāla constructed by the Faxiang School. Toward the end of this chapter, I will suggest a tentative way to deal with this question of authorship. In addition, I aim to highlight the distinctive features of Dharmapāla’s works and their potential contributions to encourage more research on this insufficiently studied figure.

Dharmapāla’s Life and Works As with almost all figures in Indian Buddhism, we know very little about Dharmapāla’s life.3 The only biography we have is preserved in Chinese sources.4 It indicates that Dharmapāla came from the city of Jianzhibuluo 建志補羅, the capital of Daluopitu 達羅毘荼 (*Draviḍa?) in southern India. He was the son of a high-ranking minister and became a monk to avoid having to marry the king’s daughter. The biography stresses his mastery of Buddhist doctrine and that he died very young. The historical accuracy of this biography is impossible to confirm. Given that Dharmapāla adopted the system of logic devised by Dignāga (c. 480–540) and was referred to by Vinītadeva5 (690–750, see Funayama 2001), we can infer that he must have lived sometime in the late sixth or seventh century. According to Xuanzang’s disciples, Dharmapāla composed four texts, one of which is a lengthy work on grammar that has long been lost (Tillemans 2008, 11–12). The other three, preserved only in Chinese translations, are all commentaries. In my discussion in the following, when I refer to “works reliably attributed to Dharmapāla,” I am referring to these three commentaries, as follows: 1. A Commentary on the Four-Hundred Verses (Dasheng guang bailun shilun 大乘廣百論 釋論; T1571, 10 fascicles; hereafter Four-Hundred Commentary): Dharmapāla’s commentary on the Four-Hundred Verses (Catuḥśataka) of Āryadeva (c. third century). 2. Jewel-Arising Treatise on the Establishment of Consciousness-Only (Cheng weishi baosheng lun 成唯識寶生論; T1591, 5 fascicles; hereafter Twenty Verses Commentary): Dharmapāla’s commentary on Vasubandhu’s Twenty Verses (Viṃśikā). 3. A Commentary on the Investigation of Cognitive Objects (Guan suoyuan lun shi 觀所緣 論釋; T1625, 1 fascicle; hereafter Investigation Commentary): Dharmapāla’s commentary on Dignāga’s Investigation of Cognitive Objects (Ālambanaparīkṣā). The Four-Hundred Commentary was translated by Xuanzang. Compared to the Twenty Verses Commentary and Investigation Commentary, which were both translated by Yijing (義淨; 635–713), it is relatively accessible. Moreover, both the Twenty Verses Commentary and the

362

Dharmapāla

Investigation Commentary are incomplete. Verses 11–15 (the refutation of the Realists’ theory of atoms [ paramāṇu]) and verse 22 (the closing verse) in the Twenty Verses Commentary are missing, as well as verses 7–8 (the last two verses) in the Investigation Commentary.6 In addition to Dharmapāla’s works mentioned previously, the CWSL is a collection of commentaries on Vasubandhu’s Thirty Verses (Triṃśikā) compiled by Xuanzang, in which we find references to doctrines professed by Dignāga, Sthiramati, and Dharmapāla, among others.7 But doubts have been raised: both Lusthaus (2002, Chapter  15 in particular) and Sakuma (2006 [2008]) have challenged Master Ji’s attribution of certain doctrines to particular Indian masters. These doubts are magnified by the fact that in the works reliably attributed to Dharmapāla, he appears to be a rather conservative commentator and interpreter, while the positions Master Ji attributes to him portray him as quite creative and bold. This has led some scholars to suspect that Xuanzang and Master Ji in fact crafted their own ideas in the name of Dharmapāla. As I shall discuss in the following, I agree that there are legitimate doubts about whether any of the doctrines cited by the CWSL represent the actual teachings of Dharmapāla.8

Features of Dharmapāla’s Works A common method employed by Dharmapāla and other Indian authors in their commentaries is to lay out carefully the content of an argument and explain why the author of the root text (that is, the work being commented upon) phrases his argument in the way that he does. Dharmapāla shares with other Indian commentators a widespread principle of Buddhist and Indian scholastic exegesis: every word in the root texts counts, and not a single word is unimportant or redundant.9 Another feature that Dharmapāla shares with other Indian commentators is that he is rather conservative in his commentaries on the root texts. Often, he does not go much beyond the philosophical arguments made in the original text. In his Twenty Verses Commentary, for instance, Dharmapāla simply repeats what has already been said in Vasubandhu’s Twenty Verses, with some re-arrangement of the arguments for the sake of refuting different opponents (see following discussion). A distinctive feature that really sets Dharmapāla apart from other commentators is that he often tries to make sense of Yogācāra texts by formulating the argument in terms of Dignāga’s logic. A clear example can be seen in Dharmapāla’s reading of the first verse of the Twenty Verses Commentary,10 which reads: This world is consciousness-only (vijñaptimātra) Because of the appearance (avabhāsana) of non-existent external objects As when someone with a timira disease sees non-existent nets of hair and the like. (T31.1591.79c12–14)11 Dharmapāla takes this statement to be a well-formulated inference (anumāna) proposed by Vasubandhu in order to establish the thesis ( pratijñā) of the entire Twenty Verses. Dignāga’s logic requires that a well-formulated inference consist of three parts: a thesis, which itself consists of a subject (dharmin) and a predicate (dharma); a reason (hetu); and an example (dṛṣṭānta).12 According to Dharmapāla, Vasubandhu’s inference can be formulated as follows: Thesis: cognitions that have matter (rūpa), etc., as their cognitive object (ālmabana) do not cognize actual external objects;

363

Ching Keng

Reason: because the cognitions merely have the appearance (avabhāsana) of nonexistent objects; Example: like a person with a timira disease who sees a net of hair. In the following, I discuss why Dharmapāla feels the need to construe this verse as establishing an inference in Dignāga’s sense. Dignāga, except in the Investigation of Cognitive Objects, rarely employs his own logical paradigm and epistemology to prove a specific Yogācāra doctrine. Dharmapāla, on the other hand, is one of the first Buddhist philosophers to apply Dignāga’s logic extensively to interpret Yogācāra texts. Similarly, his younger contemporary, Bhāviveka, uses Dignāga’s logical system in his interpretations of Madhyamaka texts. In fact, this is a movement that can be observed in the period following Dignāga, in which Buddhist commentators started to use the logical tools Dignāga elaborated in the service of their own particular positions. It effectively amounted to a marriage of epistemology ( pramāṇa) and Yogācāra, in the case of Dharmapāla, and of epistemology and Madhyamaka, in the case of Bhāviveka.13

What We Can Learn From Dharmapāla’s Works Although Dharmapāla’s commentaries may not be very original, much can be learned about the evolution of Indian Buddhist philosophy from them. Here, I highlight three areas in which Dharmapāla’s commentaries illuminate significant philosophical developments in his era.

Post-Vasubandhu Developments: How to Win a Debate First, Dharmapāla’s commentaries illustrate the ways that views about how to win a debate changed in the period between Vasubandhu and Dharmapāla. As mentioned earlier, Dharmapāla maintains that Vasubandhu’s Twenty Verses expresses a thesis in the first verse. In contrast, Kellner and Taber (2014) insist that in the Twenty Verses, Vasubandhu simply refutes the objections of his opponents without providing a positive thesis. Namely, having refuted the validity of all realist objections (including those of both the Nyāya-Vaiśeṣika and the Sarvāstivāda), Vasubandhu implicitly resorts to an “argument from ignorance” in order to reach the conclusion that things external to mind do not exist. That is to say, he first demonstrates that arguments establishing the existence of external objects fail. From this absence of sound argument, he then concludes that there can be no rational basis for believing in the existence of external objects. Dharmapāla, by contrast, maintains that Vasubandhu lays out a positive thesis beyond this “argument from ignorance.” According to Hu (2018, 120–21), Dharmapāla’s main reason for insisting that Vasubandhu has a thesis is that he sees a simple refutation of the opponents’ argument as inadequate for establishing one’s own position. Rather, both the refutation of the opponent’s thesis and the establishment of one’s own thesis are necessary. In the Investigation Commentary, Dharmapāla says, “It is not the case that one’s own thesis can be established simply by pointing out the fallacy in the opponent’s thesis” (T31.1625.889c27–28). And in the Four-Hundred Commentary, he says that one must equally establish one’s own thesis and refute the opponent’s thesis (T30.1571.243b8–11). In the Investigation of Cognitive Objects, Dignāga at least implicitly thinks both are necessary,14 as does Vinītadeva in his commentary on the Investigation of Cognitive Objects.15 One of Dharmapāla’s reasons for thinking that both establishment and refutation are ne­cessary is that pointing out an opponent’s fallacy is insufficient: while it demonstrates that the opponent’s thesis is not necessarily right, it does not prove that it is definitely wrong.16 For this reason, establishing one’s own thesis via a three-partite argument is also necessary for 364

Dharmapāla

proving that one’s thesis is right and hence the opponent’s thesis is wrong (see the Investigation Commentary, T31.1625.890a3–5). If Kellner and Taber’s reading of the Twenty Verses is correct, it would imply that, for Vasubandhu, simply dismissing the opponent’s challenges on the basis of an implicit resort to the argument from ignorance is sufficient for establishing one’s own thesis. For Dharmapāla, this is not the case. In light of the different strategies adopted by Vasubandhu and Dharmapāla, we may suspect that the mode of constructing argumentation underwent dramatic changes in the period between Vasubandhu and Dharmapāla, particularly with respect to whether refutation of an opponent’s objection is sufficient for establishing one’s own thesis.

Non-Buddhist Indian Philosophy at the Time of Dharmapāla Dharmapāla’s commentaries also contain much information about non-Buddhist schools. As a conservative interpreter, Dharmapāla primarily repeats and rearranges the root texts upon which he is commenting. This rearrangement of an argument may not be philosophically significant, but it can be very historically informative. For example, according to the report by Xuanzang’s disciple, Wengui (文軌, c. seventh century), Dharmapāla, in Chapter One of his Four-Hundred Commentary, refutes in turn the atomist theories defended by three opponents, namely the Vaiśeṣikas, the pre-Vasubandhu17 Vaibhāṣikas, and the post-Vasubandhu Vaibhāṣikas (see Keng 2019). Here, Dharmapāla’s rearrangement of Vasubandhu’s argument can provide useful information about the philosophical debates that emerged between Vasubandhu’s intellectual milieu and Dharmapāla’s. Several times in his Four-Hundred Commentary, Dharmapāla cites and harshly criticizes non-Buddhist as well as Buddhist philosophical schools. He critiques non-Buddhists (tīrthika), the six masters, and then more specifically Vaiśeṣika, Sāṃkhya, Lokāyata or Cārvāka, Nihilism, Jainism, and Grammarians. Dharmapāla also criticizes the mainstream Buddhist schools such as Sarvāstivāda and Sautrāntika, but, interestingly, he refers to them less frequently than to non-Buddhist schools.18 In this respect, Dharmapāla resembles his predecessor Dignāga and his younger contemporary Bhāviveka to the extent that they all engage in dialogue and debate with several contemporary Indian Buddhist and non-Buddhist schools in their philosophical treatises. This indicates the existence of a sort of “neutral” philosophical platform where philosophers from different traditions and backgrounds were able to meet and argue.

Dharmapāla’s Engagement With Bhāviveka and Madhyamaka The Four-Hundred Commentary offers for Dharmapāla an opportunity to stage an important debate with an opponent about the reality of the three natures (tri-svabhāva) (T30.1571.247b22– 249b3).19 The Yogācāra theory of three natures holds that our phenomenal world can be examined from three different perspectives: the imagined nature, dependent nature, and perfected nature. The imagined nature ( parikalpita-svabhāva) refers to reified concepts such as “cups,” which are generic names or misconceptualizations superimposed upon particulars (svalakṣaṇa) that defy names and concepts. The dependent nature ( paratantra-svabhāva) refers to those particulars to the extent that they arise and cease in accordance with the principle of depen­dent origination ( pratītyasamutpāda). Most Yogācāra texts subscribe to a kind of metaphysical idealism and treat particulars as merely mental representations (nimitta or ākāra) rather than things external to the mind. Finally, the perfected nature ( pariniṣpanna-svabhāva) refers to the aforementioned principle of dependent origination, usually named thusness (tathatā) in the sense that it is “the way things are,” and it governs the dependent nature of arising and ceasing 365

Ching Keng

particulars. The Yogācāra position is that the imagined nature is not real, but the dependent nature and the perfected nature are real. The dependent nature is real because Yogācāra holds the foundationalist view that, while a cup as a generic and conceptualized name (the imagined nature) is not real, there must be something real (the dependent nature) – that is, the mental representation as a particular – that serves as the foundation for the name “a cup.” The perfected nature is real because it refers to the principle of dependent origination according to which the mis-named mental representation arises and ceases. In the commentary itself, Dharmapāla does not mention the name of his opponent, but Xuanzang’s disciple Wŏnch’ŭk (圓測; 613–696) identifies this opponent as Bhāviveka (Qingbian 清辯) (X21.369.338b4–339b1, T33.1708.360b5–361a28). The opponent claims that the dependent nature does not ultimately exist, but Dharmapāla maintains that it does. Dharmapāla refutes the opponent as follows: In many places where the Tathāgata talks about the three natures, he always claims that the imagined nature does not exist, but the dependent nature and the perfected nature do. Hence, we know that when the Tathāgata teaches emptiness, his teachings about emptiness have a special intent. One must not deny the existence of all dharmas by taking at face value the words of the Tathāgata. If people understand his intent by taking his words at face value, then they would be criticizing the Mahāyāna. (T30.1571.248a29 – b3; see also Keenan 1997, 111) If people deny the existence of the dependent nature, then they are denying the existence of both defiled and undefiled dharmas. They should be named as holding a wrong view about emptiness (durgṛhītā śūnyatā) and doing harm to both their own position and the position they are refuting. (T30.1571.248c5–6; see also Keenan 1997, 114) Here Dharmapāla’s point is to insist that the dependent nature and the perfected nature are not empty (literally meaning “zero,” and hence “non-existence”). If people insist that all three natures are empty, then they would endorse nihilism. Moreover, if the dependent nature is empty, then both defiled and undefiled dharmas would have no foundation, and hence neither defiled dharmas (i.e., misconceptions such as “a cup”) nor undefiled dharmas (to see a mental representation as it really is, i.e., devoid of the misconception of “a cup”) would exist, and Buddhist cultivation would have no point. Hence, Dharmapāla insists, the doctrine of emptiness preached by the Buddha must not be understood as the non-existence of all three natures but as the non-existence of merely the imagined nature. For Dharmapāla, the imagined nature does not exist at all, but, for his opponent Bhāviveka, ordinary things do exist at the conventional level. For Dharmapāla, the dependent nature and the perfected nature ultimately exist. In contrast, Bhāviveka accepts the existence of the dependent nature only at the conventional level (Madhyamakahṛdaya verse V.71 and commentary, Eckel 2008, 274) but strongly denies the ultimate existence of unconditioned dharmas such as thusness or emptiness.20 There is much to say about this debate. First, many major Yogācāra texts assert that, at the stage of buddhahood, the dependent nature ceases.21 This implies that it is not ultimately real. Interestingly, this is Wŏnch’ŭk’s basis for thinking that Bhāviveka is close to Paramārtha since they both assert that dependent nature is ultimately unreal (T33.1708.360b11–16). That is to say, the dependent nature must be eliminated in buddhahood. In contrast, CWSL, following the Treatise on the Scripture of the Buddhaland (Fodi jing lun 佛地經論; *Buddhabhūmivyākhyā)22 366

Dharmapāla

(T26.1530.302c1–3), says that when one becomes a buddha, a dramatic transformation (the so-called “transformation of the basis” [āśraya-parivṛtti or āśraya-parāvṛtti]) takes place so that the previous defiled function of ordinary consciousness is transformed into an undefiled cognitive function (T31.1585.56b2–3). Thus, according to the CWSL, when Yogācāra texts mention the elimination of the dependent nature, they do not mean to eliminate the dependent nature as a whole but simply to eliminate the defiled aspect of the dependent nature (i.e., the cognitive function derived from the so-called “storehouse consciousness” [ālayavijñāna], which is a sort of unconscious repository of previous actions, defilements, and tendencies). The undefiled cognitive function subsists even in buddhahood. From philosophical and “buddhological” perspectives, the issue of whether the dependent nature ultimately ceases concerns how we conceive of the mind of a buddha. If the dependent nature ceases after the attainment of buddhahood, a buddha will be left with no mind because all cognitive functions belong to the dependent nature. This would mean that the activity of a buddha’s saving sentient beings would not imply that there is any cognitive function and would resemble the activity of what Mark Siderits calls a “Robo-Buddha” (Siderits 2011). In contrast, if dependent nature does not cease, a buddha will continue having a mind. The only difference between ordinary people and a buddha would be that ordinary minds always involve false discrimination (vikalpa), which generates afflictions, whereas a buddha’s mind does not. This debate is important because it points to a major difference between Dharmapāla and Bhāviveka on the one hand and Dharmapāla and his Yogācāra predecessors (Paramārtha, for example) on the other.23 This disagreement between Dharmapāla and Bhāviveka is probably the first substantial debate between Madhyamaka and Yogācāra in the history of Indian Buddhism. It is preceded only by brief critical remarks in Yogācāra texts, including the disparagement in the Stages of Yogācāra (Yogācārabhūmi) of “those who misunderstand emptiness (durgṛhītā śūnyatā)” (Wogihara 1971, 47; T30.1579.488c22–28) and the denigration of Madhyamaka implied by the passage in the Scripture on Unpacking the Profound Meaning (Saṃdhinirmocana-sūtra) on the three turnings of the dharma wheel (Lamotte 1935, 85–86, 206–7). Both passages are quite brief and do not explain exactly how Yogācāra differs from Madhyamaka. Apart from this debate, Dharmapāla criticizes Bhāviveka in his commentary on the FourHundred Verses by Āryadeva, allegedly a Madhyamaka master. This suggests that around the time of Dharmapāla, the boundary between Madhyamaka and Yogācāra was more fluid than many later Buddhist thinkers and contemporary scholars generally take it to be. In fact, I would suggest that we consider Dharmapāla’s engagement with Bhāviveka the beginning of the divide between Madhyamaka and Yogācāra.24

The Cognition of Self-Cognition25 As mentioned, some doctrines in the CWSL are attributed to Dharmapāla by Xuanzang’s disciples. Among these, some of the most distinctive doctrines do not appear in any of the commentarial works attributed to Dharmapāla. Hence, scholars have expressed doubts whether they are authentic teachings of Dharmapāla. However, despite uncertainty about their authenticity, these doctrines can tell us something about Indian Buddhist philosophy in the sixth and seventh centuries. One such example can be picked from CWSL’s discussion about meta-cognition, that is, about how to explain our ability to (sometimes) report on the occurrence of a cognition. The Yogācāra tradition maintains that cognition consists of two aspects (*bhāga, ākāra, ābhāsa, rūpa; fen 分): the image (nimitta) aspect and the seeing (dṛṣṭi or darśana) aspect. This 367

Ching Keng

conception is probably derived from the Sautrāntika representationalism, according to which we cognize external objects indirectly through mental images (ākāra). This idea is only one step away from the Yogācāra idealism, which claims that mental images originate not from external objects but from the internal processes of consciousness. According to the account given by the CWSL, it was Dignāga who developed this twoaspect theory into a theory of three aspects in his Corpus of the Valid Means of Cognition (Pramāṇasamuccaya).26 In addition to the image and the seeing aspects, there is an aspect of self-cognition (svasaṃvedana or svasaṃvitti; zizheng fen自證分), namely the cognition that the seeing aspect is cognizing the image aspect, for example, the awareness of seeing blueness.27 Somewhat surprisingly, the CWSL further maintains that cognition consists of four aspects, not three (CWSL, T31.1585.10b17–28; Treatise on the Scripture of the Buddhaland, T26.1530.303b13–20). The reason for establishing a fourth aspect, the cognition of self-cognition (zhengzizheng fen 證自證分; *svasaṃvitti-saṃvitti), is that since self-cognition cognizes and hence testifies to the functioning of the seeing aspect, another aspect is required to cognize and testify to the functioning of self-cognition. In order to avoid the problem of infinite regress, the CWSL claims that self-cognition, the third aspect, plays a special role: it cognizes and testifies to the functioning of both the second and the fourth aspects. Admittedly, this way out of the difficulty of infinite regress may sound too easy to be persuasive. We can still ask: if the cognition of self-cognition (fourth aspect) can be cognized by the self-cognition (third aspect) and hence a fifth one is in no need, then why not make it that the self-cognition (third aspect) can both cognize and be cognized by the seeing aspect (second aspect) and hence do away with the cognition of self-cognition (fourth aspect)?28 In what follows, I aim not so much to defend this position as to contextualize it in the debates between two models for meta-cognition in Abhidharma. Master Ji’s commentary on the CWSL attributes the four-aspect theory to Dharmapāla (T43.1830.242a24–26 and T43.1830.320c20–22). But the theory of four aspects does not appear in any of the commentarial works attributed to Dharmapāla. Instead, it is found in the Treatise on the Scripture of the Buddhaland, which is attributed to multiple authors, headed by *Bandhuprabha (Qinguang 親光, dates unknown), who seems to be a contemporary of Dharmapāla (X50.819.284a12–13). The aforementioned passage from the CWSL appears verbatim in the Treatise on the Scripture of the Buddhaland. This shows that, at least according to Master Ji, Dharmapāla and *Bandhuprabha maintain similar doctrinal positions. In a few other places, Master Ji also groups Dharmapāla and *Bandhuprabha together in contrast to other masters (T43.1830.230b19–26; T45.1861.252c8–253a5; X48.806.377a11 – b4). One possible explanation is that this theory originated not from Dharmapāla himself but from the group headed by Dharmapāla, which also includes *Bandhuprabha and other authors of the Treatise on the Scripture of the Buddhaland. So when Master Ji attributes this theory to Dharmapāla, he is actually referring to the group headed by Dharmapāla rather than to Dharmapāla himself. Another possibility is that given the fact that this four-aspect theory is not attested in any Indian sources, it is likely a Chinese invention (see Funayama 2020). Currently, we have no way to confirm either explanation. However, regardless of its origin, this theory can provide useful information about the evolution of Indian Buddhist philosophy. As was pointed out by Yao (2005), in Abhidharma, there were already two models for meta-cognition. According to Sarvāstivāda Abhidharma, meta-cognition cannot be achieved by the cognition itself. A cognitive event can become the object of meta-cognition (e.g., the awareness of cognizing blueness) only by means of another cognitive event. In other words, a cognition can only be cognized by a subsequent cognition to achieve meta-cognition. Thus, meta-cognition is similar to memory: it is a later cognition that brings an earlier cognition into 368

Dharmapāla

awareness. The typical metaphor provided to support this idea is that a knife does not cut itself, and thus another knife is necessary to cut the first knife. In contrast, the Mahāsāṃghika school of Abhidharma claims that meta-cognition is achieved not by a subsequent cognition because every cognition is self-cognizing. After Dharmakīrti, the Mahāsāṃghika model became the mainstream position (Yao 2005: 149). The four-aspect theory can be interpreted as an interesting and ingenious synthesis of the two different models proposed by the Sarvāstivāda and Mahāsāṃghika schools. As the Treatise on the Scripture of the Buddhaland concludes, Based on such reasoning, we conclude that cognition, despite being one, consists of multiple aspects. These aspects are neither identical nor separate from each other. Taken together, they cognize both what is inside and outside of cognition and free from the fallacy of infinite regress. (T26.1530.303b18–20) This means that each cognition is itself also meta-cognition. But cognition consists of four aspects, each being made aware by another aspect of the cognition. From the viewpoint of each individual cognition, this theory adopts the Mahāsāṃghika model, namely that each cognition is also meta-cognition. But, from the viewpoint of each aspect within a cognition, it follows the Sarvāstivāda model; that is, each aspect is cognized by another aspect. In my opinion, this theory clearly aims at reconciling the two conflicting models. Thus, whether this four-aspect theory was formulated by Dharmapāla or not, it shows that there was a tension between Sarvāstivāda and Mahāsāṃghika in their understandings of cognition and that creative solutions were in need. Dignāga’s theory of self-awareness (svasaṃvedana) differs from contemporary theories in the sense that for Dignāga, any cognition must at least implicitly include meta-cognition of itself. Dignāga argues, based on memory, that if a cognition does not include meta-cognition of itself, then at a later moment, one can never possibly recall that experience of the cognition in question.29 But some contemporary scholars think “beliefs and desires can be activated unconsciously” (Carruthers and Gennaro 2020, 5). That is, there are mental states that we are not aware of. Only those states of which we are aware are conscious states (Carruthers and Gennaro 2020, 5). The distinction between unconscious and conscious mental states also motivates, at least partly, the so-called higher-order theories of consciousness, of which we find a counterpart in Dharmapāla’s theory of four aspects. Higher-order theories in general hold that a mental state is not in itself conscious. It is another mental state, that is, the higher-order mental state, that makes conscious the first-order mental state. There are three major types of higher-order theories. The higher-order perception (HOP) theory claims that it is another perception (achieved by an inner sense scanning the first-order cognition) that makes the first-order mental state conscious. The higher-order thought (HOT) theory claims that it is a conceptualized belief that makes the first-order mental state conscious, and there is some sort of causal power (either actual or dispositional) of the first-order mental state that brings up the higher-order thought. What is common to these two theories is that one of the first-order or the higher-order mental states can occur without the other, despite some causal connection between the two. In contrast, the third theory, the selfrepresentational higher-order theory, holds that the first-order mental state possesses in itself a higher-order intentional content that makes itself conscious.30 Due to the limitation of space, I cannot review various versions of the higher-order theories and their respective strength and weakness. But I would like to draw attention to the similarities 369

Ching Keng

between Dharmapāla’s theory and Peter C. Carruthers’s “dual-content theory,” a subtype of the self-representational higher-order theory (Carruthers and Gennaro 2020, 32).31 These two are similar in two aspects. First, both theories hold that it is one and the same mental state that is both first-order and higher-order (Carruthers and Gennaro 2020, 31). Despite Dharmapāla’s claims that there are four aspects, some first-order and some higher-order, all four aspects taken together constitute one single cognitive event. Second, in contrast to the part-whole self-representational account (another subtype of the self-representational higher-order theory according to which a conscious perception is always partially conceptual), Carruthers insists that the higher-order intentional content, that is, a “seeming of the first-order content of the state,” can be entirely non-conceptual (Carruthers and Gennaro 2020, 35). This agrees with both Dignāga and Dharmapāla, who hold that self-cognition together with the cognition of self-cognition must be non-conceptual. The question of self-cognition is one of multiple areas in which Dharmapāla’s thought can engage contemporary philosophical debates.

A Solution to the Problem of the Janus-Faced Image of Dharmapāla We now have a Janus-faced image of Dharmapāla: the commentarial works reliably attributed to him make him appear quite conservative, but the theories attributed to him in the CWSL depict him as an original thinker. To make matters more complicated, there is little evidence to suggest that the positions cited in the CWSL are actually Dharmapāla’s. Given the conflicting evidence, I  suggest that the attributions of several positions to Dharmapāla in the CWSL indicate that they were held by members of Dharmapāla’s lineage, including *Bandhuprabha and Xuanzang, but not by Dharmapāla himself alone. Even if the opinions attributed to Dharmapāla were actually Xuanzang’s or Master Ji’s, as suggested by Funayama and others, they preserve invaluable information about theories of cognition in sixth- and seventh-century India. Further comparison of doctrines attributed to Dharmapāla by Master Ji with other extant sources will help us better judge the reliability of the attributions.32

Dharmapāla’s Legacy in East Asia After Xuanzang returned to China in 643, he established a scholarly lineage that was later labeled School of Dharma-Characteristics in China and Japan. In this tradition, the CWSL was treated as the definitive doctrinal text. But Xuanzang’s lineage began to decline in the Tang dynasty, and some of the main texts, including Master Ji’s commentary on the CWSL, were lost in China. After that, interest in the CWSL largely receded, except for a brief revival in the late Ming dynasty.33 In the history of Chinese Buddhism, the Indian masters Dharmapāla and Paramārtha have long been considered to hold opposing views. For Dharmapāla, the mind (here understood as the cognitive function originating from the seeds in the storehouse consciousness) is defiled in nature, but for Paramārtha, the mind is fundamentally pure. Disputes arose surrounding these different views that were expressed in terms of the opposition between the School of Dharma-Characteristics (faxiang zong 法相宗 or simply xiang zong 相宗) and the School of Dharma-Nature (faxing zong 法性宗 or simply xing zong 性宗). In the late Ming dynasty, when interest in the CWSL revived, there were attempts to reconcile these two schools. The disagreement concerning which school represents the authentic teachings of Indian Yogācāra, Dharmapāla/School of Dharma-Characteristics or Paramārtha/School of Dharma-Nature, reappeared in early twentieth-century Chinese Buddhism. Regarding this millennium-long debate, I have shown (Keng 2022) that the doctrinal source behind the School of Dharma-Nature was 370

Dharmapāla

not Paramārtha himself but the Chinese reinterpretation of Paramārtha’s original teachings through the lens of the Awakening of Faith in Mahāyāna (Dasheng qixin lun 大乘起信論; T1666). In reality, the doctrinal divide between Dharmapāla and Paramārtha may not have been as substantial as the tradition holds it to be.

Conclusion We thus have a bifurcated image of Dharmapāla: his commentarial works make him appear rather conservative, but the positions ascribed to him in the CWSL make him seem quite innovative. I have suggested that a tentative way to make sense of this bifurcation is to treat Dharmapāla not as a single person but as a group composed of Dharmapāla and his students and followers. Furthermore, even if these doctrines were Chinese inventions falsely attributed to Dharmapāla, they still provide useful information about the dynamics at play within Indian Buddhist philosophy in the sixth and seventh centuries. The study of Dharmapāla contributes to a better understanding of Indian Buddhist philosophy in at least the following aspects. First, Dharmapāla shows how to employ Dignāga’s logic to defend Yogācāra philosophy, a strategy we find neither in pre-Dignāga Yogācāra thinkers and only rarely in Dignāga. Second, Dharampāla’s reference to and criticism of his contemporary Indian schools, both non-Buddhist and Buddhist, preserve invaluable information about Indian philosophy around the sixth century. Third, Dharmapāla’s debate against Bhāviveka provides a perspective about the relation between Yogācāra and Madhyamaka on the one hand and between early Yogācāra and later Yogācāra on the other. Finally, even if some doctrines were suspiciously and retroactively attributed to Dharmapāla, they might still teach us about significant issues lying beneath the surface of Indian Buddhist philosophy around the sixth century, such as the tension between the Sarvāstivāda and the Mahāsāṃghika models for meta-cognition. Finally, I close with encouragement for further research on Dharmapāla. Based on what I have shown previously, it would not be an exaggeration to say that without adequate studies of Dharmapāla, we will remain far from reaching a clear picture of Indian Buddhist philosophy during the sixth and seventh centuries, a crucial transitional period between Vasubandhu and Dignāga on the one hand and Bhāviveka, Candrakīrti, and Dharmakīrti on the other.

Notes 1 I am grateful to Mark Siderits, Funayama Tōru, Michael Radich, Tony Cheng, Lok-Chi Chan, and Chih-Chiang Hu for their very useful comments and suggestions. 2 Most sources give 530–61 as Dharmapāla’s exact dates, but Funayama (2000) has doubts about this dating. It would be safe to simply say that he was active in the sixth century. 3 Tillemans (2008, 8 ff.) gave a detailed description of Dharmapāla’s life and works. For a more recent yet brief overview, see Moriyama (2019). 4 They are mostly brief and scattered in various Chinese sources. Here I rely on the brief biography from the Biography of Xuanzang entitled The Biography of the Tripiṭaka Master of the Great Ci’en Temple of the Great Tang dynasty (大唐大慈恩寺三藏法師傳); see T50.2053.241c11–24. 5 Cross-references to Vinītadeva can be extremely helpful for studying Dharmapāla. As mentioned in the following, the three works reliably attributed to Dharmapāla are all commentaries. It should be noted that Vinītadeva also composed commentaries on two of the same texts, namely Vasuban­ dhu’s Twenty Verses and Dignāga’s Investigation of Cognitive Objects. The Japanese translations of Vinītadeva’s two commentaries can be found in Yamaguchi and Nozawa (1953). Recently Duckworth et al. (2016, 78–104) provide an English translation of Vinītadeva’s commentary on the Investigation of Cognitive Objects. In addition, Yamaguchi and Nozawa (1953) have translated Sthiramati’s

371

Ching Keng commentaries on Vasubandhu’s Thirty Verses into Japanese. An examination of Sthiramati’s commentary is necessary to ascertain the accuracy of Master Ji’s ascription of certain doctrines to Sthiramati and Dharmapāla in the CWSL.   It is interesting to note that Vinītadeva refers to Dharmapāla’s commentary and observes that Dharmapāla’s interpretation “might be a correct understanding, but his idea is too profound for me to comprehend” (Duckworth et al. 2016, 80). In contrast, Vinītadeva says that the goal of his commentary on the Investigation of Cognitive Objects is merely “to delight beginners” (ibid., 103). Despite this difference, Yamaguchi and Nozawa (1953, 418) remark that the two masters generally agree. 6 As for contemporary scholarship on the Four-Hundred Commentary, Tillemans (2008 [1990]) includes an excellent study and annotated English translation of Chapter 4 (Refutation of Heretical Views) and Chapter 5 (Refutation of the Sense-Organs and Their Objects). Keenan (1997) provides an English translation of Chapter 10, which contains an important debate between Dharmapāla and his opponent (probably Bhāviveka, see subsequent discussion); also see Hoornaert (2004), which, however, is not a complete translation of the chapter. There is no English translation of the Twenty Verses Commentary. Ui (1963) offers a kundoku 訓讀 (Japanese way of reading Chinese texts) with footnotes. Similarly, no English translation of the Investigation Commentary is available. And, again, Ui (1951) provides a kundoku with footnotes. 7 Traditionally, the CWSL was said to be compiled by Xuanzang, who collected from the ten commentaries on the Thirty Verses. Recently, Yamabe (2020) argues that this genre of collecting the views of various masters into one volume may have already existed before Xuanzang’s time, which suggests that the CWSL might have been based on a single Indian original source. 8 As for contemporary studies of the CWSL, the best available is still the French annotated translation by Louis de la Vallée Poussin (1928–29), which was completely translated into English by Gelong Lodrö Sangpo and Gelongma Migme Chödrön (2017). Francis Cook (1999) provides another complete English translation. 9 For example, in a few places in his commentaries, Dharmapāla tries to answer the objection that a particular statement in Dignāga’s root text is pointless (anartha) or serves no purpose (e.g., T31.1625.890a13–15). 10 There are doubts about whether these first lines are verse or prose. In addition, Kellner and Taber (2014, 735–36) argue that this is neither the first verse nor an inference. My assertion that Dharmapāla takes this to be a well-formulated inference owes much to Hu (2018, 16 ff.). 11 Unless otherwise cited, all translations are mine. 12 For more details, please see the chapter on Dignāga by Kei Kataoka in this volume. 13 For Bhāviveka and the culture of debate, see Eckel (2008, 9ff.). 14 That is, Dignāga refutes three of the realist opponent’s theories in verses 1–5 and then proposes his own theory in verses 6–8. 15 Vinītadeva thinks that Dignāga’s argument consists of two parts: “It refutes an external view of the percept, and it affirms an internal one” (Duckworth et al. 2016, 52). 16 For example, if we show that the opponent’s argument commits the fallacy of indeterminacy (anaikāntika; buding 不定), this simply shows that the reason does not necessarily lead to the conclusion, not that the opponent’s thesis is definitely wrong. 17 To be more precise, “pre-Vasubandhu” and “post-Vasubandhu” mean before and after the text Abhidharmakośabhāṣya. 18 More work should be done to examine what the major philosophical disagreements are between Dharmapāla and his opponents, whether Dharmapāla’s references to his opponents are faithful, and how Dharmpāla’s refutation agrees or differs from those found in Bhāviveka and his other contemporary Buddhist scholars. 19 For more details about the three natures, please see the chapter on Vasubandhu by Jonathan C. Gold in this volume. 20 An example of this can be found in the inference proposed by Bhāviveka in his Jewel in the Hand (*Mahāyāna-hastaratna-śāstra): “From the ultimate perspective ( paramārthatas), conditioned dharmas are empty, like illusion, because they originate in dependence [on something else]. Unconditioned dharmas do not exist because they do not arise, like flowers in the sky” (T30.1578.268b21–22;). I cannot go into detail here. Readers should consult recent studies by Keenan and Eckel. For a brief discussion, see Moriyama (2019, 170). 21 Such texts include Madhyāntavibhāga I.4, III.9; Mahāyānasaṃgraha II.2; and Triṃśikā 5a. For the Mahāyānasūtrālaṃkārabhāṣya XIX.51 on the destruction of the dependent nature, see D’Amato (2005, 197).

372

Dharmapāla 22 No Sanskrit survives for this text, unfortunately. For information about its authors, see subsequently under the section “The Cognition of Self-Cognition.” 23 I would like to suggest that the Yogācāra thinkers living before Dharmapāla, including Asaṅga and Vasubandhu, all side with Bhāviveka’s view that the dependent nature does not ultimately exist; that is, it does not exist as the buddha’s cognitive function. But there is no space here to prove this point. 24 For example, Tāranātha (1575–1634) once remarked, “Before the arrival of these two masters (Buddhapālita and Bhāviveka), all adherents of the Mahāyāna remained under the same teaching” (Eckel 2008, 11). 25 Here I follow Funayama (2020) and translate the Chinese term zheng zizheng fen 證自證分 and the reconstructed Sanskrit term *svasaṃvitti-saṃvitti as “cognition of self-cognition.” The issue I am dealing with in this section is better known among Anglophone Buddhologists under the designation “self-awareness.” I did not choose the more popular translation “self-awareness” for svasaṃvitti or “awareness of self-awareness” for *svasaṃvitti-saṃvitti because it has now become standard to use self-awareness to refer to Dignāga’s specific way of explaining how meta-cognition is possible. According to Dignāga, meta-cognition is achieved without the mediation of any higher-order cognition. But, as shown subsequently, for Dharmapāla, meta-cognition is achieved through a combination of first-order and higher-order mental states. 26 As far as I am aware, most scholars in the Indian epistemological tradition ( pramāṇa) do not agree that, for Dignāga, self-cognition refers to some third aspect that cognizes and testifies to the functioning of the seeing aspect. In their interpretation, self-cognition simply means that any cognition, which consists of two aspects, is reflexive or self-cognizing in nature. See footnote 28. Thanks to Mark Siderits for reminding me to make clarifications on this point. 27 For Dignāga’s establishment of the two aspects and the reasons behind it, see the chapter on direct perception ( pratyakṣa) in the Corpus of the Valid Means of Cognition, especially verses I.11–12. Note that Dignāga’s terminology is different: he uses the term svābhāsa for the seeing aspect and arthābhāsa for the image aspect. The CWSL’s interpretation of Dignāga as holding a theory of three aspects may have been based on Dignāga’s distinction among “what is being measured” ( prameya), “the means for measuring” ( pramāṇa), and “the result of measuring ( pramāṇa-phala)” in verse I.9– 10. Given the Buddhist notion of no self, self-cognition should be formulated as “awareness of seeing blueness” rather than “awareness that I am seeing blueness.” 28 Note that the CWSL does offer an explanation for why the second aspect does not cognize the third aspect, and hence the fourth aspect is necessary. The reason is that what cognizes and testifies to the function of the self-cognition must function in the manner of direct perception (xianliang 現量; pratyakṣa), but the second aspect does not always function in that manner. See T31.1585.10b20–22. Indeed, the post-Dignāga Indian epistemologists, Dharmakīrti in particular, did take this approach to claim that self-cognition does not refer to any aspect within a cognition but simply means that any cognition, which consists of two aspects, is reflexive or self-cognizing in nature. A cognition does not become aware because of a separate or higher-order cognition. To borrow the stock example, cognition is self-illuminating, just like a lamp illuminates the objects as well as itself. For a brief summary of the views about meta-cognition from Dignāga and Dharmakīrti, see section five, “Buddhist Accounts of Meta-Cognition,” in the general introduction of Siderits, Keng, and Spackman (2020). 29 See Corpus of the Valid Means of Cognition, verse I.11d (Hattori 1968, 30). For a brief explanation of Dignāga’s memory argument for self-awareness, see section five, “Buddhist Accounts of MetaCognition,” in the general introduction of Siderits, Keng, and Spackman (2020). 30 For a survey of the higher-order theories, see Carruthers and Gennaro (2020). 31 Here I aim only to draw attention to the similar aspects between Dharmapāla’s and Carruthers’ views. I don’t mean to suggest that these two are similar in every aspect. For example, Carruthers holds that a state’s being conscious is a dispositional property, but Dharmapāla would not accept this. Again, many thanks to Mark Siderits for drawing my attention to this difference between Carruthers and Dharmapāla. 32 This is precisely the path taken by Sakuma (2006 [2008]), where he provides an excellent comparison between the doctrines attributed to Dharmapāla in the CWSL and Sthiramati’s position regarding three issues: the correspondences between the buddhas’ four kinds of cognition and ordinary sentient beings’ eight kinds of consciousness; the correspondences between the buddhas’ four kinds of cognition and the buddhas’ three bodies; and the formation of the five-gotra system. 33 For a very brief description of the reception of Yogācāra Buddhism in China and its revival in the late Ming dynasty, see Duckworth et al. 2016, xviii–xx.

373

Ching Keng

Bibliography T: Taishō shinshū Daizōkyō 大正新脩大藏經 (Tōkyō: Daizōkyō kankō kai, 1924–35). X: Shinsan Dai Nihon zoku Zōkyō 新纂大日本續藏經 (Tōkyō: Kokusho kankō kai, 1975–1989). Carruthers, Peter, and Rocco Gennaro. 2020. “Higher-Order Theories of Consciousness.” In The Stanford Encyclopedia of Philosophy,  Fall 2020 ed., edited by Edward N. Zalta. https://plato.stanford.edu/ archives/fall2020/entries/consciousness-higher/. Cook, Francis. 1999. Three Texts on Consciousness Only. Berkeley: Numata Center. D’Amato, Mario. 2005. “Three Natures, Three Stages: An Interpretation of the Yogācāra TrisvabhāvaTheory.” Journal of Indian Philosophy 33: 185–207. de La Vallée Poussin, Louis. 1928–29. Vijñaptimātratāsiddhi: La siddhi de Hiuan-Tsang. Paris: Paul Geuthner. Duckworth, Douglas, Malcolm David Eckel, Jay L. Garfield, John Powers, Yeshes Thabkhas, and Sonam Thakchoe. 2016. Dignāga’s Investigation of the Percept: A Philosophical Legacy in India and Tibet. New York: Oxford University Press. Eckel, Malcolm D. 2008. Bhāviveka and His Buddhist Opponents. Cambridge: Harvard University Press. Funayama, Tōru 船山 徹. 2000. “Two Notes on Dharmapāla and Dharmakīrti.” Zinbun 35: 1–11. ———. 2001. “On the Date of Vinītadeva.” In Le Parole e i Marmi, edited by Raffaele Torella, 309–25. Roma: Istituto Italiano per l’Africa e l’Oriente. ———. 2020. “The Genesis of *Svasaṃvitti-saṃvitti Reconsidered.” In Buddhist Philosophy of Consciousness: Tradition and Dialogue, edited by Mark Siderits, Ching Keng and John Spackman, 209–24. Leiden and Boston: Brill-Rodopi. Hoornaert, Paul. 2004. “The Dharmapāla-Bhāvaviveka Debate as Presented in Dharmapāla’s Com­ mentary to Catuḥśataka XVI.23.” Kanezawa daigaku bungakubu ronshū: Kōdō kagaku tetsugaku hen 金沢大学文学部論集. 行動科学‧哲学篇 24: 119–49. Hu, Chi-Chiang 胡志強. 2018. “Zhijue, tazhe yu luoji: Hufa Cheng weishi baosheng lun zhi zhexue yanjiu 知覺、他者與邏輯：護法《成唯識寶生論》之哲學研究” (“Perception, Others, and Logic: A Philosophical Study on Dharmapāla’s Cheng weishi baosheng lun”). PhD diss., National Chengchi University, Taiwan. Keenan, John P.  1997. Dharmapāla’s Yogācāra Critique of Bhāvaviveka’s Mādhyamika Explanation of Emptiness: The Tenth Chapter of Ta-Ch’eng Kuang Pai-Lun Shih Commenting on Āryadeva’s Catuḥśataka Chapter Sixteen. Lewiston: Edwin Mellen Press. Kellner, Birgit, and John Taber. 2014. “Studies in Yogācāra-Vijñānavāda Idealism I: The Interpretation of Vasubandhu’s Viṃśikā.” Asiatische Studien/Études Asiatiques 68 (3): 709–56. Keng, Ching. 2019. “Three Senses of Atomic Accumulation – An Interpretation of Vasubandhu’s Viṃsika Stanzas 12–13 in Light of the Abhidharmakosabhaṣya and Dharmapala’s Dasheng Guangbailun Shilun.” Journal of Indian Philosophy 47 (3): 565–601. ———. 2022. Toward a New Image of Paramartha: Yogacara and Tathagatagarbha Buddhism Revisited. London, UK; New York, NY: Bloomsbury Academic. Lamotte, Étienne. 1935. Samdhinirmocana sutra, l’explication des mystères: texte tibétain édité et traduit par Étienne Lamotte. Louvain: Bureaux du recueil Bibliothèque de l’Universite. Lusthaus, Dan. 2002. Buddhist Phenomenology: A Philosophical Investigation of Yogācāra Buddhism and the Ch’eng Wei-shih lun. London and New York: RoutledgeCurzon. Moriyama, Shinya. 2019. “Dharmapāla.” In Brill’s Encyclopedia of Buddhism, vol. II, 168–72. Leiden and Boston: Brill. Sakuma, Hidenori. 2006 (2008). “On Doctrinal Similarities Between Sthiramati and Xuanzang.” Journal of the International Association of Buddhist Studies 29 (2): 357–82. Sangpo, Gelong Lodrö, and Gelongma Migme Chödrön, with the assistance of Alexander Leonhard Mayer, trans. 2017. Vijñapti-mātratā-siddhi: a commentary (Cheng weishi lun) on Vasubandhu’s Triṃśikā by Xuanzang, translated from Chinese into French and Annotated by Louis de La Vallée Poussin. New Delhi: Motilal Banarsidass Publishers. Siderits, Mark. 2011. “Buddhas as Zombies: A Buddhist Reduction of Subjectivity.” In Self, No Self? Perspectives from Analytical, Phenomenological, & Indian Traditions, edited by Mark Siderits, Evan Thompson, and Dan Zahavi. Oxford: Oxford University Press. Siderits, Mark, Ching Keng, and John Spackman. 2020. Buddhist Philosophy of Consciousness: Tradition and Dialogue. Leiden and Boston: Brill-Rodopi. Tillemans, Tom J. F. 2008. Materials for the Study of Āryadeva, Dharmapāla, and Candrakīrti: The Catuḥśataka of Āryadeva, Chapters XII and XIII with the Commentaries of Dharmapāla and

374

Dharmapāla Candrakīrti: Introduction, Translation, Notes, Sanskrit, Tibetan and Chinese Texts, Indexes. New Delhi: Motilal Banarsidass Publishers. Ui, Hakuju 宇井伯壽. 1951. Jinna chosaku no kenkyū 陳那著作の研究 (Studies of the Works of Dignāga). Tokyo: Iwanami Bookstore. ———. 1963. “Jō yuishiki hōshōron kenkyū 成唯識寶生論研究 (A Study of the Cheng weishi baosheng lun)”. In Daijō butten no kenkyū 大乘佛典の研究 (Studies of Mahāyāna Buddhist Texts), 607–811. Tokyo: Iwanami Bookstore. Wogihara, Unrai 荻原雲來. 1971. Bodhisattvabhūmi; A Statement of Whole Course of the Bodhisattva (being fifteenth section of Yogācārabhūmi). Tōkyō: Sankibō. Yamabe, Nobuyoshi 山部能宜. 2020. “A Hypothetical Reconsideration of the ‘Compilation’ of Cheng Weishi Lun.” In From Chang’an to Nālandā: The Life and Legacy of the Chinese Buddhist Monk Xuanzang (602? – 664), edited by Shi Ciguang, Chen Jinhua, Ji Yun, and Shi Xingding. Singapore: World Scholastic Publishers. Yamaguchi, Susumu 山口益, and Jōshō Nozawa 野沢静証. 1953. Seishin yuishiki no genten kaimei 世 親唯識の原典解明 (Explication of the Yogācāra Primary Texts by Vasubandhu). Kyoto: Hōzōkan. Yao, Zhihua 姚治華. 2005. The Buddhist Theory of Self-Cognition. London and New York: Routledge.

375

22 STHIRAMATI A Yogācāra Commentator and Innovator Roy Tzohar and Jowita Kramer

Sthiramati1 (circa sixth century CE) was an Indian Buddhist scholar of the Yogācāra, best known for his extensive commentaries on some of the most important treatises of this tradition. What often goes less noticed, however, is his role in the very creation of that tradition, a function of his particular position within the school’s textual development.2 Relative to previous thinkers like Asaṅga and Vasubandhu, Sthiramati operated under a much more defined notion of the Yogācāra as a distinct school, or at least a more defined textual tradition; accordingly, his interpretive challenge – and contribution – consisted in synthesizing a varied textual corpus into a coherent and consistent worldview, adding to it in the process some original and strikingly innovative insights. In this chapter, we seek to provide the necessary background to his writings and life and to present a brief survey and evaluation of his philosophical contributions and their significance. The first section, authored by Jowita Kramer, presents the state of current research on what is known regarding his life, dates, and authorship and aims to point out some common philosophical and stylistic traits among the works ascribed to him by providing a comparative and systematic consideration of his intertextual realm. The concluding section, authored by Roy Tzohar, shows Sthiramati as an individual thinker by examining, as a case study, his lengthy comments on the opening verse of Vasubandhu’s Treatise in Thirty Verses (Triṃśikā), highlighting their innovativeness and unique contribution to Yogācāra thought.

Sthiramati’s Life, Authorship, and Intertextual Realm Most of the information about Sthiramati’s life and work available to us today comes from Chinese and Tibetan sources whose historicity is uncertain.3 A datable Indian reference to the name Sthiramati can be found in an inscription of King Guhasena (r. ca. 553–569) from Valabhī in western India, in which a monastery built by an abbot with this name is mentioned (Bühler 1877, 9 and plate II, lines 3–4; Silk 2009, 384–85). In an inscription of Śīlāditya III (661/2), Sthiramati is also said to have founded a monastery in Valabhī.4 Furthermore, the Sanskrit colophons of the Commentary on the Treatise on the Five Constituents (Pañcaskandhakavibhāṣā) (Kramer 2014a, 133), the Commentary on the Distinguishing of the Middle from the Extremes (Madhyāntavibhāgaṭīkā) (Yamaguchi 1934, 165), the Commentary on the Treatise on Mental Presentation in Thirty Verses (Triṃśikāvijñaptibhāṣya) (Buescher 376

DOI: 10.4324/9781351030908-33

Sthiramati

2007, 142), and the Commentary on the Treasury of Abhidharma [Called] True Reality (Tattvārthā Abhidharmakośaṭīkā) (Matsuda 2013, 49) name Sthiramati as the author of these commentaries. However, these do not allow us to determine his date (Kramer 2019, 456f.). Based on the Valabhī inscription of Guhasena and on the assumption that Sthiramati and Guhasena were contemporaries, Erich Frauwallner proposed Sthiramati’s lifetime as 510–570 (Frauwallner 1961, 136–37). However, taking into account that the inscription states only that the monastery was erected by Sthiramati (and not for Sthiramati), we cannot know with certainty that Sthiramati was still alive at the time of Guhasena’s donation. Some scholars have dated Sthiramati to 470–550 (Nguyen 1990, 21) because, according to Chinese sources, one of the commentaries attributed to Sthiramati in the colophon to its corresponding Tibetan translation, namely the Da Baoji jinglun (大寶積經論, T. 1523), corresponding to the Commentary on the Kāśyapa Chapter (*Kāśyapaparivartaṭīkā), was translated into Chinese already between 508 and 535 (by Bodhiruci) (Silk 2009, 382). It has also been suggested in modern studies that there may have been more than one author named Sthiramati or that works that have actually been composed by an author (or authors) with a different name over time have (erroneously) been assigned to one single author, namely Sthiramati (Silk 2009, 385; Silk 2015, 149; Kramer 2016a, 71). All together, seventeen works have been ascribed to (an author called) Sthiramati. They are extant in Sanskrit, Tibetan, Chinese, Uighur, and/or Mongolian and cover a wide range of topics. The ten commentaries among them refer to both Yogācāra and non-Yogācāra root texts (see also Kramer 2019, 457). These commentaries include 1. the Explication of the Compendium of the Abhidharma (Abhidharmasamuccayavyākhyā; available in Sanskrit, Tibetan, and Chinese) 2. the Commentary on the Treasury of Abhidharma [Called] True Reality (Tattvārthā Abhidharmakośaṭīkā; Sanskrit [chapter 3 and parts of chapters 2 and 4 missing], Tibetan, Chinese [fragments], Uighur [fragments], Mongolian) 3. the Commentary on the Treatise on the Five Constituents (Pañcaskandhakavibhāṣā; Sanskrit, Tibetan) 4. the Commentary on the Treatise on Mental Presentation in Thirty Verses (Triṃśikāvijñaptibhāṣya; Sanskrit, Tibetan) 5. the Commentary on the Distinguishing of the Middle from the Extremes (Madhyāntavibhāgaṭīkā; Sanskrit, Tibetan) 6. the Commentary on the Comments to the Ornament of [Mahāyāna] Sūtras (*Sūtrālaṃkāravṛttibhāṣya; Tibetan) 7. the Commentary on the Kāśyapa Chapter (*Kāśyapaparivartaṭīkā; Tibetan, Chinese, perhaps more correctly *Ratnakūṭasūtraṭīkā?) 8. the Commentary on the Mahāyāna Madhyamaka (Dasheng zhongguan shilun 大乘中觀 釋論; Chinese) 9. the Commentary on the Exposition of Akṣayamati (*Akṣayamatinirdeśaṭīkā; Tibetan; Braarvig 1993, cxxiii and cxxviii) 10. a commentary on the Collection of Means of Knowledge (Pramanasamuccaya; lost; Matsuda 2013, 52). The remaining seven works associated with the name Sthiramati seem to belong to the Tantric genre and are only mentioned in Tibetan sources (Cordier 1909, 138ff., 151 and 158; Kramer 2016a, 49). Sthiramati’s work has received far less attention in modern studies than the writings of other Yogācāra authors, such as Asaṅga or Vasubandhu, most likely because he is perceived as a 377

Roy Tzohar and Jowita Kramer

commentator and not as an original author in his own right. As shown in Kramer 2012, 2013a, 2013b, 2016a, 2016b, this negligence is unjustified, as some of the commentaries that are regarded as Sthiramati’s main works, such as the Commentary on the Treatise on the Five Constituents and the Commentary on the Treatise on Mental Presentation in Thirty Verses, clearly contain creative and innovative thought (as will be seen in the second part of this chapter). It is also important to note that the creative energy of commentators such as Sthiramati manifests not only in the creation of entirely new ideas but also in the reorganization of previous teachings (Kramer 2015, 327). We may assume with some certainty that the tantric works ascribed to Sthiramati were composed by a different author than the philosophical commentaries. At the same time, the Commentary on the Treatise on the Five Constituents, the Commentary on the Treatise on Mental Presentation in Thirty Verses, and the Commentary on the Treasury of Abhidharma [Called] True Reality appear very closely related and were probably composed by the same person. The three texts share many parallel passages, which make their interdependence obvious. For instance, a whole section on the mental (caitasika) factors of the Commentary on the Treatise on the Five Constituents is reused in the Commentary on the Treatise on Mental Presentation in Thirty Verses.5 In the Commentary on the Treasury of Abhidharma [Called] True Reality, Sthiramati provides an extensive commentary on the Commentary on the Treasury of Abhidharma (Abhidharmakośabhāṣya). Kazunobu Matsuda and other scholars working on the edition of the Sanskrit manuscript of the text have found a great number of parallels between the Commentary on the Treatise on the Five Constituents and the first chapter of the Commentary on the Treasury of Abhidharma [Called] True Reality, showing the close relationship of the two texts.6 At the same time, the Commentary on the Treasury of Abhidharma [Called] True Reality generally follows the traditional Abhidharmic system without referring to specific Yogācāra terminology and doctrines. However, the exploration of this text, the Sanskrit original of which came to light only recently, is still in its infancy; thus, at the present time, no definitive statements about its nature and its philosophy can be made.7 In contrast, the Commentary on the Treatise on the Five Constituents represents a comprehensive compendium of Yogācāra Abhidharma, discussing topics such as the “store mind” (ālayavijñāna) extensively. By means of the Treatise on the Five Constituents (Pañcaskandhaka), Vasubandhu produced a handy manual on the five constituents of the person (skandha) as viewed from the perspective of the Yogācāras. According to Sthiramati, the reasons for the conciseness of the work, covering merely seven manuscript folios, are to be found in the intention to meet the needs of its potential readers: the householders who do not have enough time to read extensive treatises because of their various duties and the contemplating monks who should not be distracted by reading lengthy works (Kramer 2014a, xx). In contrast to the briefness of Vasubandhu’s work, Sthiramati produced an extensive commentary ten times the length of the root-text. The Commentary on the Treatise on the Five Constituents may be subdivided into seven main parts: (1) introductory remarks; (2) explanation of the five constituents; (3) explanation of the twelve realms (āyatana); (4) explanation of the eighteen elements (dhātu); (5) reasons for explaining the constituents, realms, and elements; (6) list (mātṛkā) of qualities; (7) concluding matter. The main and longest section of the text is the second one, in which the five constituents (skandha) are described and analyzed in detail. The two most notable and important parts of this section are the explanations of “matter” (rūpa) and of “perception” (vijñāna). Although Sthiramati mostly does not explicitly mention his sources in his comments on matter, his positions are closely related to the earlier Yogācāra texts, the Commentary on the Treasury of Abhidharma and the Compendium of the Abhidharma (Abhidharmasamuccaya).8 At the same time, he occasionally develops his own views, for instance, when characterizing objects of the sense faculties, such as sounds and “information” (vijñapti), that is, karmic action that is 378

Sthiramati

perceivable by others.9 Another site of Sthiramati’s innovation is his discussion of whether there are any other types of matter that are comparable to “non-information” (avijñapti), an invisible and penetrable kind of matter related to karmic action that does not reveal itself to others. Here, he puts forth ideas that do not appear to be mentioned in the texts upon which he is commenting, and the way he discusses this problem indicates that the boundary between material and mental factors was controversial among the authors of Abhidharmic texts. Especially Yogācāra authors seem to have had some difficulties in distinguishing material from mental entities, when they tried to categorize objects which appeared to be “borderline cases” between matter and mind, such as the matter of “non-information” (avijñapti) and the matter of atoms or visualized images. The reason for these difficulties among the Yogācāras in particular might be found in the fact that it was in this milieu that the existential status of matter weakened increasingly in favor of the position of mind – a condition which finally resulted in the evolving of the theory that all external phenomena exist only as mental images (vijñaptimātratā). It is notable that Sthiramati rejects a view that clearly has its origin in the Compendium of the Abhidharma, namely the division of the invisible and penetrable kind of matter into five categories. These five are classified in the Compendium of the Abhidharma as belonging to the Abhidharmic category of “factors” (dharma), which are the object of the mental faculty (manas), that is, can only be perceived by thinking. The fivefold categorization of the Compendium of the Abhidharma includes compressed (ābhisaṃkṣepika) matter, matter of space (ābhyavakāśika), matter of commitment (sāmādānika), imagined ( parikalpita) matter, and matter produced by those with meditative power (vaibhūtvika). Sthiramati opposes the view that the categories other than avijñapti (appearing as “matter of commitment” [sāmādānika] in the Compendium of the Abhidharma) are to be classified as material. According to Sthiramati, the first two kinds, the compressed matter and the matter of space, are both matter of atoms and thus included in “color,” which is part of visible matter and an object of the faculty of sight. The objects of the last two categories, the imagined matter and the matter produced by those with meditative power, are nothing other than mental images and are therefore to be regarded as being part of the mind (and not matter). Remarkably, Sthiramati seems to generally follow a position here that in the Commentary on the Treasury of Abhidharma is ascribed to the Vaibhāṣika-Sarvāstivāda tradition. According to the latter, only the non-information (avijñapti) is to be regarded as invisible and penetrable matter. The contrary view, namely that meditative images are to be considered this kind of matter, too, was held by some early Yogācāras (as represented by some portions of the Foundation of Yoga Practitioners [Yogācārabhūmi]) and the “Sautrāntikas” (Kramer 2013b, 93), a group within the Sarvāstivāda tradition known for frequently opposing the positions of the Vaibhāṣikas. However, it is also important to note that, unlike the Vaibhāṣikas, Sthiramati rejects the view that non-information exists as a real entity (dravya). At the same time, he points out that this nonexistence of non-information as a real entity does not lead to the nonexistence of the various types of restraint (saṃvara). The concepts of restraint and non-information are inextricably linked in the Abhidharmic context. Therefore one could (wrongly) conclude that Sthiramati’s rejection of non-information as a real entity means that restraint is not really existent. The background of this statement is probably the discussion between the Vaibhāṣikas, who object the claim that non-information cannot really exist, and the “Sautrāntikas,” who defend their view that non-information cannot be a real entity. Notably, Sthiramati’s explanation of his own view partly relies on the assumption of the “store mind” (ālayavijñāna) and thus goes beyond Vasubandhu’s line of reasoning. Sthiramati understands the concept of “restraint” as an intention to restrain oneself from committing wrong deeds. This intention is produced at the time when the ordination ceremony takes place, imprinting a seed in the store mind, which constitutes the basis for future intentions of the same kind (see Kramer 2013b, 93). 379

Roy Tzohar and Jowita Kramer

Also in contrast to the Compendium of the Abhidharma, Sthiramati, in a few instances, even incorporates into the Commentary on the Treatise on the Five Constituents some aspects of the position “that everything is a (mental) presentation only” (vijñaptimātratā) – for instance, when he mentions that color, from the viewpoint of the highest reality ( paramārthatas), is not the object of the faculty of sight, because perception (vijñāna) does not have any external objects. According to Sthiramati, from the viewpoint of highest reality, the existence of color atoms, like that of shape, is not possible.10 The greatest part of the section on perception (vijnāna) of the Commentary on the Treatise on the Five Constituents is devoted to a description of the “store mind” (ālayavijnāna). Sthiramati comments extensively on the three characteristics, mentioned already in the Treatise on the Five Constituents, which mark the differences between the store mind and the actual perception ( pravṛttivijñāna) (Kramer 2014b, 315): 1. The object of the ālayavijnāna and the mode in which it is apprehended is not clearly determined (aparicchinnālambanākāra). 2. The ālayavijnāna is of one kind (ekajātīya). 3. The continuity of the ālayavijnāna is not interrupted (santānānuvrtti). Sthiramati moreover explains in detail the four arguments (also already stated by Vasubandhu) that make the existence of a store mind besides the mental stream of continuously fluctuating sense perceptions necessary.11 These four proofs are of particular interest because they differ from the eight proofs presented in the Yogācārabhūmi and the Commentary on the Compendium of the Abhidharma (Abhidharmasamuccayabhāṣya), as well as from the arguments provided in the Summary of the Mahāyāna (Mahāyānasaṃgraha).12 Of particular interest is also the Commentary on the Treatise on the Five Constituents’ explanation of the process beginning in the moment of actually perceiving an object until the moment in which intention (cetanā) leaves an imprint (vāsanā) in the store mind (Kramer 2016b, 152ff.). The Yogācāra concept of the “afflicted notion [of ‘I am’]” (kliṣṭaṃ manas) is treated more briefly in the Commentary on the Treatise on the Five Constituents than the store mind (ālayavijnāna). It is notable that neither Vasubandhu nor Sthiramati provide any proofs for the afflicted notion of “I am” comparable to those for the store mind. It appears likely that the same author who composed the Commentary on the Treatise on the Five Constituents and the Commentary on the Treatise on Mental Presentation in Thirty Verses also produced the Explication of the Compendium of the Abhidharma. This commentary on the Compendium of the Abhidharma is a compilation of the root text and another commentary, namely the Commentary on the Compendium of the Abhidharma. Neither of the colophons of the Sanskrit versions of the Commentary on the Compendium of the Abhidharma and the Explication of the Compendium of the Abhidharma mention the author. However, both commentaries are ascribed to *Jinaputra in Tibetan sources. At the same time, the Chinese tradition attributes the compilation of the Explication of the Compendium of the Abhidharma to Sthiramati. The idea that Sthiramati was involved in the production of the Explication of the Compendium of the Abhidharma appears possible, insofar as the Commentary on the Compendium of the Abhidharma (or the Explication of the Compendium of the Abhidharma) plays a central role in his Commentary on the Treatise on the Five Constituents (Kramer 2015, 295–98). However, it is not very probable that he composed the Commentary on the Compendium of the Abhidharma itself, as this commentary seems to be older than the Commentary on the Treatise on the Five Constituents (see Kramer 2016a, 50). The same person who authored the Commentary on the Treasury of Abhidharma [Called] True Reality (as well as the Commentary on the Treatise on the Five Constituents and the Commentary on the Treatise on Mental Presentation in Thirty Verses) may have also composed a no 380

Sthiramati

longer extant commentary on the Collection of Means of Knowledge. This has been suggested by Kazunobu Matsuda on the basis of a remark to be found in the Commentary on the Treasury of Abhidharma [Called] True Reality (Matsuda 2013, 52). There the reader is referred to a commentary on the Collection of Means of Knowledge (Pramāṇasamuccaya-upanibandha). A similar term, namely Pañcaskandhaka-upanibandha, is used in the Commentary on the Treatise on Mental Presentation in Thirty Verses when Sthiramati refers the reader to his Commentary on the Treatise on the Five Constituents. Therefore, Matsuda suggests understanding Pramāṇasamuccaya-upanibandha in a parallel way, that is, in the sense of a commentary on the Collection of Means of Knowledge composed by Sthiramati himself. In general, the author of the Commentary on the Treatise on the Five Constituents and the Commentary on the Treatise on Mental Presentation in Thirty Verses appears to be a welleducated and innovative scholar. He very often offers philosophically complex explanations of high quality and not just definitions of individual terms. His comments are concise and not repetitive and rarely contain redundant sentences and phrases. Unlike the author of the Commentary on the Comments to the Ornament of [Mahāyāna] Sūtras, he hardly gives any examples for the concepts he is explaining. As the commentarial style of the Commentary on the Comments to the Ornament of [Mahāyāna] Sūtras differs strongly from that of the Commentary on the Treatise on the Five Constituents and the Commentary on the Treatise on Mental Presentation in Thirty Verses, it appears doubtful that the Commentary on the Comments to the Ornament of [Mahāyāna] Sūtras was composed by the same author. The explanations of the Commentary on the Comments to the Ornament of [Mahāyāna] Sūtras, a rather strict word-by-word commentary, are philosophically less independent and elaborate, are often much shorter than the comments in the Commentary on the Treatise on the Five Constituents and the Commentary on the Treatise on Mental Presentation in Thirty Verses, and are partly repetitive. The author of the Commentary on the Comments to the Ornament of [Mahāyāna] Sūtras appears less creative and inspired than the commentator of the other two texts. The authorship of the Commentary on the Kāśyapa Chapter and the Commentary on the Mahāyāna Madhyamaka remains doubtful and requires further examination of the contents and style of these texts.13 The fact that comments on the Kāśyapa Chapter (Kaśyapaparivarta) that are found in Sthiramati’s Commentary on the Distinguishing of the Middle from the Extremes differ strongly from parallel explanations in the Commentary on the Kāśyapa Chapter as well as the Commentary on the Kāśyapa Chapter‘s rather early translation into Chinese (between 508 and 535 by Bodhiruci) raise doubts that the same author wrote both texts (see also Kramer 2016a, 50). The Commentary on the Exposition of Akṣayamati (composed by Vasubandhu, according to the Tibetan tradition) has been attributed to Sthiramati by Jens Braarvig on the basis of some passages in the Commentary on the Exposition of Akṣayamati that have parallels in parts of the Commentary on the Comments to the Ornament of [Mahāyāna] Sūtras, the Commentary on the Treatise on the Five Constituents, and the Commentary on the Treatise on Mental Presentation in Thirty Verses (Braarvig 1993, cxxiii and cxxviii). The parallels referred to by Braarvig all deal with explanations of the mental (caitasika) factors, which indeed seem to be related in the three commentaries. We do not know with certainty whether this fact proves that the Commentary on the Exposition of Akṣayamati and the two other commentaries were composed by the same person (since the possibility cannot be ruled out that the author of the former simply copied the definitions from the latter).

Innovation and Original Contribution The overwhelmingly commentarial nature of Sthiramati’s work makes pointing out his contribution as an independent thinker – beyond mere local innovations – something of a challenge; 381

Roy Tzohar and Jowita Kramer

yet, as I hope to demonstrate in this section, it is an undertaking well worth pursuing. His reliance on a great number of sources, from which he quotes often and freely; his habit of juxtaposing a spectrum of views (often without stating the preferred one); and a general Indian scholastic inclination to tone down innovation in the name of tradition – all these make it hard to pinpoint his own views and indeed give the impression of works that in their greater part are meant to explicate rather than innovate.14 And yet a close reading of his oeuvre, one that is attuned to the accumulative and dispersed manner in which his ideas appear as the commentarial genre dictates, reveals not just striking and innovative insights but also a substantive and profound underlying philosophical outlook. One distinct display of this outlook and of Sthiramati’s innovativeness as a thinker is found in his commentary on the opening verse of Vasubandhu’s Treatise in Thirty Verses (Triṃśikā). Here, arguably, Sthiramati presents what may be seen as a broadly conceived theory of meaning – one that is applicable, in his words, “both in the world and in texts” – according to which all language use is figurative (upacāra). The term upacāra, it should be noted, is mentioned by Vasubandhu in the Triṃśikā only once. Sthiramati’s disproportionally lengthy commentary on this topic indicates the significance he accords to this issue but also suggests that he seems to see this as an opportunity to set the philosophical framework for venturing into the Triṃśikā‘s unfolding claims and, by extension, for the Yogācāra worldview. To achieve this, Sthiramati incorporates ideas introduced by his predecessors and contemporaries, both Buddhists and non-Buddhists, into a novel philosophical synthesis that ties together the Yogācāra understanding of the causal activity of consciousness with a linguistic theory of sense. Space does not allow me to delve into the question of his verified and possible sources of influence for making these claims, as well as into the textual and philological support for their interpretation (I have dealt with these extensively elsewhere,15 and the reader is referred to these sources in the notes). Instead, what follows is more of a speed-tour through the essentials of Sthiramati’s arguments, with an eye to the extent and nature of their innovativeness and their ramifications for and overall contribution to Yogācāra thought. Vasubandhu’s opening verse of the Triṃśikā is as follows: The metaphor (upacāra) of ‘things’ (dharma) and of ‘self’ in its various forms, which is set in motion, [Sthiramati: “that is to say in the world and in treatises”], that is with reference to the transformation of consciousness. (Buescher 2007, 43)16 Sthiramati’s ensuing commentary does many kinds of philosophical work, much of it venturing beyond the focus or scope of the root verse. Put briefly, the commentary connects a rather formulaic scholastic understanding of metaphor – under a more general discussion of theories of meaning – with the Yogācāra model of the activity of consciousness, that is, its “transformation,” and then utilizes this model to refute two “extreme views” (one referring apparently to the Madhyamaka), culminating with a view of all language use as figurative. Sthiramati begins his commentary by clarifying that the verse uses the terms “self ” and “things” (“dharma”) as generic terms for all substantive entities (Buescher 2007, 40:4–8; Tzohar 2018, 157 n. 6). But why are these referred to as “metaphors”? His explanation points out that the “self ” and “things” are mere mental appearances (nirbhāsa), understood under the Yogācāra view by which all phenomenal objects do not exist as they appear and can be known or discussed solely in terms of mental representations (vijñapti). The commentary clarifies that whereas in ordinary language, we understand the words “self ” and “things” to refer to objectively existent phenomena, in fact we can only actually relate to the constructed mental 382

Sthiramati

appearances of a self and of things (which appear as if they were objectively and externally existent; the paradigmatic analogy here is to an eye disease, timira, that causes one to see nonexistent objects). These mental appearances, Sthiramati further explains, are an outcome of the transformation of consciousness (vijñānapariṇāma), defined by him as a causal mental development, identical to the Buddhist notion of dependent arising ( pratītyasamutpāda; in the following, I provide a more detailed explanation of this term).17 Sthiramati then proceeds to explain what renders the words “self ” or “things” metaphors. For this purpose, he uses a formulaic śāstric definition of metaphor according to which a word is used figuratively when it indicates “something that is not there,” or, in more formal terms, when a word’s primary referent is absent from the locus of reference.18 So, in the example used by Sthiramati (Figure 22.1, right-hand side), when a person of the Bāhīkas is called an “ox” (because he is similarly strong, slow witted, etc.), the word “ox” directly denotes its primary referent (an ox, whether as an individual or an implicit generic property), which is, however, absent from the locus of reference (i.e., from that which the word actually describes: the Bāhīka person); therefore, the word is said to designate the Bāhīka person only in a secondary, figurative way. Accordingly, the same applies to the terms “self ” and “things” (Figure 22.1, left-hand side): they are figurative because their primary referents (the self, things) are absent from their locus of reference, identified here as the “transformation of consciousness.” However, unlike in the śāstric stock example, “self ” and “things” – like “ox,” or any other word for that matter – are always mere figures, since according to the Yogācāra, they never exist in the manner in which they appear (being mere mental appearances) and hence are always absent from their locus of reference. Furthermore, because the self and things, we are told, stand for the entire range of subjective and objective phenomena, all of which do not exist as they appear, it follows – as Sthiramati asserts explicitly  – that all language use is necessarily figurative, and hence his theory can be characterized as “pan-figurative.” Bracketing for now the ramifications of this view, we may ask to what extent was it grounded in the work of Sthiramati’s predecessors and to what extent is it innovative, and more specifically, what are the particular points of influence and difference between his works

In Ordinary Language “Bāhīka” (as well as “self” and “things”, “ox”, etc.)

Figurative-indirect denotation

Figurative-indirect denotation

Transformations of Consciousness (causal, mental)

“Ox”

Direct denotation

Direct denotation

An appearance of a Bāhīka person

“Bāhīka”

Figure 22.1  Sthiramati’s pan-metaphorical claim in the Triṃśikābhāṣya.

383

Roy Tzohar and Jowita Kramer

and those of his predecessors? First in line, of course, is the question of how Sthiramati’s ideas stand in relation to those of Vasubandhu. As I mentioned previously, the term upacāra in the sense of figurative application appears only once in the Triṃśikā. It does figure prominently in Vasubandhu’s Commentary on the Treasury of Abhidharma (Abhidharmakośabhāṣya), though there, it is implemented mostly as a hermeneutical device for indicating assertions or phrases that need to be interpreted as carrying an implicit intent. Within this largely hermeneutical context, however, some cases in the Abhidharmakośabhāṣya seem to consider that named objects may be figures insofar as they are underlined by mental events (Gold 2015, 268 n.47), and in this restricted sense, Sthiramati may perhaps be explicating what is already implicit in Vasubandhu’s work. Nonetheless, it is only in Sthiramati’s commentaries – not just on the Triṃśikā but also on the Abhidharmakośabhāṣya – that figures are explicitly tied, theoretically speaking, to their underlying causal mental phenomena (Tzohar 2018, 126–36).19 Another important difference – if not a full-blown theoretical leap – between Sthiramati’s account and Vasubandhu’s involves the way in which the former develops the discussion of figures beyond the idea of involving a mere cognitive imposition of “things” and “self ” on the transformation of consciousness into a full-fledged linguistic figurative theory of meaning. Sthiramati facilitates this move by explicitly and consciously situating his arguments about metaphor against the background of a non-Buddhist Indian philosophical conversation on denotation (in which metaphor is distinctively linguistic),20 culminating his own discussion with some general observations on language.21 Within this framework, metaphors serve as much more than mere indications of a cognitive mistake, as they are utilized to explain how meaning is created within language. But why was this linguistic context so important to Sthiramati? A  possible explanation is found when we consider his particular position within the development of the Yogācāra textual tradition. Relative to the earlier textual corpus  – ascribed to Asaṅga, Maitreya, and Vasubandhu – Sthiramati operated under a much more defined notion of the Yogācāra as a textual tradition and a doxographically distinct school. This position brought with it a new set of philosophical challenges in need of addressing, a dynamic that is distinctly expressed in Sthiramati’s rebuttal – at the heart of his exposition of metaphor – of two “extremist” views (ekāntavāda), one of which is ostensibly that of the Madhyamaka school. For his move against the Madhyamaka, Sthiramati, putting new wine into old wineskins, unmistakably paraphrases an argument originally found in the Bodhisattvabhūmi (Dutt 1978, 31 ll. 16–20), which in turn has been cited and explicitly identified by Bhāvaviveka as a Yogācāra critique of the Madhyamaka (Tarkajvālā on the Madhyamakahṛdayakārikāḥ 5.82– 5.83ab; Eckel 2008, 281–82, 431–32). Arguably aware of this polemical history, Sthiramati uses this argument to take the Yogācāra debate with the Madhyamaka to the next stage, drawing on his figurative theory of meaning.22 To understand how his argument proceeds, we need first to clarify in more detail how Sthiramati’s pan-figurative claim is translated into a theory of meaning, and of what kind. Briefly put, Sthiramati’s pan-figurative view manages to salvage the meaningfulness of ordinary language by presenting what amounts to a figurative theory of sense (by which the meaning of language is understood primarily as a function of its use; see Figure  22.2).23 According to Sthiramati, all words are figurative because they denote objects only indirectly, referring directly to their appearances in their capacity as the transformation of consciousness. As I mentioned, the transformation of consciousness is not a static state of affairs but a dynamic causal process – indeed, it is the very Buddhist notion of dependent origination as a causal nexus underlying all phenomena. According to this scheme, a word designates not an object but rather the entire causal mental process by virtue of which the object appears in consciousness. Accordingly, the reference of the 384

Sthiramati

Meaning of a word as function of its use. Knowledge of the use pertains to the meaning. (“Transformations of Consciousness so there is an appearance of an external pot along with its designation”)

Transformations of Consciousness so there is an appearance of an external pot along with its designation

“Pot” Sense

Reference

Figure 22.2  The Yogācāra figurative theory of meaning.

word “person” is the underlying mental causal nexus that brought it about (its mental appearance), while the sense of the word is roughly the description of this causal process.24 The benefits for the Yogācāra from this creative understanding of meaning are multiple. First, it creates a kind of scholastic meta-language that “cleanses” ordinary discourse of its essentialist and reifying quality by understanding it in terms of causal mental descriptions – but does so without changing its basic vocabulary. Second, this framework allows the Yogācāra to distinguish between different levels of discourse and posit a hierarchy of meaning within the conventional realm of language. This last point requires some elaboration. Recall that under the Yogācāra theory of meaning, what a term comes to mean is constituted by an underlying causal description. These causal descriptions, however, are not static. As seen, for instance, in the Triṃśikā, causal mental explanations of phenomena presuppose an evolving description, one that increases in intricacy and subtlety and which embodies a variety of points of view on the continuum between the ordinary, deluded point of view and the complete, direct knowledge of a Buddha or an accomplished bodhisattva. So, for example, while for the simpleton (Figure 22.3), the word “pot” refers to an objectively existent pot, for an adept informed by śāstra and practice, it comes to mean the activity of the eight modes of consciousness and so on, and this continues until an exhaustive direct knowledge of causal relations (as the dependent nature) is reached, which, according to Sthiramati, marks the point of view of an advanced bodhisattva.25 Note that this account renders meaning (1) dependent upon point of view and (2) hierarchically organized, since insofar as some speakers’ causal descriptions are more accurate or insightful than others, their respective use of language is also more meaningful.26 Thus while the Yogācāra accept that all discourse is indeed conventional, its figurative nature and objective causal underpinning imply that some claims are more meaningful than others. This layered view of discourse works well, structurally speaking, with both the Yogācāra soteriology and system hermeneutics,27 but in terms of our present concern, it serves Sthiramati primarily by enabling him to defend the meaningfulness of Yogācāra metaphysical claims from the threat posed by Madhyamaka radical conventionalism.28 Sthiramati’s line of defense is to point out that the latter’s view of language as a self-referential closed system (1) is incoherent 385

Roy Tzohar and Jowita Kramer

“Pot”: “Transformations of Consciousness so there is an appearance of a pot along with its designation”

“Pot”: direct knowledge of all causal relations that brought bout the appearance, under the tatpṛṣṭhalabdhajñāna

“Pot”: Externally objectively existent pot

Figure 22.3  The Yogācāra conception of discourse.

and (2) carries a heavy price tag for Buddhists. First, while language does not truly represent reality in any respect, its denotative function and the very conditions of its meaningfulness require us to presuppose extra-linguistic underlying conditions as its “support.” This is because language, insofar as it is meaningful, must be intentional – words have to be about something; that is, they have to operate within a certain referential framework that distinguishes between designations and their alleged extra-linguistic referents. This does not mean that the referents of words are real or exist as they are designated; but it does suggest that the very principle of conceptual differentiation allowed by referential language – the principle according to which certain names designate certain objects and not others – hinges on the presence of underlying extra-linguistic restricting conditions. In light of this claim, the Madhyamaka position undermines the very conditions that render the school’s own discourse meaningful – thereby implicating it in a pragmatic contradiction. The Yogācāra critique of the Madhyamaka stance, therefore, is not (as it has often been interpreted) that it is nihilistic but that it is senseless.29 Second, though it may be polemically powerful, the Madhyamaka conventionalism entails a flattening of conventional discourse insofar as this view renders all claims – whether they are about ultimate reality or about medium-sized dry goods – equally meaningful (since they are all referenceless and hence empty). Given this, the Madhyamaka view of language is patently self-undermining also because it cannot provide any criterion in light of which the meaningfulness of their own claims might be distinguished from those of any other conventional language use (including, for instance, uninformed, naïve assertions about the “self ” or “things”). This conception of conventional discourse, however, cannot serve the Yogācāra, which presents explicit metaphysical claims – about an inexpressible underlying ontology and its causal mental modifications – and therefore needs to defend the meaningfulness of this discourse, that is, to defend the possibility of describing reality in a way that is not merely negative. This need is met by the school’s figurative theory of meaning, which allows it to present a notion of meaning and discourse according to which some descriptions of reality, although merely conventional, are more meaningful than others. Apart from responding to the Madhyamaka critique, this allows the Yogācāra to carve out a position that is exceptional in the Buddhist 386

Sthiramati

landscape: one that views ordinary language as incapable of representing or reaching reality but at the same time justifies the meaningfulness of the school’s own metaphysical discourse. The significance of Sthiramati’s figurative theory of meaning, however, does not end with this and carries broader ramifications for our understanding of the Yogācāra conception of experience and of the school’s soteriology. In the remainder of this section, I will point out, if only schematically, some deep structural affinities between Sthiramati’s understanding of linguistic meaning and his understanding of experience, particularly of intersubjective experiences of the external word as well as the experience of the enlightened mind while it abides in saṃsāra. This allows us to identify several fundamental themes running through his thought that, taken together, imply a broadly conceived theory of meaning that is not merely linguistic but also perceptual and is designed to address some of the fundamental concerns of the Yogācāra.30 An important benefit of the figurative theory of meaning as characterized previously is that it makes it possible to account for successful communication between diverse speakers – for instance, between an ordinary person and an enlightened one – despite deep discrepancies in their understanding. Specifically, this is enabled by seeing conventional discourse as incorporating different modes of language use, layered in accordance with their respective standing vis-à-vis knowledge of causality. Above all, this ability turns on an important feature that the figurative theory of meaning shares with contemporary causal theories of reference, namely that it allows for the reference of a term to remain fixed even as its meaning changes from one speaker to another. As such, it overcomes the danger of incommensurability – namely the relativistic claim that vast differences in our conceptual frameworks entail that the meaning of a common term (and insofar as meaning determines it, also its reference) may radically change between frameworks. Within the Yogācāra tradition, a version of the challenge of incommensurability – in a linguistic but also in a broader experiential sense – is palpably present in the need to account for the way in which beings at different ends of the Buddhist path (say, a fulfilled bodhisattva who is required to operate within saṃsāra and a simpleton) can converse and interact in a meaningful way despite the epistemic abyss that separates them. This issue is very clearly expressed in the Yogācāra discussions of how to bridge the bodhisattva’s nonconceptual (nirvikalpa) experience of reality on the one hand and her use of ordinary language within the realm of saṃsāra on the other. To this end, various Yogācāra thinkers, Sthiramati included, introduced a division of labor between two types of knowledge: nonconceptual or nondiscriminatory awareness (nirvikalpajñāna), and the “awareness obtained subsequent to it” (tatpṛṣṭhalabdhajñāna) – a different kind of insight that the advanced bodhisattva is said to attain following nonconceptual awareness. Sthiramati explicitly identifies the latter awareness with the dependent nature and as the state in which ultimate knowledge of causality is attained.31 This knowledge of causality, we are told by various Yogācāra sources, is what enables the bodhisattva to remain, operate, and communicate effectively in the phenomenal world. This awareness appears to enable a unique kind of conceptual (!) activity, which is, however, very different from ordinary conceptual awareness insofar as it is non-reifying (Tzohar 2018, 183–88; Tzohar 2020). Note here the similarities with Sthiramati’s figurative theory of meaning. In both cases, a certain epistemic/linguistic inherent gap – which is presupposed by the Buddhist worldview or path – is bridged by appealing to an underlying casual reality and by positing the possibility of conceiving of this underlying reality from a number of perspectives simultaneously (i.e., the dependent nature, seen from the side of the imagined in essentialist terms and from the side of the perfected without essentializing or reifying). The structural affinities between these schemes are quite striking, and once we discern them, they seem to be at work in other instances of Yogācāra accounts of experience as well. A distinct 387

Roy Tzohar and Jowita Kramer

example is the school’s – and pronouncedly Sthiramati’s – overall explanation for the possibility of intersubjective experiences under a representation-only framework (i.e., without accepting externality as objectively existent). Briefly put, these experiences are explained by first reducing them to an underlying mental causal nexus and then tracing the distinction between intersubjective and private kinds of experiences to various common and uncommon causal (karmic) strands, respectively. The focus here is indeed perceptual, but the governing principle is the same as in the case of linguistic meaning – in both cases, the reduction to causal and mental reality establishes a shared constructed realm of experience while still accounting for acute differences and discrepancies between each individual experience (Tzohar 2017c, 343–45). An analysis of Sthiramati’s treatment of these instances shows them all to be founded on a similar analytic structure. In all these cases, both agreement and discrepancy, the shared and the private, are explained in causal mental terms, yielding what can be described as a broadly conceived understanding of meaning – in the experiential and existential senses of the word – that is not merely linguistic but also perceptual. Common to all these accounts is an understanding of meaning as inevitably relational, shaped by the confluence of perspectives of a community of speakers and experiencers and implying a view of language as inherently polysomic and of experience as inevitably perspectival. This understanding of meaning, I have argued, is utilized by Sthiramati for various purposes: to defend the meaningfulness of the Yogācāra metaphysical discourse, to show the possibility of intersubjective experiences under a “mere representation” view, and to allow the school to overcome its soteriological version of the problem of incommensurability – that is, of how a bodhisattva may operate in saṃsāra and in particular how her nonconceptual experience and everyday experience and language use may be bridged (Tzohar 2018, Chapter 6). Seen against this background, Sthiramati’s lengthy commentary on the opening verse of the Triṃśikā and his account of metaphor emerges as more than just a local philosophical move: it functions as a broad strategy devised, rather ingeniously and innovatively, to deal with broader and more fundamental Buddhist concerns. Venturing beyond merely clarifying a case of cognitive or linguistic imposition (these are intertwined, his analysis tells us), his figurative theory of meaning introduces what can be seen as a “master” template for the Yogācāra conception of meaning – one that is applicable “in the world and in texts” and one that he applies throughout his commentary and develops in other works as well. Set within this context and the Yogācāra’s overall understanding of experience and meaning, his pan-metaphorical claim emerges as not merely internally coherent and polemically effective but indeed essential to the school’s salvific vision.

Notes 1 Chinese: Anhui (安慧), Jianhui (堅慧) (Silk 2015, 150–51); Tibetan: Blo gros brtan pa. 2 See Tzohar 2017b for a discussion of Sthiramati’s conception of the Yogācāra identity as a distinct school in the Madhyāntavibhāgaṭīkā, as reflected in his use of commentarial techniques for supporting a certain traditional narrative of the Yogācāra lineage of textual transmission. 3 See especially the writings of the Chinese pilgrim Xuanzang (玄奘, 600?–664) and the Tibetan historians Bu ston (1290–1364) and Tāranātha (1575–1634). 4 Diskalkar 1925, 37–40; Silk 2009, 385, n. 13. See also Kramer 2019, 456. 5 See Kramer 2015, 330–47. In some cases, Sthiramati presents the definitions in an abbreviated form in the Commentary on the Treatise in Thirty Verses, probably extracting only those phrases from his Commentary on the Treatise on the Five Constituents that appeared most relevant to him. But even though the text is not identical in these instances, it is still obvious that there is a direct relation between the comments (see also Kramer 2015, 300ff.).

388

Sthiramati 6 For studies on the Commentary on the Treasury of Abhidharma [Called] True Reality, see, for example, Matsuda 2013; Kazuo and Kramer 2020; Odani 2021. 7 Until recently, the text of the Commentary on the Treasury of Abhidharma [Called] True Reality was only available in a rather bad Tibetan translation and some Chinese and Uyghur fragments. One of the notable aspects of the Sanskrit manuscript of the Commentary on the Treasury of Abhidharma [Called] True Reality is that it seems to be one of the very oldest manuscripts preserved in Tibet. It has been dated to the eighth or ninth century by Kazunobu Matsuda based on its script. For more details, see Kazuo and Kramer 2020. 8 For details, see Kramer 2013a. 9 For details, see Kramer 2008. 10 See Kramer 2013b, 89. However, it should be noted that he does not use the term vijñaptimātratā explicitly. 11 The four arguments may be summarized as follows (Kramer 2016b, 146f.): (1) Actual perception ( pravṛttivijñāna) reappears after a person has risen from unconscious states; (2) actual perceptions appear in different modes ( prakāra) depending on different kinds of object conditions (ālambanapratyaya); (3) actual perception reappears after it has been interrupted by sleep (middha) or a swoon (mūrchā); (4) without the store mind, an individual could not (a) arise in and (b) be liberated from saṃsāra, because, on the one hand, the process of rebirth would not be possible and, on the other, the contaminations (kleśa) could not be removed. 12 For a detailed comparison, see Kramer 2016b, 146f. 13 Nguyen presents several arguments that from his point of view make it plausible that only one Sthiramati composed all but the tantric works attributed to him in traditional sources (1990, 23–83). 14 See Delhey 2016, 16ff. for an assessment of Sthiramati’s commentarial approach and style. 15 The present chapter is based on arguments that I elaborate on in greater detail in Tzohar 2017a, 2017b, 2017c, 2018. 16 All translations from the Triṃśikā and Triṃśikā-bhāṣya are my own. Translations from the first two are from Buescher’s Sanskrit and Tibetan critical edition (2007). 17 “And this metaphor of two kinds (of self and things] is applied in reference to the transformation of consciousness and not to a primary actual self or things. Why so? Since these do not exist outside of the transformation of consciousness. What is it that is termed ‘transformation’? The state of being otherwise than what it previously was. Here ‘transformation’ – which is the acquisition of the effect as something distinct from the moment in which the cause exists and simultaneous with the moment of the passing away of the cause – is a mental construct, an appearance of self, etc., and the appearance of matter, etc., all arising from the storehouse consciousness because of the incubation of latent impressions through the conceptual construct of a self, etc., and the incubation of latent impressions through the conceptual construct of matter, etc. Hence, the metaphor of things such as matter, etc., and the metaphor of self, etc., having its base in the appearance of matter, etc., and the appearance of self, etc., which due to a mental construction appear as if they were external, occurs from time immemorial, even without an external self and things, like in the case of the metaphor of net-like apparitions experienced by a person suffering from an eye-disease. And that which is not existent there in the locus of reference is figuratively designated, like when one calls the Bāhīka person an ox. . . . And thus it should be accepted that all objects are in fact unreal, due to their intrinsic nature being a fabrication. Consciousness, on the other hand, because of its being dependently originated, exists as a substrate (dravyataḥ). And again, consciousness’s state of being dependently originated is made known by the use of the term transformation’ ” (Tzohar 2018, 157–59). 18 This principle, which is constitutive of Sthiramati’s understanding of metaphor, is in fact based on the formula of the Nyāya-sūtra’s 2.2.61. . . . atadbhāve ’pi tadupacāraḥ / Tarkatirtha (1936–1944, 662). 19 A possible source for Sthiramati’s conception of this relation may be found in the Laṅkāvatārasūtra. See Tzohar 2018, 137–44. 20 Thus, in a lengthy section, Sthiramati pits his arguments about secondary denotation against the arguments of opponents from non-Buddhist schools and alludes clearly (perhaps even with parodic intent) to similar discussions in the Nyāya-sūtra and in the works of grammarians (Buescher 2007, 46:3– 48:11; Tzohar 2018, 161–66). Furthermore, Sthiramati’s account of metaphor appears to owe greatly to Bhartṛhari’s discussion of the issue, to which it shows close thematic as well as stylistic similarities (Tzohar 2018, Chapter 2). It is also possible that Sthiramati’s knowledge of these polemics was gained through the mediation of Dignāga’s works (especially the fifth chapter of the Pramāṇasamuccaya;

389

Roy Tzohar and Jowita Kramer see Tzohar 2018, 144–49). Finally, an early and important Yogācāra source of influence on Sthiramati’s thought, one that discusses upacāra in a distinctively linguistic polemical context, is the Chapter on the Meaning of Reality (Tattvārthapaṭalaṃ) of the Levels of the Bodhisattva (Bodhisattvabhūmi) and its accompanying sections in the later Collection of Clarifications (Viniścayasaṃgrahaṇī), both included in the Levels of Spiritual Training (Yogācārabhūmi) corpus. See Tzohar 2018, Chapter 3. 21 “Thus, since there is no correlation of an object with a word, language and knowledge do not exist. Thus, because of the non-existence of an expression and of a thing to be expressed, a primary object does not exist. And all that is merely figurative (gauṇa) and not a primary meaning” (Buescher 2007, 48:6–8). 22 “And because of the impossibility of a metaphor without a support (nirādhāra), the occurrence of the metaphors of the self and things should be necessarily regarded as actually being the transformation of consciousness. Due to this, the following [Mādhyamika’s] view would not withstand reasoning: ‘That the object of consciousness, just like consciousness, exists only conventionally (saṃvṛtitaḥ) and not ultimately ( paramārthataḥ).’ This view is not logically tenable since it would result in the nonexistence of both of them even conventionally, because convention without a basis (upādāna) does not make sense. In light of this, the extremist views of both types must be abandoned because they are illogical, so said the ācārya [Vasubandhu].” (Buescher 2007, 42:9-15) 23 An alternative and viable reading takes it as a figurative theory of reference, whose main purpose is a Yogācāra defense against the realist attack on its supposed Mentalism. See Gold 2007. 24 Sthiramati provides a dual characterization of the transformation of consciousness – as an underlying causal reality on the one hand and a conceptual construct on the other (for the latter, see Buescher 2007, 108:4–109:13; Tzohar 2018, 166–69) – suggesting a basic distinction between real causal activity and its imagined products, or, with an eye to their philosophical presentation, between ineffable real causal facts and constructed casual descriptions. 25 According to Sthiramati, such knowledge is attained by the bodhisattva under a “subsequent mundane pure awareness” (tatpṛṣṭhalabdhaśuddhalaukikajñāna), which follows from his previous nonconceptual transcendent awareness (nirvikalpalokottarajñāna). See Sthiramati’s commentary on Triṃśikā verse 22; Tzohar 2018, 180–82. 26 Note that under this framework – this figurative theory of sense – discourse is layered in terms of its degrees of meaning but not in terms of its truth value. This is because, within this framework, the truth of propositions was understood to be a function of the relations between expressions and their alleged referents. The latter’s ultimate non-existence on the one hand and the utter conventionality of discourse on the other hand render all propositions false from the ultimate point of view. However, the picture of conventional discourse as incorporating different modes of language-use, layered in accordance with their respective standing vis-à-vis knowledge of causality, results in some propositions being more meaningful than others. See also Tzohar 2017b, section 2. 27 A common feature to all these schemes is their progressional structure: they all facilitate a dialectical movement of understanding through which erroneous accounts of the same reality are gradually replaced by more accurate ones, up to the final knowledge of a Buddha, thereby enabling the school to correlate structurally between more accurate levels of discourse on the one hand and the gradual progress along the path on the other. 28 See note 22 for Sthiramati’s argument. It presents a set of parallel observations: the transformation of consciousness is to the metaphors of “self ” and “things” what a “support” is to a linguistic metaphor and what a “basis” (upādāna) is to the “conventional.” To a large extent, the understanding of Sthiramati’s critique of the Mādhyamika turns on what we take him to mean by these parallel observations and these specialized terms. For an analysis of this argument, see Tzohar 2018, 173–77. 29 This reading of Sthiramati’s argument is further reinforced by Vinītadeva’s later explanation of this passage in his sub-commentary, the Sum Cu Pa’i ’grel Bshad (Triṃśikā-ṭīkā*) TD. 4070 6b.3–6. 30 The essence of this move appears in Tzohar 2018, Chapter 6. 31 Sthiramati on the Triṃśikā verse 22 (Buescher 2007, 126:7ff.). For the translation and analysis of this section, see Tzohar 2018, 180–82.

Bibliography Braarvig, Jens. 1993. Akṣayamatinirdeśasūtra. Oslo: Solum. Buescher, Harmut. 2007. Sthiramati’s Triṃśikāvijñaptibhāṣya. Wien: Verlag der Österreichischen Akademie der Wissenschaften.

390

Sthiramati Bühler, Georg. 1877. “Further Valabhî Grants.” Indian Antiquary 6: 9–21. Cordier, Palmyr. 1909. Catalogue du fonds tibétain de la Bibliothèque Nationale. Deuxième partie: Index du Bstan-ḥgyur (tibétain 108–179). Paris: Imprimerie Nationale. Delhey, Martin. 2016. “The Indian Yogācāra Master Sthiramati and His Views on the Ālayavijñāna Concept.” International Association for Buddhist Thought and Culture 26: 11–35. Derge, T. 199 – ?. 1986. Bstan ’Gyur Sde Dge’i Par Ma: Commentaries on the Buddha’s Word by Indian Masters (Electronic Edn, 11 Cds). [Reproduced from Editions of the Individual Sections Published in Delhi at the Delhi Karmapae Chodhey, Gyalwae Sungrab Partun Khang, Itself a Reproduction from Prints from the 18th Century Sde-Dge Blocks], 213 vols. New York: Tibetan Buddhist Resource Center. Diskalkar, D. B. 1925. “Some Unpublished Copper-Plates of the Rulers of Valabhī.” Journal of the Bombay Branch of the Royal Asiatic Society 1: 13–64. Dutt, N. 1978. Boddhisattvabhūmiḥ, 2nd ed. Patna, India: K. P. Jayaswal Research Institute. Eckel, M. D. 2008. Bhāviveka and His Buddhist Opponents. Harvard Oriental Series. Cambridge, MA: Harvard University Press. Frauwallner, Erich. 1961. “Landmarks in the History of Indian Logic.” Wiener Zeitschrift für die Kunde Süd- und Ost-Asiens 5: 125–48. Gold, Jonathan C. 2007. “Yogācāra Strategies Against Realism: Appearances (Ākṛti) and Metaphors (Upacāra).” Religion Compass 1 (1): 131–47. ———. 2015. Paving the Great Way: Vasubandhu’s Unifying Buddhist Philosophy. New York: Columbia University Press. Kazuo, Kano, and Jowita Kramer. 2020. “The Fourth Chapter of the Tattvārthā Abhidharmakośaṭīkā: on Forbidding Intoxicating Liquor.” In Sanskrit Manuscripts in China III, edited by Birgit Kellner, Xuezhu Li, and Jowita Kramer, 107–37. Beijing: China Tibetology Publishing House. Kramer, Jowita. 2012. “Descriptions of ‘Feeling’ (Vedanā), ‘Ideation’ (Saṃjñā), and ‘the Unconditioned’ (Asaṃskṛta) in Vasubandhu’s Pañcaskandhaka and Sthiramati’s Pañcaskandhakavibhāṣā.” Rocznik Orientalistyczny 65 (1): 120–39. ———. 2013a. “A  Study of the Saṃskāra Section of Vasubandhu’s Pañcaskandhaka (with Reference to Its Commentary by Sthiramati).” In The Foundation for Yoga Practitioners: The Buddhist Yogācārabhūmi Treatise and Its Adaptation in India, East Asia, and Tibet, edited by Ulrich Timme Kragh, 920–70. Harvard Oriental Series. Cambridge, MA: Harvard University Press. ———. 2013b. “Notes on the Rūpa Section of the Pañcaskandhakavibhāṣā.” China Tibetology 21 (2): 86–99. ———. 2014a. Sthiramati’s Pañcaskandhakavibhāṣā. Beijing and Vienna: China Tibetology Publishing House/Austrian Academy of Sciences Press. ———. 2014b. “Indian Abhidharma Literature in Tibet: The Section on Vijñāna in Sthiramati’s Pañcaskandhakavibhāṣā.” In Buddhism Across Asia: Networks of Material, Intellectual and Cultural Exchange, edited by Tansen Sen. Singapore: Institute of Southeast Asian Studies. ———. 2015. “Innovation and the Role of Intertextuality in the Pañcaskandhaka and Related Yogācāra Works.” Journal of the International Association of Buddhist Studies 36/37: 281–352. ———. 2016a. “Some Remarks on Sthiramati and his Putative Authorship of the Madhyāntavibhāgaṭīkā, the *Sūtrālaṃkāravṛttibhāṣya and the Triṃśikāvijñaptibhāṣya.” Buddhist Studies Review 33 (1–2): 47–63. ———. 2016b. “Some Remarks on the Proofs of the ‘Store Mind’ (Ālayavijñāna) and the Development of the Concept of Manas.” In Text, Philosophy, and History: Abhidharma Across Buddhist Scholastic Traditions, edited by Bart Dessein and Weijen Teng, 146–68. Leiden: Brill. ———. 2019. “Sthiramati.” In Brill’s Encyclopedia of Buddhism, Volume II: Lives, edited by Jonathan A. Silk, 456–57. Leiden: Brill. Matsuda, Kazunobu. 2013. “Sanskrit Manuscript of Sthiramati’s Commentary to the Abhidharmakośbhāṣya.” China Tibetology 20 (1): 48–52. Nguyen, Cuong Tu. 1990. “Sthiramati’s Interpretation of Buddhology and Soteriology.” PhD diss., Harvard University. Obermiller, Eugene. 1931. History of Buddhism by Bu-ston (I. Part: The Jewelry of Scripture). Heidelberg: Harrassowitz. ———. 1932. History of Buddhism by Bu-ston (II. Part: The History of Buddhism in India and Tibet). Heidelberg: Harrassowitz. Odani, Nobuyicho. 2021. “An Introduction to Sthiramati’s Tattvārthā Chapter I.” In Illuminating the Dharma: Buddhist Studies in Honour of Venerable Professor KL Dhammajoti, edited by Toshiichi Endo, 295–306. Hong Kong: Centre of Buddhist Studies, The University of Hong Kong.

391

Roy Tzohar and Jowita Kramer Sakuma, Hidenori. 2006. “On Doctrinal Similarities Between Sthiramati and Xuanzang.” Journal of the International Association of Buddhist Studies 29 (2): 357–82. Schiefner, Anton. 1868. Tāranāthae de Doctrinae Buddhicae in India Propagatione Narratio. St. Petersburg: Academia Scientiarum Petropolitana. Silk, Jonathan A. 2009. “Remarks on the Kāśyapaparivarta commentary.” In Pāsādikadānaṃ: Festschrift für Bhikkhu Pāsādika, edited by Martin Straube et al., 381–97. Marburg: Indica et Tibetica. ———. 2015. Buddhist Cosmic Unity: An Edition, Translation and Study of the Anūnatvāpūrṇatvanirde śaparivarta. Hamburg: Hamburg University Press. Tarkatirtha, T. N., and A. M. Tarkatirtha, eds. 1936–1944. Nyāyadarśanam with Vātsyāyana’s Bhāṣya, Uddyotkara’s Vārttika, Vācaspati Miśra’s Tātparyaṭīkā, and Viśvanātha’s Vṛtti. Calcutta Sanskrit Series. Calcutta, India: Metropolitan Print and Publishing House. Tzohar, Roy. 2017a. “Does Early Yogācāra Have a Theory of Meaning? Sthiramati’s Arguments on Metaphor in the Triṃśikā-Bhāṣya.” Journal of Indian Philosophy 45 (1): 99–120. ———. 2017b. “Indian Yogācāra Understanding of Āgama-Pramāṇa.” In E. Freschi (guest ed.), “There in only ‘Philosophy:’ The Case of Testimony”. Special issue of Kervan: International Journal of AfroAsiatic Studies 21: 261–77. ———. 2017c. “Imagine Being a Preta: Early Indian Yogācāra Approaches to Intersubjectivity.” Sophia 56 (2): 337–54. ———. 2018. A Buddhist Yogācāra Theory of Metaphor. Oxford and New York: Oxford University Press. ———. 2020. “Contemporary Non-Conceptualism, Conceptual Inclusivism, and the Yogācāra View of Language Use as Skillful Action.” Philosophy East and West 70: 638–60. Yamaguchi, Susumu. 1934. Madhyāntavibhāgaṭīkā. Nagoya: Librairie Hajinkaku.

392

23 DEVENDRABUDDHI AND ŚĀKYABUDDHI Dharmakīrti’s First Commentators Alexander Yiannopoulos Devendrabuddhi (fl. c. 650) and Śākyabuddhi (fl. c. 675) were the first two authors – apart from Dharmakīrti himself – to compose commentaries on Dharmakīrti’s (fl. c. 625) Commentary on the Means of Knowledge (Pramāṇavārttika, henceforth PV). Devendrabuddhi is not known to have composed any additional works. Śākyabuddhi is listed as the author of two additional texts in the Tibetan Tengyur, commentaries on the Ten Stages Sūtra (Daśabhūmika Sūtra, Toh 3999) and the Summit of Mount Gayā Sūtra (Gayāśīrṣa Sūtra, Toh 3992); his authorship of these commentaries is generally accepted. Although Devendrabuddhi’s and Śākyabuddhi’s commentaries are distinct works, they are so closely intertwined that, for reasons which will be elaborated upon in the following, it is best to consider them parts of the same “hybrid text.” Tibetan tradition holds Devendrabuddhi to have been a direct disciple of Dharmakīrti (Chimpa and Chattopadhyaya 1970, 239), and this is widely accepted by contemporary scholars as well. The precise nature of the relationship between Śākyabuddhi and Devendrabuddhi is less certain, which is to say that it is unclear whether or to what extent Śākyabuddhi may also have been taught directly by Dharmakīrti or how closely Devendrabuddhi and Śākyabuddhi collaborated on their closely intertwined commentaries. In other words, while Śākyabuddhi is typically portrayed as the disciple of Devendrabuddhi, this should not necessarily be taken to mean that Śākyabuddhi had no direct contact with Dharmakīrti. It should similarly not be assumed that Devendrabuddhi’s Explanation of the Commentary on the Means of Knowledge (Pramāṇavārttikapañjikā, henceforth PVP) was composed strictly prior to and entirely independently of Śākyabuddhi’s Commentary for the Commentary on the Means of Knowledge (Pramāṇavārttikaṭīkā, henceforth PVṬ). That is, although it is common to regard “commentaries” as works composed independently of and posterior to the “root text” upon which they are commenting, the commentarial relationship is significantly more complex in the present case: at the very least, it must be understood that the root text of the PV was never intended to be even grammatically intelligible in the absence of some sort of commentary, and Devendrabuddhi may well have intended the PVP to be similarly dependent upon the PVṬ for broader exegesis and analysis. As for the texts themselves, the PV is fortunately well attested in Sanskrit. Unfortunately, and by contrast, although a few Sanskrit manuscript fragments of the PVP and PVṬ are still extant, for the most part, both texts are currently only known to be extant in Tibetan translation. In general, Devendrabuddhi’s PVP is written in an extremely close “word commentary” (tshig ’grel) style, for the most part only clarifying the literal meaning of Dharmakīrti’s terse DOI: 10.4324/9781351030908-34

393

Alexander Yiannopoulos

Sanskrit, which on its own is frequently obscure to the point of sheer incomprehensibility. Devendrabuddhi only very rarely ventures outside the bounds of providing semantic and grammatical glosses to Dharmakīrti’s root text. Śākyabuddhi’s PVṬ, by contrast, contains much more by way of extrapolation and interpretation. But given the complex relationships among Dharmakīrti’s PV, Dharmakīrti’s own Autocommentary for the Commentary on the Means of Knowledge (Pramāṇavārttikasvopajñavṛtti; henceforth PVSV), Devendrabuddhi’s PVP, and Śākyabuddhi’s PVṬ, these texts may be considered to constitute a type of aggregated “hybrid text” (McClintock 2010, 2) rather than a series of isolated or independent works. In order to understand this point, however, it is first necessary to examine these closely intertwined relationships among the four texts in question: the PV, PVSV, PVP, and PVṬ.

Commentarial Strata and the PV The genre of South Asian scholastic commentarial literature is intertextual to an extraordinary degree. Even the pretense of original work hardly exists; nearly all intellectual labor is performed in terms of the twin projects of (1) commenting upon the predecessors in one’s own tradition and (2) rebutting those in other traditions – and sometimes those in one’s own – who have rebutted one’s predecessors in one’s own tradition and so on (Ganeri 2010). It is thus a mistake, albeit a common one, to consider the underlying verses of the “root” (mūla) text complete and self-contained with the commentary as a mere supplement – that is, something superfluous, ultimately inessential, or strictly unnecessary, which is added to the purportedly original unity of the root text. While it is certainly arguable that there exist examples of this kind of relationship in South Asian literature, in the context of Dignāga and Dharmakīrti’s works specifically, such a view would fundamentally mischaracterize the relationship between the root verses and their commentaries. Consider, for example, the relationship between Dignāga’s Encyclopedia of the Means of Knowledge (Pramāṇasamuccaya, henceforth PS) and his Commentary on the Encyclopedia of the Means of Knowledge (Pramāṇasamuccayavṛtti, henceforth PSV). The root verses of the PS function more as a mnemonic aid, or as a skeleton to be fleshed out by the commentary, than as a series of grammatically intelligible Sanskrit sentences – the verses primarily serve to facilitate the memorization of arguments that are developed at least somewhat more clearly in the commentary, and the two (verse and commentary) were undoubtedly composed contemporaneously (Steinkellner 2005, IIIn1). To the extent that any “original unity” exists, then, it is located not in the verses of the PS itself, but rather in the complex textual interplay between the PS and the PSV: a “hybrid text” that may be designated as the “PS(/V).” The relationship between Dharmakīrti’s PV and his autocommentary, the PVSV, should be understood along these same lines. That is to say, in a manner precisely analogous to the PS(V), the PV and the PVSV were deliberately composed as a hybrid text: the “PV(/SV).” The key difference between the PS and PV in this regard is that Dignāga composed his own autocommentary to each of the six chapters of the PS, while Dharmakīrti’s PVSV only deals with the first chapter of his root text, PV 1.1 To repeat: Dharmakīrti’s Autocommentary for the Commentary on the Means of Knowledge only comments upon the first chapter of his Commentary. Devendrabuddhi’s Explanation of the Commentary on the Means of Knowledge thus defers to Dharmakīrti’s own Autocommentary with respect to this first chapter. That is to say, Devendrabuddhi’s PVP does not comment upon the first chapter of Dharmakīrti’s PV but rather deals only with its second, third, and fourth chapters. Devendrabuddhi thus rhetorically positions his Explanation as occupying the exact same commentarial stratum as Dharmakīrti’s own Autocommentary, effectively 394

Devendrabuddhi and Śākyabuddhi

serving as an extension or equivalent of Dharmakīrti’s own contribution. And this is precisely how Śākyabuddhi’s secondary commentary treats Devendrabuddhi’s work: Śākyabuddhi comments on Dharmakīrti’s Autocommentary with respect to the first chapter of the PV and Devendrabuddhi’s Explanation with regard to its second, third, and fourth chapters. In this way, Śākyabuddhi’s methodological choice places Devendrabuddhi’s commentary on an equal footing with Dharmakīrti’s Autocommentary. Like Devendrabuddhi’s methodological choice to eschew composing a commentary on the chapter of the PV that Dharmakīrti had himself already commented upon, this clearly indicates that Dharmakīrti’s immediate successors  – and quite possibly Dharmakīrti himself – considered the PVP the functional equivalent of a commentary composed by Dharmakīrti himself. Hence, despite the relative lack of scholarly attention paid to this intricate commentarial structure to date, in more or less the same way that Dignāga’s PS and PSV should be regarded as a single (though bipartite) work, the PV, PVSV, PVP, and PVṬ should similarly be regarded as a single (though hybrid) textual structure. With this understanding in mind, we now turn to an examination of some key issues raised by Devendrabuddhi and Śākyabuddhi in their respective commentaries on the PV. These points are noteworthy both for their philosophical relevance in terms of drawing out the underlying logic and broader implications of Dharmakīrti’s system and for their exemplification of the commentators’ general hermeneutical approach.

Unmediated Instrumental Effect As mentioned previously, Devendrabuddhi largely refrains from presenting additional content on top of a basic explanation of Dharmakīrti’s root text. For the most part, his commentary only provides syntactic and semantic glosses on Dharmakīrti’s extremely elliptical Sanskrit. Perhaps the most dramatic exception to this general rule concerns Devendrabuddhi’s introduction of a terminological distinction between the “mediated” (vyavahita) versus “unmediated” (avyavahita) instrumental effect ( pramāṇaphala),2 the latter being defined as the “resulting” ( phala) knowledge that comes about due to the “instrumental activity” (kriyā) or operation of an “instrument of knowledge” ( pramāṇa).3 But to the extent that Devendrabuddhi’s innovation here serves only to resolve an otherwise unexamined ambiguity in Dharmakīrti’s epistemological system, it illustrates Devendrabuddhi’s general approach as a commentator: namely that of maximal self-effacement. Unlike later commentators on Dharmakīrti such as Dharmottara (c. 775–850) and Prajñākaragupta (c. 800–875), who (not uncommonly for the genre) took the task of commentary as an opportunity to engage in substantively constructive projects of their own, Devendrabuddhi’s commentarial “voice” is difficult to discern. As with his decision to forego any commentary on the PV’s first chapter – seemingly because Dharmakīrti himself had already provided it – even Devendrabuddhi’s introduction of novel terminology is characterized by maximal deference to Dharmakīrti. The underlying ambiguity addressed by Devendrabuddhi’s introduction of the “(un/)mediated instrumental effect” concerns a tension or ambiguity between two meanings of the centrally important Sanskrit term artha, which in this context may mean “object” or “real thing” on one hand and “purpose,” “aim,” or “goal” on the other hand. This second meaning can be understood specifically in terms of the type of “human aim” ( puruṣārtha) that is the goal of practical activity (vyavahāra). These two meanings of artha in turn bear upon two possible interpretations of the compound arthakriyā, a key concept in Dharmakīrti’s system, which combines the word artha with the word kriyā, meaning activity. The compound may thus be interpreted as either the “causal functionality” possessed by real objects or the “purposeoriented activity” engaged in by humans and other sentient beings. 395

Alexander Yiannopoulos

The ambiguity between these two different senses of artha and arthakriyā may be illustrated in the distinction between two different types of sensory-cognitive events: the aforementioned sensory perception of fire and the spurious perception of myodesopsic “floating hairs” (keśa) or the “double moon” (dvicandra) that arises due to having crossed eyes. In the first case, fire is the object (artha) of a sensory perception; hence, when perceiving fire for the purpose (artha) of seeking out warmth, the causal functionality (arthakriyā) of the fire, consisting in the fact that the fire is warm by nature, backstops the trustworthiness of the perception of fire. “In other words, an entity’s causal characteristics, while independent of a perceiver’s expectations, are conceptualized by the perceiver within the context of expectations concerning some goal” (Dunne 2004, 259): the perception of fire is trustworthy because real fire actually does possess the causal functionality (arthakriyā) of fire, entailing that one’s purposeful activity (arthakriyā) of seeking out warmth will result in the attainment of one’s goal (artha). But the ambiguity in and of itself does not necessarily constitute a philosophical problem in this case. Indeed, the overlapping senses of the terms artha and arthakriyā on display here are almost certainly part of Dharmakīrti’s intended point: that the practical utility of a perceptual event derives from the perceptual object’s fulfillment of some practical goal. However, the situation is rather different in the case of the cognition of floating hair or a double moon. There, because it is not the case that there really exists any floating hair or second moon, the cognition technically has no artha – that is, neither any real object nor any practical purpose. As Dunne (2004, 257–58) explains in a passage that includes an excerpt from Devendrabuddhi’s PVP ad PV 3.1 (note that, in this translation, Dunne renders arthakriyā as “telic efficacy”): Although artha as “aim” and “object” may thus be distinguished, even when the term artha is best rendered as “object” it retains some connection to the notion of a purpose or goal. Such is the opinion of Devendrabuddhi, as is apparent when he comments upon Dharmakīrti’s statement at the outset of PV 3 that the illusory “hairs” or “flies” perceived by a person with cataracts cannot be considered “objects” (arthas). As noted earlier, the crux of the problem here is that Dharmakīrti might be obliged to posit not only particulars and universals as objects of knowledge (  jñeya), but also a third type, namely, illusory objects. Commenting on this problem, Devendrabuddhi raises an objection and offers his reply: Opponent: “Well, in the awareness of a person with defective vision there are false appearances such as hairs, flies, two moons and so on. Those false appearances do not accomplish any purpose (artha) at all; hence, they are not counted as particulars. But even though they are devoid of telic efficacy, they are not included among universals because they appear clearly and because they are not distributed over anything. Thus, since they are not subsumed under the categories of particulars and universals, they are another kind of object. Hence, it is not correct that there are just two kinds of objects.” Devendrabuddhi: “No, it is not the case that our view is not correct. To be specific, hairs and so on are not objects (arthas). Why? Because they are not considered to be objects (PV 3.1d). That is, persons engaged in practical action (vyavahartṛ) do not consider them to be objects. The intention of this statement is as follows. If the flies and so on that are perceived by a person with cataracts and so on were to be objects, then one would investigate the situation, asking, ‘Is it a particular, or is it the other [i.e., a universal]?’ But they are not objects in that fashion because, with regard to an awareness in which there is the appearance of flies and so on, persons engaged in practical action do not have the intention, ‘This is the object of that awareness.’ ” 396

Devendrabuddhi and Śākyabuddhi

The crux of the problem here is that Dharmakīrti only admits two types of epistemic objects, particulars and universals, which are respectively the object of nonconceptual (i.e., perceptual) and conceptual (i.e., inferential) awareness. Dharmakīrti furthermore insists that only particulars are actually real because only particulars possess causal functionality (arthakriyā). According to Dharmakīrti, only particulars can be the cause of a sensory cognition because a sensory cognition is the direct and immediate effect of the causal interaction between the sensory faculties and the sensory object. Universals are not real in this way because they do not possess causal functionality. However, Dharmakīrti also goes to great lengths to explain that the appearance of floating hair or the double moon is in fact nonconceptual (Yiannopoulos 2020, 129–56). But how can a nonconceptual cognition arise without a real object (artha) to serve as its cause? What, in other words, accounts for the “vivid” (spaṣṭa) appearance of nonconceptual sensory errors, as opposed to the nonvivid appearance of conceptual sensory errors, such as when a mirage is mistaken for water? In the latter case, the mirage does not truly have the appearance of water and may be correctly identified as a mirage; in the former case, however, there is indeed a visual cognition with the vivid appearance of two moons. While this is an enormous topic that touches upon nearly every aspect of Dharmakīrti’s ontology and epistemology, the short answer is that there are two ways to analyze such nonconceptual errors. These two methods of analysis ultimately correspond to Devendrabuddhi’s distinction between the mediated (vyavahita) and the unmediated (avyavahita) instrumental effect ( pramāṇaphala), the latter being defined as the knowledge that is the “result” ( phala) of the causal operation of the epistemic instrument ( pramāṇa). Analyzed in terms of “human aims” ( puruṣārtha) and “practical activity” (arthakriyā), such nonconceptual errors are in fact “noncognitions” (ajñāna; Yiannopoulos 2020, 276) because they have no artha. They have no artha since they only arise due to an internal defect within the sensory faculties (PV 3.293). Analyzed in terms of sheer “causal functionality” (arthakriyā), however, it must be admitted that the cognition of floating hair or a double moon does arise directly from its causes (the aforementioned defect in the sensory faculties), and it does produce effects. As such, it meets Dharmakīrti’s requirement that all real phenomena are causally efficacious by nature, and vice versa.4 In this latter sense, in fact, every cognition may be understood to possess arthakriyā, just insofar as every cognition is a real entity that arises from causes and produces effects. Thus, in the first case, a cognition (such as the perceptual cognition of fire) is “instrumental” insofar as it will eventually allow for the attainment of some practical goal, while in the second case, cognition is only “instrumental” with respect to itself as a cognition, just in terms of the fact that it is occurring at all, without any reference to any further goal or aim. As Dunne (2004, 261–62) explains, On Devendrabuddhi’s view, Dharmakīrti proposes two alternative interpretations of what constitutes the instrumental effect. He claims that “. . . there are two kinds of instrumental effects: one called a ‘human aim ( puruṣārtha),’ which is a mediated (chod pa = vyavahita) effect, and a distinctive one (khyad par can = viśiṣṭa) that is not mediated.” . . . [An] effect is mediated (vyavahita) in that there is an “interval” (vyavadhāna) between the functioning of the instrument (the karaṇa, i.e., the pramāṇa) and the effect. In other words, the effect is separated (vyavahita) from the instrument by virtue of other intervening causes and conditions that must be in place for the effect to occur. The effect is thus “remote” (vyavahita) relative to the instrument, and since some obstruction can therefore occur between the functioning of the instrument and the production of the effect, the effect may not necessarily occur, even if the instrument functions correctly. . . . 397

Alexander Yiannopoulos

In contrast to the mediated effect, the unmediated effect is “distinctive” in that, as Śākyabuddhi notes, it “necessarily occurs.” This is so because the effect is not separated (avyavahita) from the instrument; it is, in fact, identical to the instrument itself. As such, it is not remote (avyavahita); instead, the effect is actually simultaneous with the instrument. In other words, the instrumental cognition ( pramāṇa) is itself the instrumental effect ( pramāṇaphala). In Devendrabuddhi’s account, in other words, the mediated instrumental effect – construed in terms of attaining a human aim ( puruṣārtha) – involves a temporal delay between the operation of the instrumental cognition ( pramāṇa) and the occurrence of the instrumental effect ( pramāṇaphala). For example, in this account, the initial sensory perception of fire would be the instrumental cognition, and the subsequent experience of its warmth would be the instrumental effect. This, indeed, is basically the only way to make sense of Dharmakīrti’s model of “instrumentality” ( prāmāṇya) in practical terms, since the attainment of practical goals in the world is the overarching context for pramāṇa discourse generally (Dunne 2004, 45–49), and the mere sensory perception of fire does not yet constitute the fulfillment of any practical goals (such as getting warm). However, there is a problem: this interpretation is irreconcilable with Dignāga’s and Dharmakīrti’s insistence that the instrumental cognition (i.e., the pramāṇa) is ontologically identical to the instrumental effect (i.e., the phala). In the Buddhist system as articulated by Dignāga and defended by Dharmakīrti, it is not permissible to assert any ontological distinction or temporal delay between the operation of the pramāṇa and the existence of the phala.5 Hence, the potential problem here is that, as Dunne (2004, 268 n. 78) notes, the underlying issue is that “Dharmakīrti’s extensive treatment of the instrumental effect ( pramāṇaphala) in PV 3 [in the context of reflexive awareness] does not correspond well to his account of instrumental cognition in PV 2.1–6 [in the context of practical human aims].” This is why Devendrabuddhi felt the need to make the distinction between the mediated and unmediated instrumental effect in the first place. In Devendrabuddhi’s account, only the unmediated instrumental effect is ontologically identical with the instrumental cognition; the mediated instrumental effect, being temporally subsequent, is ontologically distinct from the instrumental cognition. As with the case of floating hairs, however, in general, any cognition that is analyzed in terms of the unmediated instrumental effect cannot be understood to possess any causal functionality or telic efficacy (arthakriyā) other than its own momentary existence. When analyzing a cognition in terms of the unmediated instrumental effect, in other words, by definition, there can be no reference to any subsequent acquisition of what is desired, since any such acquisition would be “mediated” in precisely the sense at stake. This type of analysis is crucially important to Śākyabuddhi’s account of reflexive awareness, to which we now turn.

“Mere Reflexive Awareness” (svasam  vedanamātra) and False Imagism Reflexive awareness (svasaṃvitti or svasaṃvedana) is among the most important key terms in Dharmakīrti’s system, the topic of nearly half of the Commentary’s “Perception Chapter” ( pratyakṣapariccheda). And yet nowhere does Dharmakīrti provide an explicit definition of this centrally important term. Nevertheless, the basic idea is simple: reflexive awareness designates that feature of a cognition which immediately and reflexively (which is to say, nontransitively) presents the contents of cognition to itself. Grammatical reflexivity is central to the philosophical issue here: svasaṃvitti has not infrequently been translated as “self-awareness,” but this translation is at best misleading and at 398

Devendrabuddhi and Śākyabuddhi

worst simply incorrect. The confusion centers around the reflexive pronoun sva, which does not mean “self ” in the sense of a “soul” or “ego” (more commonly designated by ātman), but only designates that the accusative object of the verbal action is the same entity as the nominative subject; as in: “He holds himself well,” or “I hurt myself.” More rarely, Dharmakīrti does use ātman in compound with the root √vid (i.e., ātmavit) to designate reflexive awareness; but here as well, ātman is only functioning as a reflexive pronoun, rather than in any way that might imply some type of metaphysical “self ” that is aware of “itself.” Indeed, Dharmakīrti takes great pains to construct his epistemological system in such a way so as to avoid postulating any such agentive “self,” which would be absolutely incompatible with any type of Buddhist philosophy. Thus, the meaning of reflexive awareness is not “self-awareness” in the sense of the mind somehow taking itself for a transitive object or deliberately reflecting upon its own contents. As MacKenzie (2007) notes, reflexive awareness is not at all the same thing as reflective awareness. On the contrary, reflexive awareness designates the fact that the contents of cognition are always immediately, which is to say “unmediatedly” (avyavahita), present to the cognizing mind. Indeed, it is precisely by virtue of being unmediated in this exact way that the present moment of awareness constitutes the “unmediated instrumental effect” described previously. As Dunne (2004, 276) explains: Reflexive awareness . . . is an exception to the rule that the object of a perception ( pratyakṣa) is the cause of the image in the perceptual cognition. Hence, if it is to be considered instrumental, it clearly cannot be considered in terms of a mediated instrumental effect, since this would require an appeal to causality that is not possible in the case of reflexive awareness. And it likewise [is] difficult to imagine how reflexive awareness can fit into a scheme of human goals; hence, unlike a cognition in which appears the accomplishment of one’s aim, one cannot appeal to some unmediated experience of telos, except the restricted telos of knowing that one knows. . . . Thus, if we are to accept reflexive awareness as a trustworthy cognition . . . we must again turn to an alternative interpretation of arthakriyā, namely, that reflexive awareness is reliable in that it reveals the mere fact of experience, which is the same as saying that it reveals the mere causal efficiency (arthakriyā) of awareness. In other words, the “instrumentality” of pure reflexive awareness can no longer be understood in terms of the accomplishment of some ordinary human goal, precisely because (in this rarefied context) reflexive awareness is understood to be entirely unmediated. That is to say, because in this specific context epistemic instrumentality ( prāmāṇya) is no longer construed in terms of a cognition’s (i.e., a pramāṇa’s) ability to facilitate the acquisition or avoidance of some extramental “epistemic patient” or “object of knowledge” ( prameya) at some later time, but rather is strictly construed in terms of the cognition’s ability in the present moment to bear upon itself just as a cognition, there is no longer any question of the cognition being practically useful for any goal besides providing information about its own contents. Thus, as Dunne (2004, 317) explains: If we trust Śākyabuddhi’s opinion, the ultimate pramāṇa would be the pure, nondual, reflexive awareness of the mind itself. But while this ultimate instrumental cognition is the means to Dharmakīrti’s final soteriological goal, it is not useful for practical action in the world (i.e., saṃsāra). If the ultimate instrument of knowledge is indeed some pure form of reflexive awareness, then there are no longer external objects – or even mental content – on which to act. Hence, it would seem that conventional perceptions and inferences are eventually left behind, but in terms of cognitions for 399

Alexander Yiannopoulos

practical action in the world, these conventional cognitions are as trustworthy and irreproachable as any such cognition could ever be. The question or problem here arises because, according to Dharmakīrti, every ordinary cognition appears dualistically, such that the “object-image” or “aspect of the apprehended” (grāhyākāra) is necessarily accompanied by the first-person sense of subjectivity or “aspect of the apprehender” (grāhakākāra). Dharmakīrti refers to this necessity as the “rule that phenomenal subject and object must appear together” (sahopalambhaniyama). But in keeping with the Yogācāra tradition, and with Buddhist philosophy more generally, Dharmakīrti also repeatedly asserts that phenomenological duality as such constitutes a fundamental error or distortion (bhrānti) within the nature of cognition. This, then, is the issue: given that any dualistic cognition is necessarily and by nature erroneous, but also that any objective cognitive content (i.e., any grāhyākāra) must by necessity appear dualistically as part of a “structure of apprehended and apprehender” (grāhyagrāhakasthiti), how is it possible for any cognitive content to exist in the final analysis? Put slightly differently: if, as would seemingly have to be the case, the absence of phenomenological duality necessitates the absence of what would ordinarily be meant by “phenomenal content” tout court, how or in what sense could such a nondual cognition be useful for ordinary practical action in the world? Historically, the major dividing line within the post-Dharmakīrti Indian Buddhist epistemological tradition ran through the answer to this question. On one side were the “True Imagists” (satyākāravādins), who maintained that cognitive “images” or “content” (ākāra) could somehow remain, even in the absence of duality. On the other side were the “False Imagists” (alīkākāravādins), who argued to the contrary that the absence of duality entails the absence of cognitive content, leaving only “mere reflexive awareness” (svasaṃvedanamātra). Although this debate was not fully developed until centuries after Dharmakīrti, perhaps not until as late as 1000 CE, the False Imagist perspective may be observed in embryonic form within Śākyabuddhi’s remarks. In particular, Śākyabuddhi’s extensive comments concerning the extremely important verse PV 3.212 are a rich and important source for this perspective. This verse reads: This part of cognition – which is situated as though external – is different from the internal determination (i.e., the part that is situated as though internal). Hence, the appearance of difference in an actually undifferentiated awareness is cognitive distortion (upaplava). || 212 || Devendrabuddhi’s comments to this verse are sparse. This should come as no surprise, since, as discussed previously, Devendrabuddhi’s overriding concern as a commentator was fidelity to Dharmakīrti’s root text. In the case of the mediated versus unmediated instrumental effect, for example, Devendrabuddhi attempted to smooth out – with only the bare minimum of original contributions – an issue that constitutes perhaps the most glaring potential contradiction within Dharmakīrti’s epistemological system taken as a whole. But the issue here does not really present the same type of contradiction, because Dharmakīrti frequently specifies that duality just is cognitive error. Dharmakīrti’s silence on this topic may be frustrating to his readers, but does not in and of itself necessarily constitute an exegetical dilemma, insofar as there is no contradiction or incoherence to sort out. Accordingly, Devendrabuddhi apparently did not feel the need to introduce any new theoretical elaborations or terminological distinctions concerning this issue. Śākyabuddhi, on the other hand, in keeping with his penchant for extrapolation from and interpretation of Dharmakīrti’s root text, takes this extremely important verse as an opportunity to make more extensive comments. To begin with, he notes that “the bodhisattvas who 400

Devendrabuddhi and Śākyabuddhi

have realized that phenomena are selfless only know mere reflexive awareness (rang rig pa tsam = svasaṃvedanamātra), which is devoid of duality” (Dunne 2004, 406). In other words, according to Śākyabuddhi, all that remains following the collapse of phenomenological duality is mere contentless reflexive awareness. Furthermore, in keeping with what would later become the False Imagist position, Śākyabuddhi also asserts in the same passage that “dualistic awareness is not real; rather, it is established through cognitive error.” Śākyabuddhi then expands upon these brief remarks with a very interesting and lengthy digression on the nature of nonduality, subjectivity, and reflexive awareness. This excursus is framed as a response to a common recurring objection against the False Imagist perspective – namely that if neither phenomenological duality nor dualistic cognitions are actually real, then beings should constantly be experiencing “suchness” (tathatā) or nondual reality-as-such, since this would be the only thing that is in fact available to experience (Dunne 2004, 407): Someone objects: “If the object and subject do not exist, then what would be left but the suchness of awareness itself? Cognitively myopic beings do not experience anything but the objective and subjective cognitive images. If they were to experience something else, they would see suchness. That being the case, beings would be effortlessly liberated.” Śākyabuddhi answers this objection – which, it should be remembered, occurs in the context of comments to a verse that does not directly reference reflexive awareness at all – by articulating a distinction (only ever implied in Dharmakīrti’s root text) between reflexive awareness and subjectivity: [With] the word “subject” we do not mean to express reflexive awareness – the internal cognition that arises in various forms such as the pleasant and the unpleasant – such that by expressing it with the term “subject” we would be saying that it does not exist. Rather, we mean the following. Cognitive appearances such as blue seem to be external to awareness; [however,] there is ultimately no object that is distinct from awareness itself, and since that object does not exist, we say “the subject does not exist”. . . . Since an agent and its patient are constructed in dependence upon each other, these two (i.e., subject and object) are posited in dependence on each other. The expression “subject” does not express mere reflexive awareness, which is the essential nature of cognition itself. The essential nature of cognition is not constructed in mutual dependence on something else because it arises as such from its own causes. The essential nature of cognition is established in mere reflexive awareness. Since it is devoid of the above-described object and subject, [reflexive awareness] is said to be nondual.6 A complete analysis of this remarkable passage would lie outside the scope of this brief treatment; for a more thorough consideration of Śākyabuddhi’s distinction between reflexive awareness and subjectivity, see Yiannopoulos (2020, 388–413). In short, however, it should be noted that the key point here is the fact that reflexive awareness is “the essential nature of cognition itself ” – what Dharmakīrti elsewhere (PV 3.302) terms the “similar nature” (sadṛśātman) of every cognition, just insofar as it is a cognition. By contrast, phenomenal subjectivity is just an erroneous artifact of ignorance, posited in mutual dependence upon phenomenal objectivity. In concrete terms, what Śākyabuddhi is saying here in response to the opponent’s objection is that, so long as appearances appear, those appearances must be understood as fundamentally 401

Alexander Yiannopoulos

distorted or mistaken, precisely insofar as their objective appearance necessitates the concomitant appearance of subjectivity; because, even when they are properly understood within the idealistic framework of Yogācāra as “mind only” (cittamātra), these appearances are necessarily bound up with the fundamentally distorted structure of duality. “Mere reflexive awareness” (svasaṃvedanamātra), on the other hand, does not present any such properly objective phenomenal content. Indeed, later False Imagists such as Ratnākaraśānti (c. 1000 CE), drawing heavily upon Śākyabuddhi’s line of reasoning, would explicitly name this state “luminosityonly” ( prakāśamātra), as “luminosity” ( prakāśa) is a synonym for reflexive awareness, even in Dharmakīrti’s Commentary (see, in particular, PV 3.327–32). Hence, even though every cognition is identical in terms of having the essential, luminous nature of nondual “mere reflexive awareness,” it is not the case that cognition ordinarily appears as mere nondual luminosity. On the contrary: despite not actually having a dualistic nature, cognition ordinarily appears as though there were a discrete phenomenal subject and phenomenal object. Therefore, in the absence of nondual yogic practice (i.e., yogācāra) aimed at eliminating this distorted structure of duality, it is impossible to have a definite determination or ascertainment (niścaya) of the suchness of one’s own experience, which is to say, of the fact that experience as such is by nature devoid of duality. As Ratnākaraśānti succinctly summarizes in his Pith Instructions for the Ornament of the Middle Way, referencing the exact same objection raised by Śākyabuddhi previously (Yiannopoulos 2012, 227–28): One might ask: “Since this emptiness of duality is beginningless, doesn’t that mean that we should have been enlightened from that time, at the very beginning?” This does not follow, because even though one has some kind of experience, still the emptiness of duality on the part of that experience is not ascertained. In this way, we may see that Śākyabuddhi’s line of interpretation, venturing rather further afield from Dharmakīrti’s root text than Devendrabuddhi’s extremely conservative approach, proved quite influential for the later Indian Buddhist epistemological tradition. While a detailed examination of Śākyabuddhi’s comments would lie outside the scope of this brief treatment, it is worth sketching out a few key takeaways from the preceding discussion. First, and arguably foremost, it must be emphasized that Dharmakīrti’s discussion of the nondual nature of awareness dovetails – almost certainly deliberately – with Mahāyāna Buddhist approaches to contemplative practice. Indeed, it would not overstate the matter to note that Dharmakīrti’s presentation of “luminosity” ( prakāśa), particularly as elaborated upon by Śākyabuddhi, provides a theoretical and epistemological foundation for advanced nondual contemplative practices such as Mahāmudrā. From this perspective, the nascent False Imagist view that may be discerned in the text is not simply an epistemological stance, but also provides an ontological and epistemological account of the key point of meditative equipoise (samāhita), and to some extent even an account of the nature of Buddhahood itself, as phenomenally contentless (i.e., nirākāra) luminosity. We may conclude on this note by highlighting yet again how, far from merely representing the opinion of two subsequent authors, the commentaries of Devendrabuddhi and Śākyabuddhi should be regarded as integral to the PV itself, forming together with it a complex hybrid-textual structure. This is not necessarily to say that the False Imagist perspective is the only possible interpretation of Dharmakīrti, but (per Devendrabuddhi) it is certainly the most grammatically and syntactically straightforward reading of Dharmakīrti’s Sanskrit, as well as (per Śākyabuddhi) the most hermeneutically concise and consistent interpretation of it.

402

Devendrabuddhi and Śākyabuddhi

Notes 1 The order of the chapters in the PV has been a matter of some controversy in India and Tibet as well as in the contemporary scholarly literature. For a discussion of this issue, see Yiannopoulos 2020, 11–14. 2 See PVP ad PV 2.3b – d, translated in Dunne 2004, 383–85. 3 For an overview of pramāṇa theory in terms of the relation between “means” and “result,” see Patil 2009, 37–40; Dunne 2004, 49–50; Yiannopoulos 2020, 159–65. 4 See Dunne 2004, 84–97. 5 For more detail concerning this point, see Yiannopoulos 2020, 168–72. 6 Translated in Dunne 2004, 407.

Bibliography Devendrabuddhi. 2002. “Pramāṇavārttikapañjikā (tshad ma rnam ’grel kyi ’grel pa).” In Bstan ’gyur (dpe bsdur ma), edited by krung go’i bod kyi shes rig ’jug lte gnas kyi bka ’bstan dpe sdur khang, translated by Subhutiśrī and Dge ba’i blo gros, vol. 98 (tshad ma che), 3–862. Beijing: krung go’i bod kyi shes rig dpe skrun khang. Dharmakīrti. 1960. “Pramāṇavārttikasvopajñavṛtti.” In The Pramāṇavārttikam of Dharmakīrti: The First Chapter with the Autocommentary, edited by Raniero Gnoli. Roma: Istituto Italiano per il Medio ed Estremo Oriente. ———. 1979. “Pramāṇavārttika.” In Bukkyō ninshikiron no kenkyū: Hōshō cho “Puramāna Vārutika” no genryōron, edited by Hiromasa Tosaki. Tōkyō: Daitō Shuppansha. ———. 2007. “Pramāṇaviniścaya.” In Dharmakīrti’s Pramāṇaviniścaya: Chapters 1 and 2, edited by Ernst Steinkellner. Sanskrit Texts from the Tibetan Autonomous Region. Wien and Beijing: Verlag der Österreichischen Akademie der Wissenschaften; China Tibetology Publishing House. Dunne, John. 2004. Foundations of Dharmakīrti’s Philosophy. Boston: Wisdom Publications. Chimpa, Lama, and Alaka Chattopadhyaya. 1970. Tāranātha’s History of Buddhism in India. Delhi: Motilal Banarsidass. Ganeri, Jonardon. 2010. “Sanskrit Philosophical Commentary.” Journal of the Indian Council of Philosophical Research 27: 187–207. MacKenzie, Matthew. 2007. “The Illumination of Consciousness: Approaches to Self-Awareness in the Indian and Western Traditions.” Philosophy East and West 57 (1): 40–62. McClintock, Sara. 2010. Omniscience and the Rhetoric of Reason: Śāntarakṣita and Kamalaśīla on Rationality, Argumentation, and Religious Authority. Somerville, MA: Wisdom Publications. Patil, Parimal. 2009. Against a Hindu God: Buddhist Philosophy of Religion in India. New York: Columbia University Press. Ratnākaraśānti. 2002. “Madhyamakālaṃkāropadeśa (dbu ma rgyan gyi man ngag).” In Bstan ’gyur (dpe bsdur ma), edited by krung go’i bod kyi shes rig’jug lte gnas kyi bka’bstan dpe sdur khang, translated by Śāntibhadra and Śākya’od, vol. 78 (sems tsam hi), 604–25. Beijing: Krung go’i bod kyi shes rig dpe skrun khang. Śākyabuddhi. 2002. “Pramāṇavārttikaṭīkā (tshad ma rnam ’grel kyi ’grel bshad).” In Bstan ’gyur (dpe bsdur ma), edited by krung go’i bod kyi shes rig ’jug lte gnas kyi bka ’bstan dpe sdur khang, translated by Subhutiśrī and Dge ba’i blo gros, vols. 98–99 (tshad ma je; nye), 955–1802 and 1–714. Beijing: Krung go’i bod kyi shes rig dpe skrun khang. Steinkellner, Ernst, ed. 2005. Dignāga’s Pramāṇasamuccaya, Chapter  1: A  Hypothetical Reconstruction of the Sanskrit Text with the Help of the Two Tibetan Translations on the Basis of the Hitherto Unknown Sanskrit Fragments and the Linguistic Materials Gained from Jinendrabuddhi’s Ṭīkā. www.oeaw.ac.at/fileadmin/Institute/IKGA/PDF/forschung/buddhismuskunde/dignaga_PS_1.pdf. Yiannopoulos, Alexander. 2012. “Luminosity: Reflexive Awareness in Ratnākaraśānti’s Pith Instructions for the Ornament of the Middle Way.” M.A. Thesis, Rangjung Yeshe Institute at Kathmandu University. ———. 2020. “The Structure of Dharmakīrti’s Philosophy: A  Study of Object-Cognition in the Perception Chapter ( pratyakṣapariccheda) of the Pramāṇasamuccaya, the Pramāṇavārttika, and Their Earliest Commentaries.” PhD diss., Emory University.

403

24 CANDRAKĪRTI Gardener of Sky-Flowers1 Mattia Salvini

The Teacher, Gautama, and the perfect Buddhas that abide in the same Dharma-nature; the entire host of the Victors’ sons, and the Dharma they have spoken – to these I prostrate, and then bow to him, the eye for embodied beings to see the Buddha’s Words without end, he who composed this Madhyamaka, out of compassion – Nāgārjuna.2 Candrakīrti (attributed), Praise of the Madhyamaka Treatises

Introduction Candrakīrti (ca. seventh century) is a prime exponent of Madhyamaka thought, the Buddhist “Middle Way” system free from the extremes of existence, nonexistence, eternalism, and materialist denial of rebirth. He presents his philosophy primarily in the context of commenting upon Nāgārjuna’s work, to defend his Mahāyāna view through reason and appeal to authoritative texts, against an array of opposing interpretations. He moves with clarity and ease between scholastic technicalities and pithy instructions to point at the inconceivable nature of things and infuses his arguments with literary beauty and suggestion. Candrakīrti speaks like a debater, a monastic, and a meditator. Little independent historical evidence regarding his life survives, but these features of his personality are represented in later hagiographical accounts (Chimpa, Chattopadhyaya, and Chattopadhyaya 1990, 198– 209). Blurring the boundaries between exegesis, literature, philosophy, and advice to those on the path, his writings present a vision of reality through metaphors, images, and other literary devices as well as through explicit argumentation. In this vein, Candrakīrti’s clarity lies in how he prepares and affects the interlocutor’s mind, making it receptive to key Madhyamaka themes presented in striking and ingenious ways. 404

DOI: 10.4324/9781351030908-35

Candrakīrti

Candrakīrti’s Nāgārjuna The “Madhyamaka” system (literally, “relating to the Middle”) is named after Nāgārjuna’s Root Stanzas of the Madhyamaka (Mūlamadhyamakakārikā). Candrakīrti is a Mādhyamika philo­ sopher who presents his work as the elucidation of Nāgārjuna’s intent. He poetically praises Nāgārjuna by explaining the two parts of his name: he is a water-dragon (nāga) who “gained birth in the ocean of the wisdom of the Perfect Buddha”; that is, he is a bodhisattva with direct insight into reality. Like the epic hero Arjuna, he is an impeccable archer, vanquishing saṁsāra with the arrows of his insightful speech (MacDonald 2015, Vol. II, 7; Salvini 2016, 419–21). Candrakīrti writes that Nāgārjuna composed his work in order to explain the system of the Perfection of Wisdom, an important set of Mahāyāna sūtras (MacDonald 2015, Vol. I, 116–17), out of compassion for sentient beings who would have difficulty understanding it. This determines the focus of Candrakīrti’s philosophy on the harmony between basic Buddhist categories relating to sentient experience and the illusory nature of that very experience  – the recurrent, perhaps even main, theme of the Perfection of Wisdom literature. This literature has been misunderstood, according to Candrakīrti, by Yogācāra philosophers (Li 2015). Candrakīrti agrees with his opponents that the mind has causal prominence when compared to non-sentient causal factors. On the other hand, he argues that the Yogācāra view of an ontological primacy of the mind is an untenable reification of awareness, incompatible with the intent of either the sūtras themselves or of Nāgārjuna’s treatises.

Candrakīrti’s main philosophical works The Introduction to the Madhyamaka (Madhyamakāvatāra) (de La Vallée Poussin 1912; Huntington 1989; Li 2015; Lasic, Lee, and MacDonald 2022), together with its autocommentary, is a comprehensive outline of the bodhisattva path, organized in accordance with the Mahāyāna gradation of ten levels (bhūmi) – from the entrance into the state of “Noble Individuals” on the first level (when a practitioner is said to see reality directly for the first time) to complete buddhahood. Within this text, the sixth chapter, the longest of all, is devoted to the cultivation of the Perfection of Wisdom and constitutes an influential presentation of Candrakīrti’s Madhyamaka understanding. Its definitions of the two truths is of particular significance and informs the writing of later Madhyamaka authors who followed in its wake, such as Prajñākaramati and, arguably, Śāntideva. While the Introduction to the Madhyamaka is not quite literally a commentary on any of Nāgārjuna’s works, it is still an exegesis of his thought, as Candrakīrti implies at the beginning of chapter six. This section contains a long refutation of the Yogācāra view of mind. According to Candrakīrti, Yogācāra philosophers mistook the “-only” in the Buddha’s expression “mind-only” (citta-mātra) to be a refutation of materiality, implying an ontology wherein only mental entities are real. In contrast, Candrakīrti argues, that expression was meant by the Buddha as a refutation of any other “agent” (kartr. ) apart from the mind, especially, in terms of an agent of creation, that is, an omnipotent singular creator. For Candrakīrti, the mind is the agent of creation, that is, the creator; at the beginning of each cosmogonical cycle, the collective accumulated karma of sentient beings in that particular universe determines the formation of the environment that they will inhabit – a standard Buddhist perspective on cosmogony. His commentary In Clear Words (Prasannapadā; de La Vallée Poussin 1903–13; Ruegg 2002; MacDonald 2015) is a detailed gloss on all the chapters of Nāgārjuna’s Root Stanzas on the Madhyamaka. The title is an indication of Candrakīrti’s preferred style. The term prasanna may refer to clear, well-settled water, unagitated by “the wind of speculation” (tarka-anila) – an allusion to his opponent Bhāviveka’s fondness for presenting arguments in a manner clearly 405

Mattia Salvini

influenced by Dignāga’s innovations in the formalized articulation of inferences. Candrakīrti did not endorse the new format, as he rather preferred Buddhapālita’s own focus on pointing out unwanted consequences (a sort of reductio ad absurdum). The extent to which this diffe­ rence in argumentative style may have any broader ontological implications is a matter of considerable debate in the later Madhyamaka tradition (see, for example, Dreyfus and McClintock 2003). This commentary refers to the Introduction to the Madhyamaka several times, which therefore must have been composed earlier. The commentary on Nāgārjuna’s Sixty Verses on Reasoning (Yuktiṣaṣṭikāvr․ tti; ScherrerSchaub 1991; Loizzo and The AIBS Translation Team 2007; Li and Ye 2014) is a helpful complement, offering several additional clues to better understand Candrakīrti’s thought on a number of important and difficult points in the interpretation of Nāgārjuna. In particular, Nāgārjuna’s text contains a verse (34) that has been read by a number of Yogācāra authors as supporting an interpretation of Nāgārjuna according to their own philosophy. Candrakīrti offers a completely different understanding of this key passage, in no way suggesting that it could support a “mind-only” reading. The Commentary on the Seventy Verses on Emptiness (Śūnyatāsaptativr․ tti; Erb 1997) is a relatively short commentary. In the root text, Nāgārjuna clarifies his understanding of how conventional dependent arising ( pratītyasamutpāda) relates to ultimate non-arising (anutpāda) and is thus of great importance in assessing the sense in which Nāgārjuna declares that there is no difference between saṁsāra and nirvāṇa. Candrakīrti’s commentary engages with Abhidharmic matters and shows, for example, his acquaintance with the doctrine of six causes (Erb 1997, 242). Āryadeva is one of the early great masters of Madhyamaka, reputed to be a direct student of Nāgārjuna. Candrakīrti composed an elaborate commentary on his Four Hundred Verses on the Yoga Practice of the Bodhisattvas (Bodhisattvayogācāracatuḥśatakaṭīkā; Shastri 1914; Suzuki 1988, 1989, 1994; Tillemans 1990; Lang 2003; Johnson 2012). Since the root text is itself a very wide-ranging philosophical treatise, this commentary offers insights into aspects of Candrakīrti’s thought that go beyond more habitual topics: kingship and its likely downfalls, the importance of contemplating death, the necessity of avoiding harmful actions, the unreliability of romantic attraction, and so on. It also offers additional glimpses of Candrakīrti’s vast erudition and establishes once again his familiarity with Vasubandhu’s works. The Treatise on the Five Aggregates (Pañcaskandhaprakaraṇa; Lindtner 1979; Li and Kanō 2015) is an independent short manual of Abhidharma. It follows the classification of entities into five realities (discussed subsequently) and yet does not perfectly match usual Sarvāstivāda presentations. This treatise engages with Abhidharmic analysis to show that, when brought to its logical conclusion, it culminates in the type of wisdom that establishes emptiness. A somewhat neglected work in contemporary scholarship (and some even doubt its authorship), it offers important clues in reconstructing Candrakīrti’s overall philosophical perspective, especially in order to understand his approach towards other strands of Buddhist thought. The Seventy Verses on Refuge (Triśaraṇasaptati) (Sorensen 1986; Skilling and Saerji 2012; Kanō and Li 2014) differs slightly from other works of Candrakīrti in terms of its main emphasis and confirms his familiarity with several Buddhist schools and his willingness to engage with their position. This work is also important for the Tibetan understanding of the meanings of the Three Jewels (the Buddha, the Dharma, and the Saṅgha) and of taking refuge in them, which is the foundational Buddhist vow common to lay practitioners and monks. This shorter treatise complements the Introduction to the Madhyamaka by offering a more basic discussion on the initial stages of Buddhist practice. Last, at least two important texts on Tantra are ascribed to Candrakīrti (Chakravarti 1984; Campbell 2009, 149–54; Hong and Tomabechi 2009). While this attribution is disputed by 406

Candrakīrti

modern scholars, it had a significant impact on the later reception and contextualization of his philosophy.

Candrakīrti’s philosophical grammar It is not indeed the case that words, as if they were people with clubs and snares, cause the person who speaks them to lose one’s freedom: rather, when they have the proper capacity, they function in conformity with what the speaker wishes to express.3 Candrakīrti, In Clear Words

Buddhist philosophers in South and Southeast Asia were active participants in the intellectual environment of their times (Skilling 2019, 214), wherein not only was Sanskrit the idiom of scholarly communication, but Sanskrit traditional grammar (vyākaraṇa, “analysis”) also defined a common conceptual language, being the first example of how to reflect on sound logical connections. Grammar affects the manner of philosophical exposition and argumentation and offers basic templates to conceptualize and express causal processes: sentential syntax is analyzed as expressing primarily an activity and the factors that participate in it (the agent, the object, the instrument, etc.). Thus, when the nature of causality and the limits of language are discussed, traditional grammar becomes relevant. Sanskrit traditional grammar involves linguistic analysis in a sense broader than in our common usage of the term “grammar”. The first role of Sanskrit grammar, when employed in philosophical texts, is to offer pertinent etymologies, clarifying the intent of sentences that have “great meaning” (mahārtha), while discarding incompatible exegetical positions. Another role of grammar, in Candrakīrti, is to discuss the implied presuppositions that make a linguistic expression comprehensible. This type of discussion often relies on the dialectic between an activity and the factors that make it possible (the agent, the instrument, the object, etc.), accor­ding to the analysis Sanskrit grammarians apply to sentences and even single parts of speech. One instance of such analysis is when Candrakīrti examines a sentence describing commonly perceived events in the world, such as motion. He shows that these events are illusory by discussing the grammatical implications of their formulation. For example, Nāgārjuna examines motion as follows: does motion necessitate an agent of motion in order to be comprehensible? If so, is that agent a “goer” or a “non-goer”? If the agent is a “goer,” by virtue of which action of going is the agent a “goer”? Is it the same action of going that he is then performing? Or a different one, that establishes the existence of a “goer” who is actually not going, but may then “go”? and so on. Candrakīrti expands on Nāgārjuna’s analysis by highlighting some of the conceptual underpinnings of the arguments, such as the insight that factors such as “agent” are “capacities” (śakti), that is, roles with respect to single specific actions rather than things (dravya) that may relate to multiple actions. This distinction, taken from the grammatical tradition, makes better sense of Nāgārjuna’s argument, where the root text states that once there are two actions of going, it follows that there must be two agents (an argument found in the second chapter of the Stanzas on the Madhyamaka, verse 6). If we took the term “agent” in the verse to refer to a continuing entity rather than to a role in respect to a specific action, Nāgārjuna’s argument would have no force. Alternatively, Candrakīrti may take up an instance of “bad philosophy,” wherein an opponent contradicts the shared presuppositions that make an expression comprehensible, in the hope of evading analysis and positing an irrefutable entity. Such “bad” linguistic expressions are akin to writing “this sentence is not in English,” hoping to avoid any possible criticism of the sentence’s content, by claiming that it is not analyzable any further. This is how 407

Mattia Salvini

Candrakīrti criticizes Dignāga’s account of particulars or, more literally, “own-characteristics” (svalakṣaṇa), according to which an ultimate momentary entity is simultaneously the instrument and the object of its own definition. The literal sense of “particular” implies that any particular is characterized by itself only and not by a property shared with anything else, since such momentary entity is literally called “own-characteristic,” and is absolutely unique. However, an expression like “own-characteristic,” when used in Dignāga’s sense, contradicts the very grammar that would make it comprehensible, since its form implies instrument and/or object, while Dignāga implies that its referent escapes any analysis in terms matching such grammatical categories. That is, Candrakīrti points out that Dignāga is employing words that imply syntactical roles while denying that he is using them as expressing those roles (“agent,” “object,” “instrument,”etc.; MacDonald 2015, Vol. I, 265; Vol. II, 262–63; see especially 262 n. 500). Candrakīrti’s usage of grammar to uncover presuppositions may be a distant echo of the grammarian’s principle that usage is taken “from the world” (lokataḥ, Joshi and Roodbergen 1986, 25). Candrakīrti appeals to expressive cogency, that is, to the necessary coherence between what the linguistic forms presuppose and what they are intended to express. This principle of linguistic coherence relies directly on his view that whatever exists can exist only in relation to something else. Furthermore, this principle is what makes such a view of reality expressible; it offers precious clues to understand what, for Candrakīrti, is the “virtuous” circle allowing the conceptual elements of philosophical reflection to function in support of an understanding of their own empty nature. Candrakīrti also appeals to grammatical categories when explaining the intent of Madhyamaka refutations. He does so by distinguishing between a negation that implies the acceptance of its opposite ( paryudāsa) versus a negation that simply negates without positing anything ( prasajya). This distinction originates in Sanskrit grammar and is used by other philosophers, too. Candrakīrti introduces it in the context of a broader discussion about the nature of language, which, according to him, does not oblige the speaker to uphold any particular ontological view. He uses this grammatical distinction of two types of negation in an ingenious way. He applies it to the context of philosophical debate to explain how a Madhyamaka thinker should refer to entities or positions that are the subject matter of refutation without being in any way committed to their existence. Sanskrit grammarians (some more explicitly than others) apply a methodological bracketing regarding the ontological status of the referents of those linguistic expressions that they are analyzing. Candrakīrti appeals to this general principle of linguistic analysis to highlight that a philosopher may argue without being forced into committing to any counter-position. Moreover, he employs the distinction between two types of negation to highlight that negating one type of ontological position does not imply being committed to an alternative one, even to the extent that negating that “something exists” does not entail upholding that “it doesn’t exist” – nor that “I am uncertain as to whether it does or it does not exist.” In the quote at the beginning of this section, Candrakīrti extends an insight derived from traditional grammar to bear upon the nature of language in general and more specifically of philosophical and argumentative language, making it possible to speak of reality and unreality in a uniquely Madhyamaka way. In this sense, Sanskrit traditional grammar is the key to understanding Candrakīrti’s views (or “non-views”) on ontology.

Candrakīrti’s Abhidharma Where the Guide explains the act of seeing as being due to an assemblage, The intelligent say that it is on the level of an approximation of the ultimate.4 From the Sūtra on Moving within Existence 408

Candrakīrti

Abhidharma is the Buddhist phenomenological investigation into entities (dharmas), wherein causally efficient entities are primarily categorized as either mental or material. Abhidharma furthermore discusses the relationship between those two spheres and analyzes how conceptualization of spatial and temporal extension arises in the mental continuum. “Abhidharma” is a very broad term, often referring, very generally, to the employment of a recurrent set of analytical categories within Buddhist texts of any genre; in other contexts, it refers to Buddhist philosophical schools who drew specific ontological conclusions regarding such categories. Candrakīrti quotes the above verse to elucidate the role of Abhidharmic categories – some of which he just explained to be ultimately illusory. The passage describes an intermediate step between more ordinary/common-sense ways of speaking, where one may say “I see through the eye,” and the ultimate truth of emptiness wherein there is no “seeing” to be spoken of. The verse represents the description of causal processes in terms of the Abhidharma: a moment of visual consciousness comes about due to a causal assemblage of the object, the sense-faculty, and a prior moment of consciousness; it does not involve a temporally extended “person” as the agent of seeing. This account is described in the verse as the “level of an approximation of the ultimate”: it is not an ultimate description, since it is not from the perspective of emptiness and still involves entities, and yet it is not simply tantamount to ordinary and common conventions. In other words, for Candrakīrti one goes from “I see something” (ordinary speech) to “visual consciousness arises due to causes and conditions” (Abhidharma account) to “neither seeing nor non-seeing could ever possibly take place” (ultimate view). While the first two steps may not represent ultimate reality in Madhyamaka terms, they are not treated as equal. That the last two steps are connected is also implicit in the Mahāyāna terminology of “two-fold selflessness”: “selflessness of person” refers to the level of Abhidharmic analysis, ascertaining that there is no personal identity beyond the impermanent constituents of mind and body; “selflessness of dharmas” means that these constituents are also illusory. Both are called “selflessness” and both go beyond the ordinary conventions, conceptualizing persons performing activities. Candrakīrti offers a rationale for viable conventions, applicable to both the ordinary and the more refined types, as discussed previously. For Candrakīrti, comprehensible linguistic expressions consist of mutually dependent elements, not isolated or self-referential items to be potentially coordinated later. This parallels his stance towards those defining traits that are employed in philosophical treatises in order to categorize entities. In Buddhist Abhidharma, definitions make entities recognizable for the purposes of “insight meditation” (vipaśyanā), a form of phenomenological refinement that for Candrakīrti culminates in the insight that those definitions are, in ultimate analysis, only provisionally viable and context bound. Thus, Abhidharma is linguistically viable, just like ordinary speech, but, being a more refined form of analysis, Abhidharma also goes beyond ordinary speech and offers the basic rationale for Candrakīrti’s investigations into ontology. This dual viability maps onto the Buddhist tradition’s general view that the Buddha spoke in terms of “two truths” conventional and ultimate. In the Abhidharma, the two truths are interpreted as pertaining to two levels of existence: existence only by way of a convention and ultimate existence that does not depend on conceptually created conventions. The two levels of reality are sometimes distinguished in terms of analysis. For instance, according to Vasubandhu, what can be broken into constituent parts, either in actuality (like a pot) or through analysis (like minuscule groups of atoms arising together), is only conventionally existent; what cannot be analyzed into more basic constituents (like atoms of materiality or moments of mind) is ultimately existent (Pradhan 1967, 334). This idea is so central to the Abhidharma project that in some ābhidharmika schools the term “part” (avayava) is a synonym for an ultimately existent entity, to be contrasted with the “part-possessor” or whole (avayavin) that is only a concept referring to those ultimately existing entities and does not 409

Mattia Salvini

exist on its own. This is unlike, for example, in the non-Buddhist Nyāya-Vaiśeṣika system, where the part-possessor is regarded as having a reality beyond and besides the parts. It is important to note that, according to the Abhidharma, when a part-possessor, such as a “table” is taken to be something real, this is not a perceptual mistake, that is, not the result of some sensory malfunction, but rather it is the result of a conceptual mistake, that is, the habit, whether innate or cultivated through inaccurate philosophical views, of taking one’s recurrent conceptual syntheses as matching something “out there.” Thus, while the Nyāya-Vaiśeṣika system may seem to be in harmony with a more “common sense” view of reality, this, for the Buddhist, is only in as much as it reifies the habitual mismatches between sensory and conceptual activity, and does not therefore represent “ordinary perception” in the sense of the way images, sounds, and so on appear to the mind of even ordinary, untrained people. Rather, it represents a theoretical solidification of concepts that ordinary people already may have about those appearances. The Abhidharma discusses the more fundamental level of reality in terms of defining traits, regarded to be ultimately valid. For example, materiality includes “four great elements” the first among which is the earth-element; the defining trait of the earth-element is “hardness.”5 The question then arises as to the ontological status of those defining traits: are they identical to (first solution) or different from (second solution) the entity that they characterize? Both solutions offered in the Abhidharma presuppose the reality of the characterized entity. Candrakīrti applies Nāgārjuna’s examination of the characteristic and the characterized entity to offer a solution discarding the ultimate reality of either the characteristic or the entity it characterizes (de La Vallée Poussin 1903–13, 172–73). Thus, Candrakīrti’s solution is that, while conventionally we may need to speak of both the earth-element and the hardness that characterizes it, it is precisely this need to conceptualize both that shows how they are correlatives with no ultimate, independent existence. They are not just two ways of speaking of the same entity: they are mutually dependent concepts that make our experience of entities possible while showing that there is no real entity in the first place once we cognize that they are not conceivable independently. Abhidharma analyses reach a level where further analysis requires taking into consideration the mutual dependence of the elements under analysis, which are therefore, for the Madhya­ maka, proven unfindable as separate independent entities and thus existent only by way of our concepts. This last step differs from mereological analysis of larger entities, that is, coarser conventions, which can take several steps before reaching the type of impasse that discloses the unreality, rather than the just the reducibility, of the object under examination. This gradation from ordinary speech to Abhidharma and eventually to ultimate emptiness can be observed, for example, in Candrakīrti’s analysis of the locus of motion, meant to show that motion is illusory. He starts by showing that the foot of someone who walks is not a real singularity, but rather, it can be analyzed into parts, starting from dividing it into toes and so on down to infinitesimal elements of existence. Those infinitesimal parts may be supposed to be the ultimate elements that make up the foot, and this is where the Abhidharma would stop. For the Madhyamaka, however, they can be further analyzed by applying a conceptual alternative: is it really partless? Then how does it produce the extended objects that we seem to perceive? Indeed, if it is partless, it is unextended, like a geometrical point, and it is therefore difficult to establish extension on the basis of a collection of unextended entities. This alternative relates to a more fundamental level than the conceptual proliferation of extended objects (de La Vallée Poussin 1903–13, 93). It deals with the conception of minimal entities existing without extension and thus not analyzable into parts. When collected into larger groups, these entities would then be the basis for the perception of extension, according the Abhidharma. To put it differently, the special feature of Abhidharmic categories (of the type that Candrakīrti endorses as valid analytical conventions) is that they are only one step removed from the 410

Candrakīrti

final step of analysis before the resultant “non-perception” by the mind. Concepts of coarser extended entities such as persons or tables are several steps removed from that finer step: they do not involve interdependence between parts or moments, and they do not even match the “selflessness of persons,” the level at which one speaks of causal processes but not anymore of personal identities. I can analyze a table first as a collection of parts of a table, such as the upper surface, the lower surface, the legs, and so on down to minimally extended entities in space – that is, partless elements (atoms in the Buddhist philosophical sense). At that point, however, the analysis will have to shift to a mode unavailable to the Abhidharma, revealing that the atom is untenable: analysis has exhausted itself in not finding the initial object of analysis anywhere. Another way to express this is to say that (most) Abhidharmic categories are viable ways to understand the causal elements of experience. On the other hand, the categories of the nonBuddhists are invalid analyses of ordinary perceptions and speech, since they rely on explanatory elements (such as a permanent “self ”) that exist neither conventionally nor ultimately, being posited as permanent and substantially real. Said another way, the non-Buddhists confuse sensory data and conceptual synthesis. In other words, non-Buddhist categories start from common experience and go in the wrong direction of reifying those experiences. Abhidharmic analysis, on the other hand, goes in the right direction of dissolving concepts about the reality of entities. While the Abhidharma schools may not be willing to continue the analysis until the very ideas of “perception” and “entities” are shown to be illusory, in some sense they have, according to Candrakīrti, done half of the Madhyamaka job. Candrakīrti highlights how the person-less categories of Abhidharma, the most basic of which are called “aggregates, bases, and entrances” (skandha, āyatana, and dhātu) are a ne­cessary didactic tool in disclosing the nature of things: Just like, although they had destroyed a sense of “I” and “mine,” the Buddhas, the Bhagavats, in accordance with the world, said “I” or “mine,” for in this way they could communicate the meaning that they wished to express; in the same way the aggregates, bases, and entrances were spoken of due to a task to be accomplished. For, without distinguishing those, there would be no means to introduce the world to reality (tattva). (Li and Ye 2014, 142, my translation) The last sentence is particularly important. Although the analysis of a “person” relies on bodily and mental constituents that are regarded as ultimately illusory, they need to be employed so as to show precisely that they, too, and not only the “person,” are illusory. In addition to miscellaneous passages where he takes a position in Abhidharmic debates, Candrakīrti composed an independent work offering an overview of Abhidharma, the Treatise on the Five Aggregates. The overall framework of this treatise is the Sarvāstivāda classification of entities into “five realities”: form/materiality; mind; mental states; causal factors that belong neither to materiality nor to the mind; and unproduced, permanent entities. Candrakīrti is acquainted with several Abhidharmic traditions. He can thus argue that even when Nāgārjuna’s analysis mentions only the categories familiar to a specific school, its applicability goes far beyond. For example, when Nāgārjuna analyzes causality in terms of “four conditions” in the first chapter of his Root Stanzas of the Madhyamaka, Candrakīrti explains that other classifications of causes and conditions can be subsumed under this four-fold scheme (Salvini 2014). This point is, I think, worthy of note: Candrakīrti is attentive to the specific technicalities of Nāgārjuna’s Abhidharmic underpinnings, and yet he is keen to show that the Madhyamaka refutations are meant to apply far more broadly. In the previous example, this implies that the 411

Mattia Salvini

refutation of the “four conditions” should be taken as a template to refute any variant of the view that a real effect arises from a real cause different from itself.

Candrakīrti and the sūtras Without understanding the signless, you said, there is no liberation: Therefore, that has been taught by you, in its entirety, in the Mahāyāna.6 Nāgārjuna, Praise of the One Beyond the World

In Clear Words is filled with quotes from both non-Mahāyāna (“sūtras from the Āgama”) and Mahāyāna sūtras (de La Vallée Poussin 1903–13, 548–49). Candrakīrti considers “emptiness” the ultimate intent of the Buddha’s teachings and argues that it has been taught in both Mahāyāna and non-Mahāyana texts. While the Mahāyāna sūtras explicitly use the term “emptiness,” Candrakīrti argues that the sense of non-Mahāyāna passages describing all things as illusory is precisely the same. Showing familiarity with non-Mahāyāna sūtras from different schools, Candrakīrti, like Nāgārjuna, is keen to offer a common ground for the Madhyamaka perspective. After quoting a very famous passage that he interprets as teaching emptiness, Candrakīrti remarks that “this is found in the treatises and the sūtras of all the schools” (de La Vallée Poussin 1903–13, 549). Candrakīrti mentions different ways in which the Buddha’s words had been collected by different schools (Sorensen 1986, 50–51). By taking into account how sūtras were preserved in different Buddhist lineages, his work displays his work displays an inclusive hermeneutics in his approach to texts from other Buddhist traditions. Writing with the awareness of a varied audience, Candrakīrti prioritizes rational argumentation over authoritative claims, and his hermeneutics allows for alternative positions on important topics, such as, for example, differing Abhidharma accounts of experience. This flexibility is likely connected to his attitude towards his sources – hermeneutics and reflection on the nature of experience become closely intertwined. Candrakīrti’s exegetical standard regarding the Buddha’s words relies on the familiar dichotomy between the interpretable meaning and the definitive meaning. He quotes two Mahāyāna sūtras to the effect that whichever passages teach emptiness are to be considered as articulating the ultimate meaning, while others, taught for the sake of introducing to the path and speaking in terms of persons, living beings, and so on, require interpretation (MacDonald 2015, Vol. I, 207–9). This exegetical principle determines his use of both Mahāyāna and nonMahāyāna sources when supporting arguments in favor of emptiness. For instance, the Rice Stalk Sūtra contains an elaborate discussion of dependent arising. Candrakīrti relates this sūtra to the overall structure of the Root Stanzas on the Madhyamaka (Salvini 2011), according to the following progression: first, the misunderstanding about causality and entities must be cleared away, which then allows dependent arising to be taught as the template to understand repeated birth, and this, in turn, results in the relinquishing of views. These three steps are therefore, for Candrakīrti, a common feature of the sūtras and of the foundational treatise of his system, which highlights a key feature of Madhyamaka thought: the proper understanding of causality goes hand in hand with a proper understanding of its illusory nature.

Two truths All entities bear two natures, obtaining their state by force of right seeing and false seeing; 412

Candrakīrti

the domain of those who see rightly is reality, the domain of those who see falsely is explained as conventional truth. Even those with false seeing are, indeed, accepted as being of two types: those with clear faculties and those with defective faculties. Compared to the cognition of those with good faculties, the cognition of those with defective faculties is considered to be false. What the world obtains as perceptible for the six faculties when they are not damaged, that is truth from the perspective of the world itself, while the rest, that is wrongly imagined, is false from the perspective of the world itself. Things thoroughly imagined on their own account by the Tīrthas, who are bewildered by the sleep of ignorance, and things imagined in illusions, mirages, and so forth have indeed no reality even according to the world.7 Candrakīrti, Introduction to the Madhyamaka As we have seen, the theme of two truths is well developed in Buddhist traditions preceding Candrakīrti. Not surprisingly, he relies on Abhidharmic discussions to interpret Nāgārjuna’s own two-truths system. According to this system, conventional truth must be relied upon in order to teach ultimate truth. I understand Candrakīrti’s presentation of the two truths as having three salient features: 1. only ultimate truth is truth in a primary sense, while conventional truth is truth only in a secondary and circumscribed sense (i.e., for those who cannot see ultimate truth); 2. Abhidharmic categories are valid conventions and a stepping stone towards ultimate truth; non-Buddhist categories are invalid even as conventions; 3. ultimate truth does not entail any positing of existence or non-existence (or of any other properties) as real properties. By applying Nāgārjuna’s refutations of entities, Candrakīrti demonstrates that if we accept the Abhidharmic principle that what cannot resist analysis exists only conventionally and not ultimately and apply it consistently, it follows that nothing exists ultimately. There is nothing that resists analysis: we saw this point earlier in the discussion about the interdependence between a characteristic and the entity that it characterizes, as well as in the examination of the locus of motion and the supposedly ultimate elements of materiality that make up a foot. It therefore follows that ultimate truth is not ultimate existence, and what is more, any predication of exis­ tence or non-existence makes sense only within a conventional context – one which relies on a network of concepts, and on the presupposition that perception really occurs and validates something. Thus, conventional “truths” are always based on a false superimposition. Candrakīrti harmonizes Nāgārjuna’s statements about the ultimate non-reliance on existence or non-existence with an ontological principle taken from the Abhidharma. This procedure goes beyond the limits of Abhidharmic analysis to show that while we may talk of “ultimate truth,” talking of “ultimate existence” is, at best, an occasional metaphor employed for a specific purpose; thus, the whole idea that conceptual conventions are predicated upon real existent entities collapses. The two truths are truths in a very different sense – the conventional truth is “true” for those who have not understood its falsity, whereas the ultimate truth 413

Mattia Salvini

is true for those who have realized the falsity of the first, and its truth consists precisely in realizing the falsity of conventional “truth.” Thus, it is first of all useful to recognize that only ultimate truth is, properly speaking, truth. The second point, that is, distinguishing between ordinary conventions, Abhidharmic conventions, and non-Buddhist conventions, is where Candrakīrti goes far beyond Nāgārjuna’s text that only speaks of the “truth of the conventions of the world” and the “ultimate truth.” Candrakīrti offers a more elaborate system, spelling out the reasons for such a subdivision. In Clear Words presents a detailed analysis of the expression “truth of the conventions of the world.” The term here translated as “conventions” (saṁvr. ti) is analyzed in three different ways: it can mean “complete obscuration” of the nature of reality; mutual dependence between entities; and convention, very broadly understood as linguistic expression, transactional usage, and behavior/attitude towards something (all possible meanings of conventions of the world or vyavahāra) (Salvini 2019). What I have translated as “world” (loka) is also a semantically rich term, and Candrakīrti clearly employs different nuances that are not easily captured by any single English term. First, loka may mean the “world” of ordinary experience. But it can also mean a “person” in that world, conceptually imputed upon the “five aggregates,” that is, the finer elements into which Buddhist texts analyze the mental and material constituents of living beings. Last, Candrakīrti takes loka to mean a “perceiver,” and this allows him to introduce a distinction not explicitly found in Nāgārjuna: some people in the world do not qualify even as loka, since they are not genuine “perceivers,” and these “false-perceivers” are of two types – those with impaired sense faculties and those with distorted views. We know from related passages on “false conventions” in the Introduction to the Madhya­ maka quoted previously that “those with distorted views” refers to non-Buddhist philosophers. Candrakīrti reiterates in several places that the scholastic definitions and categories used by non-Buddhist thinkers are to be refuted even conventionally, unlike (most of) the Abhidharmic definitions, that must be accepted as valid conventions. This point is raised even in the context of criticizing Bhāviveka’s preferred method of argumentation, by which the latter refutes certain principles of non-Buddhist views by arguing that they cannot exist “ultimately.” For Candrakīrti, the qualifier “ultimately” is not only redundant but misleading, for Mādhyamikas should refute non-Buddhist categories even at the level of conventional truth. The reasons for this have been highlighted previously: non-Buddhist categories go in the wrong direction, reifying ordinary perceptions instead of refining them towards an understanding of emptiness. For instance, neither the Vaiśeṣika nor the Sāṁkhya systems can make sense of continued causality involving sentient “persons” without positing a permanent self, which is a mistake at the level of conventions themselves. For the Vaiśeṣika, the self is not sentient, but it is the agent of all actions. For the Sāṁkhya, the self is not active, but it is the only sentient element, which prompts everything else into activity. In both cases, the entire system, and many of the categories within it, stand in some necessary relationship with the permanent self and are in fact geared towards a direct realization of the permanent self on the part of the Vaiśeṣika or Sāṁkhya practitioner. This means that without a permanent self, these two systems would collapse. Since a permanent self is a mistaken concept even conventionally, all the other cate­ gories in their systems cannot be accepted even conventionally. The refutation of the Cārvāka system, which also does not accept a permanent self, works differently. The Cārvāka are “materialist” in the specific sense that they consider sentience derivative and dependent upon the material constituents of the body, so that the continuum of consciousness for them is “cut-off ”at death – from which they take the appellation, in Buddhist texts, of “upholding cutting-off ” (ucchedavādin). This is one of the two extremes that Buddhist teachings steer away from, even according to the very first line of the Stanzas on the 414

Candrakīrti

Madhyamaka, and indeed is often cited as the worst possible mistake. Candrakīrti imagines an opponent raising the following objection against the Madhyamaka: if, for the Madhyamaka, in ultimate analysis, there is no rebirth, no karma, no person, and so on, isn’t Madhyamaka a type of “nihilism” (nāstika), that is, not different from the Cārvāka? Candrakīrti answers by explaining that the Madhyamaka accepts rebirth and so on at a conventional level, and the same statement “there is no rebirth” has therefore a completely different sense when uttered by a nihilist as opposed to a Madhyamaka. Furthermore, if someone were to say that “ultimately, it is the same” since neither birth nor re-birth exist ultimately, Candrakīrti would respond that this does not matter, as there is a difference in the cognizer. The example he offers is that of two people who both report that a third person has committed theft; one of them has witnessed the theft, and the other has not. Thus, although what the sentence corresponds to may be true, one of the speakers is uttering a lie, for the context of the utterance implies that they have directly witnessed the event. Those who deny rebirth both ultimately and conventionally resemble the liar of that story: although they have not experienced the nature of things, they imply that they have by speaking about it. When we consider that for Candrakīrti, the level of rebirth and so on is the level of Abhidharmic analysis, we can see that the passage once more emphasizes the necessity of going through the Abhidharmic step on the way to the ultimate analysis. We could also use a different example: stating the conclusion that entities are empty without understanding their interdependence in Abhidharmic terms is akin to writing the correct result of a mathe­ matical problem by chance, due to a wrong procedure and a series of calculation mistakes. Candrakīrti does not spell out the reasons for rejecting the nihilist view. He even states that Nāgārjuna does not go into a detailed refutation of the nihilist view, because it had already been successfully refuted in the Abhidharma (Scherrer-Schaub 1991, 121–22). This may indicate one of the reasons Candrakīrti does not offer any detailed refutation of the materialist view at the conventional level; even most of his own opponents would not have taken that view very seriously, and Abhidharmic analysis is perfectly sufficient to discard materialist misconceptions. It should also be noted that a materialist perspective would not be easily harmonized with the Madhyamaka view of conventional reality as consisting of ultimately empty concepts. The materialist (in the sense delineated previously) gives an epistemic and ontological priority to the material substratum of concepts, making all concepts existentially dependent upon the material substratum, and this may not square well with Madhyamaka views on interdependence. In other words, this type of materialism could be accused by the Mādhyamika of falling into a type of essentialist reification of matter, avoided by the Abhidharma dualist classifications and by its refinement in phenomenological observation. Another way to understand the two truths is in terms of mistaken perception, unmistaken perception, and non-perception. Mistaken perception, such as the perception arising due to defective sense-faculties, or the conceptual superimpositions of the non-Buddhist (a self, a primordial “nature,” etc.), is a wrong type of convention and thus does not even qualify as a conventional “truth.” Unmistaken perception, such as the perception that arises from well-functioning sensefaculties, as well as ordinary speech and the analytical categories of the Abhidharma, that are in harmony with unmistaken perception, constitutes conventional truths. At the level of ultimate truth, however, perception is itself discarded as a mere convention that cannot withstand analysis, and thus Madhyamaka authors (including Candrakīrti) often speak of “non-perception” and “non-seeing” as the end result of Madhyamaka analysis. This gradualism may allow us to make better sense of the role of the Abhidharma within Candrakīrti’s system: the Abhidharma offers the most refined possible “truth” as long as truth relates to perception, and going beyond the Abhidharma means discarding perception as a parameter of truth altogether. Thus, while it cannot be said that Abhidharmic categories are ultimate, their status cannot be compared to that 415

Mattia Salvini

of the ordinary conceptualization of extended entities, such as chariots and persons, which are liable to be analyzed into constituent parts even within the context of perception. Abhidharma, furthermore, plays a key role in Candrakīrti’s epistemology of the two truths. It should be noted that Candrakīrti distinguishes between the functioning of the sense faculties and further conceptualization. It should be clarified, moreover, that Candrakīrti counts the “sense faculties” as six rather than five, that is, the five well-known sense perceptions and mental perception or the faculty for the mind to know objects. Mental perception and further conceptualization, therefore, should be distinguished, and Candrakīrti does so by distinguishing between the “object of perception” (grāhya) and the “object of conceptualization” (vikalpita). This helps us understand in which sense Abhidharmic categories can still be said to relate to (mental) perception, while non-Buddhist categories (such as the part-possessor, or the “self ”) are excluded from it (despite apparently conforming and expanding upon what ordinary people may take to be “perception”). In other words, it seems that for Candrakīrti, there is a clear difference between “the perception of ordinary people” and “what ordinary people may take to be perception,” since ordinary people may have very little ability to distinguish between the cognitions that result from the functioning of the six faculties and the further layers of conceptualization that occur on that basis. We should also not assume that for Candrakīrti all “ordinary people” are exactly at the same level in all respects; the term “ordinary people,” rather, here refers to anyone who does not have a direct realization of reality, irrespective of the refinement and soundness of their view. There is another sense in which Abhidharmic analysis, as opposed to ordinary speech, may be said to be “on the level of an approximation of the ultimate”: it emphasizes dependent arising. Candrakīrti’s own definition of conventional truth, as we have seen, is presented in Abhidharmic terms, relying on the scheme of six sense-faculties. This scheme, as Candrakīrti’s commentary clarifies, is closely connected with the analysis of causality: in this case, the object (visible entities) and the sense-faculty (the eye) are the cause for the arising of a specific cognition (visual consciousness). This framework is an “approximation to the ultimate” because it is precisely dependence, when extended so as to include the mutual dependence of conceptual correlatives, that discloses the emptiness of all entities, that is, their unfindability as independent ultimates. Abhidharmic categories are the results of the analysis of experience in terms of dependent arising based on perception, while Madhyamaka analysis is based on an extended application of the same principle, wherein the distinction between notional and existential dependence collapses. Thus, for the Madhyamaka, it is not only true that “effects depend on causes” but also that “causes depend on effects,” since there could be no notion of a cause without having in sight the notion of an effect. For example, a father is not a “father” until we know that he has children. This, once again, makes causes and effects viable as conventions precisely because they allow for a step of analysis that shows their mutual dependence and thus their ultimate emptiness. If causes or effects were to be non-empty, it would entail their having a non-contingent essence precluding their functioning within causal processes through time. Accepting the categories of the Abhidharma as a finer level of analysis does not entail discarding all expressions referring to spatio-temporally extended entities; these would remain “valid fictions,” that is, valid conventional truths, as their bases remain a finer level of perception at the border with non-perception rather than a philosophical reification of sensory objects. Ordinary speech is not in contradiction with Abhidharmic categories; rather, it is a helpful shorthand in ordinary interactions. Thus, Mādhyamikas argue that for something to be dependently arisen, it must be empty of independent existence, that is, of the type of ultimate existence that entails fixity of nature and precludes participation in causal change. More directly, Candrakīrti says that the term 416

Candrakīrti

“emptiness” refers to dependent arising rather than to non-existence. He suggests that someone may even say that emptiness “exists,” by a conventional superimposition, in order to help those listeners who may otherwise become alarmed (de La Vallée Poussin 1903–13, 264). I understand this “conventional superimposition” as being based on a common feature of emptiness and existence: both are not non-existence. Candrakīrti’s metaphorical qualification of emptiness forcefully clarifies that his view of ultimate truth does not entail non-existence. Rather, emptiness or non-arising is the very nature of all entities, since “it is never otherwise,” “it does not depend on something else,” “is not fabricated,” and this ultimate nature “neither exists nor doesn’t exist,” for, as we have seen, existence and non-existence cannot be ultimate properties of ultimately ascertainable entities.

Candrakīrti’s epistemology ( pramān a) Candrakīrti discards Dignāga’s system of two-fold epistemic validation (direct perception and subsequent inferential validation). Instead, he endorses a system of four valid sources of knowledge: direct perceptual evidence, subsequent inferential validation, tradition, and comparison (MacDonald 2015, Vol. I, 274–75). These four epistemic sources are also accepted as valid in the Nyāya system. We should, however, be cautious in understanding this as implying that Candrakīrti is adopting non-Buddhist philosophical categories. This epistemic scheme may have been perceived to be at a mundane enough level to be acceptable by a number of different philosophical traditions (as they would all accept, for example, the medical system of tridoṣa, occasionally referred to in philosophical treatises).8 Furthermore, Dignāga did not exactly discard tradition and similarity but rather showed how they should be understood as instances of inferential validation. He explained “subsequent inferential validation” as the conceptual ascertainment that a certain perceived logical mark is concomitant with a nonperceived, inferable entity. Thus, the inference that “there is fire on that hill, because there is smoke” is “subsequent validation,” because it rests on a prior perceptual ascertainment of the concomitance between smoke and fire. Tradition can function as a valid source of knowledge only when this rests on a valid inference regarding the reliability of the source of that tradition: for example, the Buddha must be proven to offer reliable reports even when facts can’t be independently ascertained by direct perception, and if he cannot be proven thus reliable, his textual tradition will not be a valid source of knowledge. Dignāga’s reduction of other epistemic means to just two, that is, direct perception and inferential validation, is predicated upon an ontology of the particular (lit. “own-characteristic”) vs the universal (lit. “commonality-characteristic”) rejected by Candrakīrti on grounds discussed earlier. Accordingly, the reduction of all possible epistemic means to two and only two is not justified, and it should not be surprising that Candrakīrti would re-instate two additional epistemic sources, in agreement with some pre-Dignāga Buddhist texts that also followed a scheme of four epistemic means (MacDonald 2015, Vol. II, 289–91, n. 541). Even though Candrakīrti rejects Dignāga’s epistemic framework, reflections on the nature of know­ ledge play an important role in his philosophy. We have seen that Candrakīrti’s own definition of valid conventional truths focuses on perception and is primarily concerned with unimpaired cognition. Nonetheless, Candrakīrti discards the possibility that the perception of ordinary people could function as a reliable epistemic means to validate ultimately existing entities. Some later Tibetan Mādhyamikas interpret Candrakīrti so as to harmonize his system with a significant part of Dignāga’s epistemological innovations. This is achieved by pointing to passages where Candrakīrti speaks of two possible objects of knowledge, direct and indirect, and thus qualifies his refutation of a two-fold epistemic scheme as pertaining only to the 417

Mattia Salvini

ensuing ontological claims and not to the two epistemic means in themselves (Ruegg 2000, 272–81; Vose 2009, 9; Thakchoe 2013).

Conclusions This overview of Candrakīrti’s thought offers tentative solutions to some difficult points in the interpretation of his system and situates itself among long-standing debates about Candrakīrti’s intent. Several aspects of Candrakīrti’s philosophy deserve more attentive representation in English, starting from his engagement with the Abhidharma and with non-Mahāyāna sources in general. To some extent, both later tradition and modern scholarship have tended to focus on Candrakīrti’s relationship with a specific epistemological tradition rather than unpacking his employment of the Abhidharma and its basic categories. This may not be a significant concern for a readership well trained in Abhidharmic details and their implications, but modern rea­ ders may need the Abhidharmic underpinnings of Candrakīrti’s system to be more explicitly spelled out. This is even more so in the case of Sanskrit grammar and for the likely implications of his literary style. Candrakīrti’s philosophical concerns may fruitfully overlap with those of some contemporary Western philosophers and be (partly) comprehensible in terms of ontology, epistemology, or logic. Yet, Candrakīrti’s intellectual project has a different horizon, different priorities, and a very different conceptual idiom, not always fitting so easily and naturally into those categorizations. While he offers cogent and sophisticated insights into the nature of reality, he does so independently of, and very differently from, Western intellectual modes (especially scientific empiricism) and may be called philosophical in a rather different sense, or at least a broader one, than what many contemporary readers may have come – perhaps somewhat narrowly – to expect. Thanks to his massive erudition and literary refinement, Candrakīrti is a bright window into the world of South-Asian thought. Candrakīrti delves into the conceptual grammar of Buddhist thought, at the limits of what we can speak of, primarily with respect to “exis­tence” and “nonexistence,” perception and non-perception, and their mutual dependence. As an especially lucid and engaging exponent of the Madhyamaka, his critique of essentialism has, arguably, few matches in any tradition. His reflections on the nature of language and conceptualization are counterintuitive, provocative, and deeply rewarding.

Notes 1 Acknowledgements: I  thank Guru Bhikṣu Vāgindraśīladhvaja for important clarifications on key points of the Madhyamaka view. Part of this work was funded by the Australian Government through the Australian Research Council Discovery Projects funding scheme, within the project “Ancient Today” (DP DP220100370); I thank the ARC, and my team-mates – especially, Yasmin Haskell. 2 śāstāraṃ praṇipatya gautamam ahaṃ taddharmatāvasthitān | sambuddhān sakalaṃ jinātmajagaṇaṃ dharmaṃ ca tair bhāṣitam | cakṣurbhūtam anantabuddhavacanasyālocane dehinām | yo ‘muṃ madhyamakaṃ cakāra kṛpayā nāgārjunas taṃ name || Madhyamakaśāstrastuti 14, Pandey 1994, 155. 3 na hi śabdāḥ dāṇḍapāśikā iva vaktāram asvatantrayanti, kiṃ tarhi satyāṃ śaktau vaktur vivakṣām anuvidhīyante || MacDonald 2015, 163. 4 sāmagryā darśanaṁ yatra prakāśayati nāyakaḥ | prāhopacārabhūmiṁ tāṁ paramārthasya buddhimān || Bhavasaṁkrāntisūtra, de La Vallée Poussin 1903–13, 120. 5 See, for example: kharaḥ pr. thivīdhātuḥ | Abhidharmakośabhāṣya, Ejima 1989, 12; bhūtānāṁ catuṣṭ vakhakkhaṭādilakṣaṇāvadhāraṇāt sūtre | Abhidharmakośabhāṣya, Ejima 1989, 37; Candrakīrti also refers to hardness as the defining trait of the earth element: iha bhāvānām anyāsādhāraṇam ātmīyaṁ yat svarūpaṁ tat svalakṣaṇaṁ, tadyathā pṛthivyāḥ kāṭhinyaṃ . . . MacDonald 2015, Vol 1, 248.

418

Candrakīrti 6 animittam anāgamya mokṣo nāstīti tvam uktavān | atas tvayā mahāyāne tat sākalyena deśitam || Lokātītastava 27, Lindtner 1982, 138. 7 samyagmṛṣādarśanalabdhabhāvaṃ rūpadvayaṁ bibhrati sarvabhāvāḥ | samyagdṛśāṁ yo viṣayaḥ sa tattvaṁ mṛṣādṛśāṁ saṁvr. tisatyam uktam || mṛṣādṛśe ’pi dvividhā matā hi spaṣṭendriyā doṣavadindriyāś ca | sadindriyajñānam apekṣya mithyā jñānaṁ mataṁ doṣavadindriyāṇām || vinopaghātena yad indriyāṇāṁ ṣaṇṇām api grāhyam avaiti lokaḥ | satyaṁ hi tal lokata eva śeṣaṁ vikalpitaṁ lokata eva mithyā || ajñānanidrāpracalāyamānais tīrthair yathāsvaṁ parikalpitā ye | māyāmarīcyādiṣu kalpitāś ca teṣām atattvaṃ khalu lokato ’pi || Madhyamakāvatārakārikā 6.34 – 26, Li 2015, 7 and see Salvini 2019, 666 n. 12 (containing an emendation to Li’s edition also followed here). 8 See, for example, Abhidharmakośabhāṣya 3.44.

References Campbell, John R. B. 2009. “Vajra Hermeneutics: A Study of Vajrayāna Scholasticism in the Pradīpoddyotana.” PhD diss., Columbia University. Chakravarti, Chintaharan, ed. 1984. Guhyasamājatantrapradīpoddyotanaṭīkāṣaṭkoṭivyākhyā. Tibetan Sanskrit Works Series, no. 25. Patna: Kashi Prasad Jayaswal Research Institute. Chimpa, Lama, Alaka Chattopadhyaya, trans., and Debiprasad Chattopadhyaya, ed. 1990. Tāranātha’s History of Buddhism in India. New Delhi: Motilal Banarsidass (Reprint: First Edition, Simla 1970). de La Vallée Poussin, Louis, ed. 1903–13. Mūlamadhyamakakārikās (Mādhyamikasūtras) de Nāgārjuna avec la Prasannapadā Commentaire de Candrakīrti. St. Petersbourg: Académie Impériale des Sciences. ———. 1912. Madhyamakāvatāra par Candrakīrti. Traduction tibétaine. Bibliotheca Buddhica 9. St.Pétersbourg: Académie impériale des sciences. Dreyfus, Georges B. J., and Sara L. McClintock, eds. 2003. The Svātantrika-Prāsaṅgika Distinction. Boston: Wisdom Publications. Ejima, Yasunori, ed. 1989. Abhidharmakośabhāṣya of Vasubandhu, Chapter  1: Dhātunirdeśa. Tokyo: Sankibo Press. Erb, Felix, ed. and trans. 1997. Śūnyatāsaptativr.tti. Stuttgart: Franz Steiner Verlag. Hong, Luo, and Toru Tomabechi, eds. 2009. Candrakīrti’s Vajrasattvaniṣpādanasūtra (Vajrasattvasādhana). Beijing and Vienna: China Tibetology Publishing House/Austrian Academy of Sciences. Huntington, C. W., trans. 1989. The Emptiness of Emptiness. Honolulu: University of Hawaii Press. Johnson, Dennis A. 2012. Refuting the Conditioned: The Saṁskr. tārthapratiṣedha of Candrakīrti’s Catuḥśatakaṭīkā: Introduction, Translation and Summary together with a Critical Edition of the Tibetan Translation. Wien: University of Vienna (unpublished Magister der Philosophie thesis). Joshi, S. D., and J. A. F. Roodbergen, eds. and trans. 1986. Patañjali’s Vyākaraṇa-Mahābhāṣya. Paśpaśāhnika. Introduction, Text, Translation and Notes. Pune: University of Poona. Kanō, Kazuo, and Xuezhu Li. 2014. “Sanskrit Verses from Candrakīrti’s Triśaraṇasaptati Cited in the Munimatālaṁkāra.” China Tibetology 1: 4–11. Lang, Karen C., trans. 2003. Four Illusions: Candrakīrti’s Advice for Travelers on the Bodhisattva Path. Oxford: Oxford University Press. Lasic, Horst, Li, Xuezhu, and MacDonald, Anne, ed. 2022. Candrakīrti’s Madhyamakāvatārabhāṣya Chapters 1 to 5. Critically and diplomatically edited by Horst Lasic, Xuezhu Li & Anne MacDonald on the basis of preparatory work by Helmut Krasser. Beijing/Vienna: China Tibetology Publishing House/Austrian Academy of Sciences Press. Li, Xuezhu. 2015. “Madhyamakāvatāra-kārikā: Chapter 6.” Journal of Indian Philosophy 43 (1): 1–30. Li, Xuezhu, and Kazuo Kanō. 2015. “Critical Edition of the Sanskrit Text of the Munimatālaṁkāra, chapter 1 (fol. 48r4–58r5): Explanation of skandha, dhātu, and āyatana based on Candrakīrti’s Pancaskandhaka.” The Mikkyo Bunka 234. Li, Xuezhu, and Shaoyong Ye. 2014. Yuktiṣaṣṭikākārikā: Editions of the Sanskrit, Tibetan and Chinese Versions, with Commentary and a Modern Chinese Translation. Shanghai: Zhongxi Book Company. Lindtner, Chr, ed. 1979. “Candrakīrti’s Pañcaskandhaprakaraṇa.” Acta Orientalia 40: 87–145. Lindtner, Chr, ed. and trans. 1982. Nagarjuniana: Studies in the Writings and Philosophy of Nagarjuna. Copenhagen: Institut for indisk filologi. Loizzo, Joseph J., and The AIBS Translation Team, trans. 2007. Nāgārjuna’s Reason Sixty with Candrakīrti’s Reason Sixty Commentary. Edited by Robert A. F. Thurman, Thomas F. Yarnall, and

419

Mattia Salvini Paul G. Hackett. Critical Editions by Joseph John Loizzo and Paul G. Hackett. New York: American Institute of Buddhist Studies, Columbia University. MacDonald, Anne, ed. and trans. 2015. In Clear Words: The Prasannapadā, Chapter One. Vienna: Austrian Academy of Sciences Press. Pandey, J. S., ed. 1994. Bauddhastotrasaṁgraha. New Delhi: Motilal Banarsidass. Pradhan, Prahlad, ed. 1967. Abhidharma-Koshabhāṣya of Vasubandhu. Patna: K. P. Jaisawal Research Institute. Ruegg, David S. 2000. Three Studies in the History of Indian and Tibetan Madhyamaka Philosophy. Wien: Universitat Wien. ———, trans. 2002. Two Prolegomena to Madhyamaka Philosophy. Wien: Universitat Wien. Salvini, Mattia. 2011. “The Nidānasamyukta and the Mūlamadhyamakakārikā: Understanding the Middle Way Through Comparison and Exegesis.” Thai International Journal of Buddhist Studies 2: 57–95. ———. 2014. “Dependent Arising, Non-Arising, and the Mind: MMK1 and the Abhidharma.” Journal of Indian Philosophy 42 (4): 471–97. ———. 2016. “A Clear and Learned Guide in Reading Candrakīrti’s Prasannapadā.” Bulletin de l’École française d’Extrême-Orient 102: 415–35. ———. 2019. “Etymologies of What Can(not) Be Said: Candrakīrti on Conventions and Elaborations.” Journal of Indian Philosophy 47: 661–95. Scherrer-Schaub, Cristina A., ed. and trans. 1991. Yuktiṣaṣṭikāvr. tti. Commentaire à la soixantaine sur le rasoinnement ou du vrai ensaignement de la causalité par la Maître indien Candrakīrti. Bruxelles: Institut Belge de Haute Études Chinoises. Shastri, H., ed. 1914. “Catuḥśātikā of Āryadeva.” Memoirs of the Asiatic Society of Bengal 3 (8): 448–514. Skilling, Peter. 2019. “Many Lands of Gold.” In The Golden Land Suvarnabhumi. The New Finding for Suvarnabhumi Terra Incognita, edited by Bunchar Pongpanich and Somchet Thinapong, 191–217. Bangkok: GISTDA and BIA. Skilling, Peter, and Saerji. 2013. “Candrakīrti and the Pūrvaśailas: A Note on Triśaraṇasaptati v. 51.” Annual Report of the International Research Institute for Advanced Buddhology at Soka University for the Academic Year 2012 16: 267–72. Sorensen, Per K. 1986. Candrakīrti, Triśaraṇasaptati: The Septuagint on the Three Refuges. Wien: Wiener Studien zur Tibetologie und Buddhismuskunde. Suzuki, Kōshin. 1988. “A Study of Candrakīrti’s Bodhisattvayogācāracatuḥśatakaṭīkā II, A Critical Edition of the Tibetan Text with the English Translation.” Sankōbunka-kenkyūjo-nenpō 20: 1–26. ———. 1989. “A Study of Candrakīrti’s Bodhisattvayogācāra-catuḥśatakaṭīkā I, A Critical Edition of the Tibetan Text with the English Translation.” Taishōdaigaku Daigakuin Kenkyūronshū 13: 1–12. ———, ed. 1994. Sanskrit Fragments and Tibetan Translation of Candrakīrti’s Bodhisattvayogācāraca tuḥśatakaṭīkā. Tokyo: Sankibo Press. Thakchoe, Sonam. 2013. “Prāsaṅgika Epistemology: A  Reply to Stag tsang’s Charge Against Tsongkhapa’s Uses of Pramāṇa in Candrakīrti’s Philosophy.” Journal of Indian Philosophy 41: 535–61. Tillemans, Tom J. F., ed. and trans. 1990. Materials for the Study of Āryadeva, Dharmapāla and Candrakīrti. Wien: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. Vose, Kevin A. 2009. Resurrecting Candrakīrti. Disputes in the Tibetan Creation of Prāsaṅgika. Somerville, MA: Wisdom Publications.

420

PART 6

Middle-Period Commentators (Eighth–Ninth Century)

MIDDLE-PERIOD COMMENTATORS Introduction to Part 6

Following the development and articulation of Madhyamaka and Yogācāra as distinct and coherent philosophical schools, with thinkers from both traditions engaging Dignāga and Dharmakīrti, in the eighth century, we see efforts to take stock of the diversity of philosophical positions. These efforts took different approaches, some critical and some syncretic, to address the hermeneutical and philosophical issues that such diverging orientations and interpretations posed to the coherence of the Buddhist tradition. At times, this meant presenting a path leading through a hierarchy of views from the lowest to the most liberating wisdom. At other times, it meant offering new interpretations that could bend the meaning of root texts in creative ways. It could also involve the creation of new conceptual and exegetical tools aimed at solving problems posed by Buddhist and non-Buddhist opponents. We see one version of a philosophical synthesis together with the development of original philosophical positions in Śubhagupta (eighth century), who draws on the atomic theories found in Vaibhāṣika and Sautrāntika traditions even as he is firmly rooted in the logic and epistemology of Dignāga and Dharmakīrti. Serena Saccone explores how Śubhagupta, while mostly interpreting and elaborating on Dharmakīrti’s thinking, develops his own ontological and epistemological views. This is especially true in his Verses on the Demonstration of External Objects (*Bāhyārthasiddhikārikā), in which he defends a form of externalism in contrast to the ontological idealism of Dignāga and Dharmakīrti even as he employs their methods of argument. In the same text, Śubhagupta argues against both the Sautrāntikas and the idealists and defends the view that sense cognitions are not shaped by the image of their objects. Śubhagupta’s student, Dharmottara (eighth century), wrote his own independent treatises, but he is primarily known for his commentaries on Dharmakīrti’s writings on epistemology. David Nowakowski, in his chapter, focuses on Dharmottara’s commentary on Dharmakīrti’s Epitome of Reason (Nyāyabindu). While Dharmottara presents himself as faithfully interpreting Dharmakīrti’s thought, in resolving tensions and constructing a more coherent and systematic account, Dharmottara nonetheless introduces novel ideas. In particular, Nowakowski spotlights how Dharmottara’s epistemology and philosophy of mind blur Dharmakīrti’s clear distinction between perception, inferential reason, and inference-for-others. While he draws on Dharmakīrti’s resources, Dharmottara’s epistemology is much more cognizer relative, as he emphasizes the ways in which both perception and inference are themselves dependent on the particular conditions of individual cognitive agents. This rejection of Dharmakīrti’s DOI: 10.4324/9781351030908-37

423

Middle-Period Commentators

more impersonal account of cognitive processes proved influential for some later Buddhist philosophers. Ryusei Keira, in his chapter, discusses Jñānagarbha (eighth century), author of the Commentary on the Distinction Between the Two Truths (Satyadvavibhaṅgavṛtti). Following Bhāviveka and informed by Dharmakīrti’s epistemology, Jñānagarbha is perhaps the first Indian Buddhist philosopher to explicitly integrate Yogācāra views into a Madhyamaka framework. For example, Jñānagarbha employs the idea of “mind-only” (cittamātra) to argue for the selflessness of external objects. Keira argues that this moment of Yogācāra thought illustrates Jñānagarbha’s gradualism, as it is only one stage in what will ultimately be a Madhyamaka account that shows the selflessness of mind. Keira is particularly interested in Jñānagarbha’s account of the two truths, how it is different from the account of the two truths found in other prominent Madhyamaka thinkers, and the diverse Tibetan interpretations of Jñānagarbha’s distinction between the conventional and the ultimate. Śāntarakṣita (ca. 725–788) is said to have been Jñānagarbha’s student and wrote a subcommentary ( pañjikā) on his teacher’s Commentary. He also employs Dignāga and Dharmakīrti’s logic and epistemology and is the first prominent Mādhyamika philosopher to produce a commentary on a text by Dharmakīrti. Allison Aitken, in her chapter on Śāntarakṣita, traces his syncretic method that was to become influential for much Indian and Tibetan Buddhist philosophy. This syncretic method employs and then rejects one view after another, culminating in the Madhyamaka relinquishing of all views. For example, in his Ornament of the Middle Way (Madhyamakālaṃkāra), Śāntarakṣita rejects the direct realist theory of perception associated with Vaibhāṣikas by using arguments from the Sautrāntika representationalist theory of perception. But then he adopts the Yogācāra idealist ontology – which, together with the Yogācāra understanding of cognition as reflexively aware, he believes to be conventionally true – to critique that same Sautrāntika representationalism. In the end, while the Yogācāra position may be helpful for freeing us from misguided views of the reality of external objects, Śāntarakṣita employs it instrumentally and leaves it behind with his famous articulation of the neither-one-nor-many argument for the Madhyamaka ultimate truth of the emptiness of intrinsic nature, including the emptiness of the Yogācāra concept of mind. Continuing this lineage, Śāntarakṣita’s prolific student Kamalaśīla (ca. 740–795) shows even greater enthusiasm for integrating Yogācāra and Madhyamaka thought, often bolstered by elaborate arguments rooted in Dharmakīrti’s epistemological theories. In the following chapter, Sara McClintock presents an overview of Kamalaśīla’s oeuvre, showcasing his range from philosophical treatises to sūtra commentaries to meditation and monastic training manuals to epistles. Like his teacher Śāntarakṣita, Kamalaśīla is unusual in this volume in that we have historical records testifying to his existence as the result of his sojourn in Tibet, where he traveled at the invitation of the king and where he engaged in teaching and debate. Some of his works appear to have been written in Tibet and may also therefore represent his attempt to transmit the full range of scholastic knowledge of Nālandā, the famed monastic university in North India with which he is associated. Throughout his many writings, however, a consistent theme is the importance of understanding the conventional realities that allow reason to function. McClintock argues that Kamalaśīla sees even wisdom as a “necessary illusion” which the bodhisattva must embrace if they are to be of help to the world. Another of Śāntarakṣita’s students, Haribhadra (eighth century), also draws on diverse Buddhist philosophical traditions in a syncretic fashion. Pierre-Julien Harter, in his chapter, focuses on Haribhadra’s engagement with the Perfect Wisdom (Prajñāpāramitā) tradition. Haribhadra exemplifies those commentators who want both to appear subservient to texts and to manifest a creativity that opens new philosophical horizons. According to Harter, we see this 424

Introduction to Part 6

in the alliance Haribhadra forges between the sūtras of Perfect Wisdom and the Ornament of Realizations (Abhisamayālaṃkāra), of which he became a champion. This exegetical innovation is at the same time a philosophical innovation since it updates the Perfect Wisdom by giving it what Harter calls a “voice” and by giving to the Buddhist path a treatise (śāstra) and a place within philosophical debates. Śāntideva (eighth century) is the Indian Buddhist author whose work is most widely read by contemporary scholars as a contribution to moral philosophy. Both of Śāntideva’s surviving texts – the Introduction to the Practice of Awakening (Bodhicaryāvatāra) and the Compendium of Trainings (Śikṣāsamuccaya) – make significant contributions to understanding the moral dimensions of mindfulness, concentration, virtue, well-being, and the relation between metaphysics and ethics. Stephen Harris, in his chapter, is particularly interested in Śāntideva’s analysis of the virtues of generosity, patience, and wisdom and its role in the cultivation of the other virtues of a bodhisattva. He also focuses on bodhicitta, or “the mind dedicated to full awakening” and the vow to become a bodhisattva. Harris emphasizes the ways in which, throughout Śāntideva’s account of cultivating the virtues of a bodhisattva, we see the convergence of altruism and self-interest, that attending to the good of the other is simultaneously to attend to our own most fundamental good. In the final chapter in this section, Shinya Moriyama turns to another commentator on Dharmakīrti, Prajñākaragupta (eighth–ninth century). Moriyama focuses on the most distinctive elements of Prajñākaragupta’s thought in ontology, epistemology, and philosophy of religion. He begins with Prajñākaragupta’s theory of “backward causation,” according to which future entities function as causes of present entities. Moriyama then discusses Prajñākaragupta’s ontological idealism, which equates existence with being perceived and which informs his epistemology. Then Moriyama turns to Prajñākaragupta’s expansive account of the Buddha’s omniscience, including the knowledge of all past and future entities. He ends with a discussion of Prajñākaragupta’s doctrine of nonduality, the culmination of the Buddhist path of wisdom.

425

25 ŚUBHAGUPTA An Externalist Outsider Within the Dharmakīrtian Tradition Serena Saccone

Introduction Śubhagupta (Tib. Dge [b]srung[s]) was an eighth-century philosopher in what is referred to as the Buddhist tradition of logic and epistemology ( pramāṇa). Much of his intellectual work is related to the discussion of Dharmakīrti’s theories. However, unlike some figures in this tradition, he is not strictly an exegete of the master’s works. Rather, he systematizes, elaborates on, and, in some cases, even refutes Dharmakīrti’s doctrines. Śubhagupta’s thinking is distinguished by his syncretic method of drawing on the views of different Buddhist traditions to conceive his own original ontological and epistemological theories. This may be why his doctrinal affiliation was, and has continued to be, debated.1 Śubhagupta’s innovative views are especially apparent in the Verses on the Demonstration of External Objects (*Bāhyārthasiddhikārikā), his magnum opus, which remains his most valuable work by far. Particularly, in this text, he tries to validate, within the “logico-epistemological” tradition, a form of externalism and a form of nirākāravāda (namely the doctrine that cognitions are not shaped by the images of their objects). It is this specific theory on externalism (bāhyārthavāda) on the ontological level (as well as the nirākāravāda standpoint on the epistemological level) that is his most distinctive position. Śubhagupta’s novelty as an author is also reflected in his adopting the genre of short digests, arguably for the first time within the Buddhist epistemological tradition. He composed brief “manuals for debate,” organizing them by single subjects that were relevant at the time. These writings are indeed synopses of arguments and theories that are already found (in part or in whole) in works by previous Buddhist authors, especially Dharmakīrti. However, Śubhagupta articulates (and reworks) them as effective proofs to use against non-Buddhist opponents and to establish Buddhist truths. Most likely, these compendia were intended for pedagogical reasons, in order to instruct the monks. At the same time, they could also have been compiled with the aim of fostering the monks’ own rational conviction in the truths defended therein. This holds true especially with regard to Śubhagupta’s four other extant works, though less so for the Demonstration of External Objects. This “literary genre” was continued by subsequent authors, such as Jitāri (tenth century). As for specific contributions to the logico-epistemological tradition, Śubhagupta also innovated by stimulating further refinement of the theories of other Buddhist philosophers, both DOI: 10.4324/9781351030908-38

427

Serena Saccone

contemporaries and successors, who responded to his criticisms. One example of such innovation can be found in the “heterodox” views that are expounded in the Demonstration of External Objects. This text serves to refute some of the principal doctrines of what we could call the “mainstream” logico-epistemological tradition, particularly the “idealistic” (vijñānavāda) turn represented by the Vasubandhu (ca. 350–430) -Dignāga (ca. 480–540) -Dharmakīrti (ca. 550–660?) lineage. However, his other extant works contain little innovation in this regard and appear to follow the mainstream ideas of this lineage and the Dharmakīrtian tradition of the period. The Demonstration of External Objects has been the focus of study and investigation by a few contemporary scholars, but his four other works have garnered very little attention. This is probably the reason Śubhagupta is regarded as a fierce adversary of the views of the “mainstream” logico-epistemological lineage and particularly of Dharmakīrti. And yet, it would be a mistake to overlook his doctrinal debt to the latter.

Biography Little is known about Śubhagupta’s life. He certainly lived after Dharmakīrti and was most likely an earlier contemporary of Śāntarakṣita (ca. 725–788), as he refers to the former’s views and is quoted by the latter. There are different conjectures among contemporary scholars regarding his exact dates (on this, see Saccone 2018, 2019a). However, the dates provided by Frauwallner (1961, 147), that is, 720–80, are the most likely, because they take into account the possibility that he was Dharmottara’s (ca. 740–800) teacher. The Tibetan historian Tāranātha (1575–1634) reports that Śubhagupta lived at the time of Dharmapāla (ed. Schiefner 1868, 166, 15–17), whose reign has been dated ca. 775–812 (Sanderson 2009, 87). Accordingly, he most likely lived in the mid-eighth century, perhaps in northeastern India.

Intellectual Activity Singularity Within the Tradition: Śubhagupta’s Relation to Other Authors Based on Pocock’s theory of “contextualism” in defining an author’s intellectual activity, I believe that Śubhagupta’s singularity as a thinker must be determined primarily through the way his works and ideas were received, interpreted, and used within the tradition in which he was active (i.e., the Buddhist tradition of logic and epistemology). In other words, his intellectual biography requires delineating his unique contributions to the history of Buddhist thought. Such contributions are particularly evident in analyzing the writings of his contemporaries, such as Śāntarakṣita and Kamalaśīla (ca. 740–795). The two authors appear to follow in Śubhagupta’s footsteps, both – to some extent – doctrinally and formally, in their extensive and doxographical magna opera, the Compendium of Truths (Tattvasaṃgraha) and the Commentary on the Compendium of Truths (Tattvasaṃgrapañjikā), respectively. For instance, it seems plausible that, with reference to doctrines connected to the debates against Brahmanical opponents, Śāntarakṣita and Kamalaśīla followed Śubhagupta’s elaboration of arguments. This is the case with the proof of the possibility of omniscience for human beings, which is connected with the reliability of scriptures written by human authors. Omniscience here is construed as the knowledge of those truths that are the object of scriptures and are extrasensory. Śāntarakṣita and Kamalaśīla’s debt to Śubhagupta holds true also with regard to the refutation of a creator God understood as a real and permanent entity. 428

Śubhagupta

The number of parallel arguments emerging from the comparison of the Compendium and the Commentary and some of Śubhagupta’s writings, such as the Verses on the Demonstration of the Omniscient One (*Sarvajñasiddhikārikā) and the Verses on the Refutation of God (*Īśva­ rabhaṅgakārikā), is remarkable. If, as I  hypothesize (Saccone 2019b, Forthcoming), all of Śubhagupta’s works were composed before those two, then Śāntarakṣita and Kamalaśīla owe a great deal to their earlier contemporary Śubhagupta. This would be all the more evident in Kamalaśīla’s work. The latter thinker explicitly refers to Śubhagupta by name many times, calling him “venerable” (bhadanta) and “master” (*ācārya/slob dpon) in two different works. Some of the formulations and overall ideas that are present in the three authors are, nevertheless, part of a common background for many Buddhist thinkers; certainly, most of them are to be found, either condensed or in full, in some previous Buddhist authors, particularly Dharmakīrti. However, to the best of my knowledge, the specific way Śubhagupta articulates these notions to design proofs against non-Buddhist opponents is unique (Saccone 2019b, Forthcoming). This is particularly the case for the demonstration of the Buddha’s omniscience (Saccone Forthcoming). Accordingly, I  believe that the parallel arguments that are found in Śāntarakṣita’s and Kamalaśīla’s works betray the direct influence of Śubhagupta’s ideas. On the other hand, Śāntarakṣita and Kamalaśīla also make a point of refuting the epistemological and ontological views that are found in the Demonstration of External Objects. They do so especially in the Examination of External Objects chapters of the Compendium and the Commentary. Śāntarakṣita and Kamalaśīla are the representatives of a mainstream current within the logico-epistemological school that defended the view of “consciousness-only” (vijñaptimātratā). According to the latter standpoint, everything can be explained as the mere occurring of consciousness, one after the other. Śubhagupta, on the other hand, endorses an externalist position on ontology and epistemology, being the obvious target of their criticism. Kamalaśīla, in particular, shapes his defense of the Dharmakīrtian “invariable co-apprehension” (sahopalambhaniyama) argument (see next section and chapters of this volume on Dharmakīrti, Ratnakīrti, and Devendrabuddhi and Śākyabuddhi) as a refutation of Śubhagupta’s attacks on it, thus testifying to the significance of those attacks in the period. This provides additional evidence of how Śubhagupta’s ideas stimulated the further refinement of theories within the tradition. Śubhagupta’s main intent with his digests appears to be that of compiling “manuals for debate” – namely instructions on how to defeat the opponents as well as (perhaps) developing a personal and rational conviction in Buddhist truths. In this sense, he sets a precedent for the chapters of the Compendium and the Commentary, with their similar scope. That this was Śubhagupta’s goal may be confirmed by the third verse of his Demonstration of the Omniscient One. This verse serves as a declaration of intent informing at least four of his works, if not also his magnum opus: “And that fool who says ‘everything is produced by God (Īśvara)’ should be clearly told the following arguments, in an assembly in order to dispel ignorance.”2 Here, Śubhagupta is explicitly introducing his text as a summary of arguments to use against opponents in an assembly in order to dispel ignorance (perhaps to be intended as both that of the opponents and that of the Buddhist monks themselves). Among the other authors within the tradition of logic and epistemology, Śubhagupta is also doctrinally associated with Arcaṭa and Dharmottara. Dharmottara describes Śubhagupta (and Dharmākaradatta, another name for Arcaṭa) as a celebrated master and declares that his know­ ledge is, to some extent, the source of all well-conceived views expounded in his Commentary on the Ascertainment of the Means of Knowledge (Pramāṇaviniścayaṭīkā).3 To the best of my knowledge, there is no source of information regarding Śubhagupta’s relation with Arcaṭa. Regarding Śubhagupta’s association with Dharmottara, there are a few ancient sources that testify to their connection to a certain extent (on these sources, see Saccone 2019a, 459). In his 429

Serena Saccone

History of Buddhism, for instance, Bu ston (1290–1364) states that Dharmottara was the pupil of Arcaṭa and Śubhagupta.4 In modern scholarship, the subject of the relationship between Śubhagupta and Dharmottara as well as Arcaṭa, though it merits analysis, has not been thoroughly investigated (for a more extensive bibliography on this, see Saccone 2018, 2019a). In fact, Matsumoto (1980, 278–77) regards Śubhagupta’s and Dharmottara’s teacher-pupil (guruśiṣya) relationship as impossible due to Dharmottara’s use of the same proof as Kamalaśīla while defending the “invariable co-apprehension” argument against Śubhagupta’s refutation.

Writings Only five works, all of which are found in the Tibetan Canon, are explicitly ascribed to Śubhagupta. These are the (1) Verses on the Demonstration of the Omniscient One (*Sarvajñasiddhikārikā) (Tōhoku no. 4243), (2) Verses on the Demonstration of External Objects (*Bāhyārthasiddhikā­rikā) (Tōhoku no. 4244), (3) Verses on the Investigation of the Vedic Revelation (*Śrutiparīkṣākā­rikā) (Tōhoku no. 4245), (4) Verses on the Examination of the Exclusion from Others (*Anyāpohavicārakārikā) (Tōhoku no. 4246), and (5) Verses on the Refutation of God (*Īśvarabhaṅgakār­ikā) (Tōhoku no. 4247). They are preserved in Tibetan only, but many Sanskrit fragments have been found in the texts of other authors, particularly in Kamalaśīla’s Commentary. Śubhagupta’s extant works are all written in verse (likely providing the monks a mnemonic device for debates) and center on single themes. Those were arguably the most controversial of their time: Vedic revelation (śruti), the existence of an omniscient being, the reality of God, and the theory of exclusion (apoha) from others. Another work attributed to him, namely the Demonstration of Selflessness (*Nairātmyasiddhi) (Frauwallner 1957, 100), is now lost. Steinkellner (1985, 216–18) also regards the Demonstration of the Other World (*Paralokasiddhi) and its commentary, both lost, as his. Based on the existence of prose passages in Kamalaśīla’s Commentary that are introduced as direct quotations of Śubhagupta, Frauwallner (1933) had surmised the existence of another nonextant work of his, that is, an auto-commentary on the Demonstration of External Objects, which was never translated into Tibetan. I have argued against that assumption (Saccone 2014, 2018). The chronology of Śubhagupta’s works has not been established with certainty. Three works of his, the Examination of the Exclusion, the Investigation of the Vedic Revelation, and the Demonstration of the Omniscient One, appear to form a trilogy. Each of them begins with the word “therefore” (de’i phyir), referring to topics that had been treated in an earlier writing. Due to the character of the arguments they contain, as well as the comparison with analogous arguments in some chapters of the Compendium and the Commentary, the chronological succession of the three could be in the order listed previously.5 The Verses on the Refutation of God and the Demonstration of External Objects are Śubhagupta’s only writings that are not connected to his earlier works. Accordingly, one may conjecture that the Verses on the Refutation of God is the earliest (among those that are known to us). The Demonstration of External Objects, in turn, being his magnum opus and his most original writing, is arguably later, if not the latest.

The “Orthodox” Works As mentioned previously, the Refutation of God starts with a verse explicitly introducing it as a summary of arguments to use against opponents in an assembly in order to dispel ignorance. This work is constructed as a compendium of proofs against those who maintain that the universe is created by God. Śubhagupta’s opponents can be identified as representatives of the Brahmanical traditions of Nyāya and Vaiśeṣika, as well as of some Śaiva sources, while God is intended mainly as Śiva.6 Śubhagupta does not show any striking doctrinal originality in this work; he 430

Śubhagupta

primarily refers back to some argumentations and views already developed by other Buddhist philosophers in the past – such as those of Dharmakīrti (especially in the Commentary on the Means of Knowledge) and Vasubandhu (in the Commentary on the Treasury of the Abhidharma). At the same time, the specific use and articulation of the evidence can be viewed as his own original contribution. Many parallel formulations of these arguments are also found in Śāntarakṣita’s Compendium and Kamalaśīla’s Commentary. Here, Śubhagupta is mainly interested in rebutting those properties of permanence and unity that his opponents attribute to God – a God that creates a universe that is impermanent and manifold. The creation of the universe involves a succession of different phases, and a permanent and unitary being is, ex hypothesi, always identical to itself; accordingly, such succession would not be possible. Another undesirable consequence would be the impossibility of destroying ignorance, which, being created by an eternal cause, would be permanent. This would make any spiritual path towards liberation, whether Buddhist or nonBuddhist, aimless. The upholder of the existence of a creator God tries to avoid the contradiction of a permanent God creating an impermanent universe by affirming that God’s will is not permanent. However, Śubhagupta responds that a quality (i.e., will) of a substratum that is permanent must itself be permanent. Otherwise, the substratum itself cannot be admitted as such, since, having that quality only momentarily and not always, it would change based on its presence or absence. Issues arising with the concept of God’s will are also analyzed. If God is admitted by the opponents as permanent and always identical to itself, it is impossible to explain why his will to create arose at a certain point in time and was not always there since the beginning. Moreover, Śubhagupta argues, if God has a purpose, he is lacking something and accordingly is not perfect. If he does not have a purpose, on the other hand, he acts irrationally. According to the hypothetical sequence of works, the Refutation of God was followed by the “trilogy.” These three works center on ideas related to the philosophy of language and concept formation; they elaborate greatly on Dharmakīrti’s views as found in the Commentary on the Means of Knowledge. This is evident in the first text, the Examination of Exclusion, which discusses the theory of exclusion (apoha), the theory of linguistic and conceptual exclusion already found in Dignāga and elaborated upon in Dharmakīrti. Some things, though ultimately being different from each other, produce the same effects, such as similar cognitions; this is why they are all determined through the same concept (and denoted by the same word) – not because they have the same nature. The Investigation of the Vedic Revelation7 is intended as a refutation of the Vedic revelation (as synonymous with veda, see Eltschinger 1999, 48), in other words, a refutation of the authority of the Brahmanical scriptures. Exactly as in the Demonstration of the Omniscient One, Śubhagupta reuses (with slight modifications) some of the linguistic theories expounded in the chapter entitled Inference for Oneself (Svārthānumāna) in Dharmakīrti’s Commentary on the Means of Knowledge. This is done in order to refute some Brahmanical (namely Mīmāṃsā’s) linguistic conceptions that lie at the core of the justification of the Veda’s authority. The Mīmāṃsā school defends the idea that the Veda does not have any author (apauruṣeyatā) and its words, being eternal, are connected to their object by their own nature. Against this view, Śubhagupta argues that words are potentially fit to refer to any object. They refer to the speaker’s intended meanings, which are concepts and not external objects. The understanding of a thing by means of a word arises due to a convention, not because of any natural relation between that thing and the word. Thus, since words found in the Veda, like all words, are dependent on human-based conventions and intentionality, the Veda cannot be viewed as reliable by itself, that is, merely because it is not produced by human beings. Similarly, in the first part (vv. 1–12)8 of the Demonstration of the Omniscient One, which is devoted to the proof of the Buddha’s omniscience, Śubhagupta employs some of Dharmakīrti’s 431

Serena Saccone

linguistic theories. He aims to provide arguments against the same Brahmanical opponents (exponents of the Mīmāṃsā) and their theory of the absence of a human author of the Veda. Śubhagupta links the authoritativeness ( prāmāṇya) of the Veda to the omniscience of its author. He aims at demonstrating the necessary relationship between the truth of scriptures and the perception of extrasensory objects (atīndriyārthadarśana) by their author. Scriptures, being based on language, must necessarily have an author, since the meanings of words are just concepts intended by the mind of a speaker and are understood through human convention. Accordingly, their reliability cannot but be grounded in their being produced (kṛtakatva) by an author and in that author being a person ( puruṣatva). That person must also be admitted as being omniscient (sarvajña) in the sense of seeing truths that are beyond ordinary perception and are, ipso facto, precisely the object of scriptures. In part 2 (vv. 13–25),9 Śubhagupta introduces an actual proof of the Buddha’s omniscience. He is established as being the only omniscient person among many other omniscient beings who are admitted by other traditions as the authors of their scriptures. This is because: (i) he is the only one teaching about selflessness (and doing so efficaciously, that is, by adjusting his teachings to various people’s spiritual needs); and (ii) he knows and teaches true, that is, effective, ritual utterances (mantras). Śubhagupta employs the knowledge (and teaching) of the power and effects of mantras as evidence that the Buddha has direct perception of extrasensory things. Since the power of the mantras cannot be perceived or inferred, only someone who has that type of supersensible perception can know about them. Śubhagupta is the first to attest to the use of the teaching of mantras as proof for the Buddha’s omniscience. Before him, Dharmakīrti had already introduced that type of knowledge as a demonstration of extraordinary perception in human beings. However, Śubhagupta is the first (and one of the few) philosophers to revisit that proof in his writings and explicitly relate it to the Buddha’s omniscience. In the last verse, he states that, even when moral faults are purged, a zealous cultivation is not enough to attain the omniscience related to the knowledge of mantras. The source of that superior knowledge, Śubhagupta argues, is the Buddha, who teaches it and is established as possessing perception of extrasensory truths. In this sense, I believe that this text presents itself as the attempt of an author mostly concerned with logic and epistemology ( pramāṇa) to put the knowledge of mantras into relation (albeit indirectly) with the Buddhist soteriological path.

The “Heterodox” Work: The Verses on the Demonstration of External Objects The most original and innovative theories of Śubhagupta, however, are found in the Demonstration of External Objects.10 This work is intended to demonstrate the reality of external objects of cognitions against a form of “idealism,” that is, the Vijñānavāda (literally the doctrine of the [reality] of consciousness) tradition of Buddhism and, particularly, the doctrine of “consciousness-only,” that is, the true existence of consciousness alone (vv. 2–3). The trademark doctrines in this work are a form of externalism (bāhyārthavāda) as well as the notion that the images of sense objects are external and do not shape their cognitions (nirākāravāda). Śubhagupta’s externalism is construed as the conception that material things, consisting of atoms, are truly existent and are the support of sense cognitions. This is based on an atomic theory that draws on both the Buddhist Vaibhāṣikas’ and Sautrāntikas’ views on atoms. In the first part of the Demonstration, Śubhagupta is interested in proving the reality of infinitesimal material particles that are singular (*eka), that is, devoid of parts, and ultimately (“substantially,” *dravyataḥ) existent. These atomic particles are perceived by direct perception but

432

Śubhagupta

never singly. However, once they are apprehended, one superimposes some concepts, such as indigo and other conceptual images, on them. Accordingly, the images that manifest when we experience sense cognitions, such as a multicolored rug, and appear as if they were external are conceptual in nature. They are not verily apprehended by those sense cognitions, which are ultimately not shaped by the image of their objects (vv. 37–44). The proof of atoms’ singularity (45cd–56) is formulated as a refutation of Vasubandhu’s attacks on the atomic theory attributed to the Vaibhāṣikas of Kashmir by the author himself in his Twenty Verses (vv. 12–14 and commentary). Atoms are finally established to be unitary, that is, devoid of parts and identical to themselves. This is the conditio sine qua non for their aggregation, because the multiplicity of entities is admitted as tenable only if those entities are established as units. They are continuous and proximate but do not touch each other. Like the Vaibhāṣikas from Kashmir, Śubhagupta sees their continuity as their having nothing between them (such as light) and as being in opposition to each other. After demonstrating this, he turns to proving that atoms are real substances (dravya), in the sense that they are causally efficient. Atoms come to existence as aggregating since only with other atoms do they have the distinctive characteristic of supporting each other in performing one activity, that is, in being causally efficient and causing their own cognition. Atoms are “tied” to each other through this power of being a substance; accordingly, they cause their own cognition and are apprehended as such (vv. 57–58; see Saccone 2015). Śubhagupta’s position on how atoms aggregate was likely influenced by Dharmakīrti’s view in the Commentary on the Means of Knowledge and its developments in Devendrabuddhi’s and Śākyabuddhi’s commentaries.11 Śubhagupta’s theory that sense cognitions are not shaped by the image of their objects runs counter to both the Sautrāntikas and the Vijñānavādins but is particularly tailored to opposing the tenets connected with the “invariable co-apprehension” argument. This is an argument advanced by Dharmakīrti in the Ascertainment of the Means of Knowledge (particularly v. 1.54) but is already present in a different shape in the Commentary on the Means of Knowledge.12 It is meant to prove the nondifference between a cognition and its object since they are always necessarily perceived together. The latter, being proof of “consciousness-only,” is largely defended by the subsequent authors within the “mainstream” Buddhist logico-epistemological tradition. In contrast to this, Śubhagupta argues that, on the one hand, a cognition cognizes its object because it has the nature (svabhāva) of apprehending it, and, on the other, no sense cognition can occur without an external object. The nature (svabhāva) of a certain cognition is intended as the nature that belongs to that specific elementary particle (dharma) and nothing else. A cognition limits itself to illuminating the object without assuming its form. Since these two, object and cognition, arise in continuity, they are always found together, but this is not because they are identical, as the idealists would argue (vv. 66, 81, 82, 89). The original Sanskrit fragments of verses 66, 81, and 89 are quoted by Kamalaśīla in the part of his Investigation of External Objects chapter where he intends to defend the “invariable co-apprehension” argument. Kamalaśīla appears to view his proof of the latter argument especially as a response to Śubhagupta’s criticism. Thus, according to Śubhagupta, the cognitive process requires two elements in order to occur: a cognition, which is the only apprehender, and an object (*viṣaya), which possesses a form to be apprehended. Cognitions are compared to pure “light” since they have the nature of making their objects known. For this reason, a cognition and its object depend on each other and are part of the same causal complex, namely that particular perceptual process. Śubhagupta concludes that this is why a cognition and its object are necessarily perceived together; the invariability (niyama) is due to their causal relationship, not to their nondifference. In the Demonstration of External Objects, we read:

433

Serena Saccone

“ There is no other apprehender besides cognition, [and] there is no sensory cognition without objects. And therefore, the [fact of] being aware [of them] together is not because of the nondifference of[, for example,] an indigo [thing] and its cognition.” (v. 66) But it is precisely the preceding [causal] complex that can make the instant of the object cognized, in the same way that a visual form [is lit up] by light. It is by this that there can be [their] being brought to awareness together. (v. 81) Since cognitions and [their] objects always arise in continuity, then, the term [“co-perception”] is [used], because it is like that; however, in reality there is no co-perception [of them]. (v. 82) In Kamalaśīla’s corpus, particularly in the Commentary on the Compendium of Truths and in the Commentary on the Ornament of the Madhyamaka, one finds many accounts of Śubhagupta’s viewpoints, which Kamalaśīla presents in prose passages as quotations of his. These accounts are historically relevant insofar as they testify to the reception of Śubhagupta’s views in the tradition; this is useful in order to assess the intellectual distinctiveness of Śubhagupta himself. A rather exhaustive account of Śubhagupta’s position on the nature of cognitions and the  cognitive process is provided in Kamalaśīla’s Commentary on the Ornament of the ­Madhyamaka, and is “copied” verbatim in the “Chapter on Suchness” from Haribhadra’s Abhisamayālaṃkārālokā.13 The text reads, Śubhagupta says: “And every cognition of an ordinary person occurs as devoid of that impression that is the image of an indigo [thing]. Regarding this [cognition,] the undesirable consequence of the loss of [its] singularity does not [arise], since it does not [truly] have a variegated form. The fact that this [cognition] is determined as having the nature of experiencing an indigo [thing] and so on, is because it has the nature of experiencing an indigo [thing] and so on; however, [it is] not because it is endowed with the form of an indigo [thing] and so on. To explain: an image (*ākāra) is said to be only [the] aspect (*prakāra) of apprehending the object-support (*ālambana), but not [to be] the form of that object-support [Abhisamayālaṃkārālokā: but not the similarity (of the cognition with the object)]. [That] indigo [thing] and so on, which is cognized as appearing as if it were external, is not an image of cognition. However, the cognizer, experiencing a cognition [that is] the awareness of an indigo [thing] and so on, conceptually determines the indigo [thing] and so on with such an image [Abhisamayālaṃkārālokā: due to ignorance, (the cognizer) conceptually determines an indigo (thing) and so on, like this, with an external image].”14 Here, Śubhagupta is reported as defining an image (ākāra) in cognition as the cognition’s own aspect ( prakāra) of apprehending its object-support (ālambanagrahaṇaprakāra), in contrast with an interpretation that takes the cognitive image to be the form of that object shaping a cognition. This appears to be an echo of Vasubandhu’s Commentary on the Treasure of the Abhidharma (Abhidharmakośabhāṣya) (Vasubandhu 1975, 401 [ad 7.13b]). In his sub-commentary, Yaśomitra attributes this position to the Sautrāntikas (Yaśomitra 1936, 629). However, Śubhagupta’s position differs significantly from that of the Sautrāntikas. Unlike him, 434

Śubhagupta

they think that, for a cognition, apprehending an object means precisely its having the image of that object. According to Śubhagupta, instead, the cognition has the characteristic, that is, the nature, of cognizing an object, and, in this sense, prakāra should be intended as its “manner,” “mode of being.” Therefore, the image (ākāra) of a cognition, for Śubhagupta, is not an image in the cognition but rather the characteristic, that is, the mode of being itself, of the cognition as apprehending the object. This could be compared to the light having the characteristic of illuminating objects without taking their images or being shaped by them. In the Demonstration of External Objects, at least two ways of construing an aspect-image (ākāra) can be found: aspect-image in the sense of prakāra, that is, the aspect or mode of the cognition that apprehends a specific object, and aspect-image in the sense of the physical form that belongs to the external object only and does not shape the cognition internally. Śubhagupta expresses this in v. 84: A cognition is said [to be] brought to awareness since it is endowed with the nature of being an awareness. The object is brought to awareness because it generates a cognition having [its] appearance [as an object]. Against the “invariable co-apprehension” argument, Śubhagupta argues that there is not only one aspect-image when someone cognizes an object, but two. Those two aspect-images (one belonging to cognition, the other to the object) are both experienced, separately, one after the other. Verses 77, 78, and 79 explicitly counter the idea of the existence of a single image. If there were only one image, it would have to be either that of the object only or that of the mere cognition only. However, Śubhagupta argues that neither possibility is logical. Instead, in perception, we are aware of both the aspect-image ( prakāra) of the cognition as cognizing the object and of the aspect-image (ākāra) of the object. This is still purely at a perceptual level. Since these two are experienced one after the other, they are wrongly determined as one aspect-image, although they are not (v. 82). Moreover, for Śubhagupta, such apprehending (grahaṇa) of the object-support (ālambana) is certainly not an activity. In v. 92, he implicitly reaffirms the claim that when the sūtras say “cognition cognizes,”15 and so on, they do not mean that cognition is an agent or does something. Cognition is an apprehender because that is its nature but not by virtue of any activity. A cognition, like every elementary particle (dharma), is in fact devoid of activity and only conventionally talked about as having one. A very similar concept was advanced in Vasubandhu’s Commentary on the Treasury of the Abhidharma and reported as a Sautrāntika idea. According to the latter view and Śubhagupta’s, a visual cognition arises depending on sight and visual forms. However, this visual cognition is devoid of activity. It consists only of elementary particles and can be explained through the relation of cause and effect. Metaphors like “eye sees, cognition cognizes” are thus not literally true but are used intentionally for the sake of everyday activity (vyavahāra). Nevertheless, one should not get attached to them (Vasubandhu 1975, 31 [ad 1.42]). Thus, once again, Śubhagupta elaborates on a Sautrāntika principle and reworks it to fit his personal standpoint. His take on the cognitive process is evidently an instance of his syncretic reuse of different standpoints from diverse traditions. A cognition is devoid of activity but has the nature of being the cognition of a specific object-support (ālambana). However, it does not assume the image of that object-support, which comprises the atoms serving as the actual cause of their own perception and, as such, being grasped by sense cognitions. The image (ākāra), in the sense of the physical form of those atoms, remains external. In the moment immediately following the non-conceptual direct perception, conceptual construction “interprets” the atoms through a conceptual image, such as that of an extended object, which is superimposed on them (v. 36). This conceptual image appears as if it were external, does 435

Serena Saccone

not manifest in sense cognitions, and is manifold. However, sense cognitions are not manifold, since they do not ultimately possess those images. In sum, with the Demonstration of External Objects, Śubhagupta breaks with the tradition by re-elaborating upon it and proposing original ideas deviating from its baseline doctrine. In this sense, while he employs some ideas of Vasubandhu, Dignāga, and Dhamakīrti, he refutes their “idealistic” standpoints.16

Conclusions Though largely overlooked by contemporary scholarship, Śubhagupta represents a pivotal and innovative figure within the Buddhist tradition of logic and epistemology. His influence and importance in the eighth century are manifested by Śāntarakṣita and Kamalaśīla, two prominent figures of the tradition, who attribute great significance to the refutation of his externalism and his theory of the absence of images in cognition. They do so in the chapters of their Compendium of Truths and Commentary that are devoted to the demonstration and defense of “consciousness-only.” In spite of the presence of important and valuable non-Buddhist opponents, such as Kumārila Bhaṭṭa, they prefer to address their criticism to a contemporary, and fellow Buddhist, scholar-monk. This is particularly true with regard to Kamalaśīla. Three, perhaps four, of Śubhagupta’s works were conceived as one set of texts. All of them are deeply indebted to the Dharmakīrtian tradition and drew on the works of previous Buddhist authors. Assuming that all his works were composed prior to Śāntarakṣita and Kamalaśīla’s Compendium of Truths and Commentary, Śubhagupta was most likely the first Buddhist author within the logico-epistemological tradition to adopt the genre of digests – a genre that would prove popular among later authors. Moreover, the specific statements of argumentation and elaboration of doctrines (predominantly taken from previous Buddhist authors) that are found in his digests demonstrate his unique innovations and contributions to the field. Nevertheless, the most groundbreaking element of his thought remains his attempt to integrate views from the tradition in a syncretic fashion. This was done in order to reaffirm a more “externalist” account of reality in contrast to the mainstream Buddhism idealism. From a careful analysis of Śubhagupta’s corpus in its entirety, we thus uncover the figure of an innovative thinker who played a crucial role in the history of the Buddhist tradition of logic and epistemology both by systematizing it and by subverting it.

Notes 1 Śubhagupta’s doctrinal affiliation has been defined in different ways in ancient sources and among contemporary scholars. In particular, it is debated whether he was a Vaibhāṣika or a Sautrāntika. On this, see Saccone 2018, 2019a. 2 On this verse, see Saccone Forthcoming; Eltschinger and Ratié Forthcoming. 3 On this verse, see Steinkellner 2006, 201, 204. 4 See Bu ston 1988, 162. Bu ston is almost certainly referring to the previously mentioned colophon of the Commentary on the Ascertainment of the Means of Knowledge. 5 On this, see Saccone 2019b, 457. 6 On this, see Eltschinger and Ratié Forthcoming. 7 On this work, see Eltschinger 1999. 8 On the first part of this work, see Saccone 2019b. 9 On the second part of this work, see Saccone Forthcoming. 10 For an edition, English translation, and analysis of parts of this work, see Saccone 2018. For its first English translation, see Shastri 1967. 11 See Pramāṇavārttika, chapter on perception, 195–96; Saccone 2015, 126–28.

436

Śubhagupta 1 2 See Pramāṇavārttika, chapter on perception, 387–390abc. 13 This is introduced by Kamalaśīla as a prose quotation of Śubhagupta and by Haribhadra as the objection of an unidentified opponent. The passage is not found in any extant work that is explicitly ascribed to Śubhagupta and neither explicitly refers to nor is connected with any verses of the Demonstration of External Objects, which is the only work devoted to epistemological issues with regard to sense cognitions. This is not the only instance where Haribhadra appears to be reusing materials from the Commentary on the Ornament of the Madhyamaka. Accordingly, one is led to believe that Haribhadra is copying Kamalaśīla here, not quoting a lost work of Śubhagupta. 14 See Kamalaśīla 1985, 163; Haribhadra 1932–1935, 632–33. 15 See yat tarhi “vijñānaṃ vijānāti” iti sūtra uktaṃ kiṃ tatra vijñānaṃ karoti (Vasubandhu 1975, 473– 74 [Pudgalavādapratiṣedha chapter]) 2). 16 The targets of his refutation are mostly the Twenty Verses and its commentary by Vasubandhu and the Investigation of the Object-Support and its commentary by Dignāga, as well as the “invariable co-apprehension” argument expressly formulated by Dharmakīrti, especially in the Ascertainment of the Means of Knowledge.

Bibliography Arcaṭa. 1949. Hetubinduṭīkā of Bhaṭṭa Arcaṭa with the Sub-Commentary Entitled Āloka of Durvekamiśra, edited by Sukhlal Sanghavi. Baroda: Baroda Oriental Institute. Bu ston. 1988. Bu ston chos ’byung, edited by Rdo rje rgyal po. Beijing: Krung go’i bod kyi shes rig dpe skrun khang. Dharmakīrti. 1979 [1985]. Pramāṇavārttika Pratyakṣa. Bukkyōninshikiron no kenkyū. edited by Hiromasa Tosaki, 2 vols. Tokyo: Daitōshuppansha. Eltschinger, Vincent. 1999. “Śubhagupta’s Śrutiparīkṣākārikā (vv. 10cd  – 19) and Its Dharmakīrtian Background.” In Dharmakīrti’s Thought and Its Impact on Indian and Tibetan Philosophy. Proceedings of the Third International Dharmakīrti Conference (Hiroshima, November 4–6, 1997), edited by Shōryū Katsura, 47–61. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Eltschinger, Vincent, and Isabelle Ratié. Forthcoming. “A  Buddhist Refutation of the Existence of a Creator God: Śubhagupta’s Īśvarabhaṅgakārikā.” In Gedenkschrift Helmut Krasser, edited by Vincent Eltschinger, Parimal Patil, Leonard Van der Kuijp, and Chizuko Yoshimizu. Hamburg Buddhist Studies. Hamburg: Hamburg University Press. Frauwallner, Erich. 1933. “Dignāga und anderes.” In Festschrift für Moriz Winternitz, edited by Otto Stein and Wilhelm Gambert, 237–42. Leipzig: Otto Harrassowitz. ———. 1957. “Zu den buddhistischen Texten in der Zeit Khri-sroṅ-lde-bstan’s.” Wiener Zeitschrift für die Kunde Süd- und Ostasiens 1: 95–103. ———. 1961. “Landmarks in the History of Indian Logic.” Wiener Zeitschrift für die Kunde Süd- und Ostasiens 5: 125–48. Funayama, Toru. 1995. “Arcaṭa, Śāntarakṣita, Jinendrabuddhi, and Kamalaśīla on the Aim of a Treatise (Prayojana).” Wiener Zeitschrift für die Kunde Südasiens 39: 181–201. Haribhadra. 1932–1935. Abhisamayālaṃkār’ālokā Prajñāpāramitāvyākhyā (Commentary on the Aṣṭasāhasrikā Prajñāpāramitā) by Haribhadra. Edited by Unrai Wogihara. Tōyō Bunko Publications Series D, vol. 2. Tokyo: The Tōyō Bunko. Hermann-Pfandt, Adelheid. 2008. Die lHan kar ma – Ein früher Katalog der ins Tibetische übersetzten buddhistischen Texte. Kritische Neuausgabe mit Einleitung und Materialien. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Kamalaśīla. 1985. Madhyamakālaṃkāra of Śāntarakṣita with His Own Commentary or Vṛtti and with the Subcommentary or Pañjikā of Kamalaśīla. Edited by Masamichi Ichigō. Kyoto: Buneidō. Matsumoto, Shiro. 1980. “Sahopalambhaniyama.” Sōtōshū kenkyūin kenkyūsei kenkyū kiyō 12: 298–65. Saccone, Serena. 2014. “Śubhagupta on the Cognitive Process.” Journal of Indian Philosophy 42 (2/3) . In Birgit Kellner, Sara McClintock (eds.): ākāra in Buddhist philosophical and soteriological analysis, pp. 377–399. ———. 2015. “The Conception of Atoms as Substantially Existing in Śubhagupta.” Journal of the International Association of Buddhist Studies 38: 107–37. ———. 2018. On the Nature of Things. A Buddhist Debate on Cognitions and Their Object. Wien: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. ———. 2019a. “Śubhagupta.” In Brill’s Encyclopedia of Buddhism, vol. II, edited by Jonathan A. Silk, R. Bowring, V. Eltschinger, and M. Radich. Leiden and Boston: Brill: 458–462.

437

Serena Saccone ———. 2019b. “Of Authoritativeness and Perception: The Establishment of an Omniscient Person (Against the Mīmāṃsakas).” In Wind Horses. Tibetan, Himalayan and Mongolian Studies, edited by Andrea Drocco, Lucia Galli, Chiara Letizia, Giacomella Orofino, Carmen Simioli. Napoli: Università degli Studi di Napoli “L’Orientale”, Dipartimento Asia, Africa e Mediterraneo, Series Minor LXXXVIII: 455–484. ———. Forthcoming. “Apology for Omniscience. An Eighth-Century Demonstration of the Buddha’s Sarvajñatva.” In: Gedenkschrift Helmut Krasser, edited by Vincent Eltschinger, Parimal Patil, Leonard Van der Kuijp, and Chizuko Yoshimizu. Hamburg Buddhist Studies. Hamburg: Hamburg University Press. Sanderson, Alexis. 2009. “The Śaiva Age – The Rise and Dominance of Śaivism During the Early Medieval Period.” In Genesis and Development of Tantrism, edited by Shingo Einoo, Tokyo: Institute of Oriental Culture, University of Tokyo: 41–350. Schiefner, Anton. 1868. Rgya gar chos ’byung (Tāranātha): Tāranāthae de Doctrinae Buddhicae in India Propagatione Narratio. Contextum tibeticum e codicibus petropolitanis edidit Antonius Schiefner. St. Petersburg: Academia Scientiarum Petropolitana. Shastri, N. A. 1967. “Bāhyārthasiddhikārikā.” Bulletin of Tibetology 4 (2): 1–96. Steinkellner, Ernst. 1985. “Paralokasiddhi-Texts.” In Buddhism and Its Relation to Other Religions: Essays in Honour of Dr. Shozen Kumoi on His Seventieth Birthday. Kyoto: Heirakuji Shoten: 215–224. ———. 2006. “Miszellen zur erkenntnistheoretisch-logischen Schule des Buddhismus IX: The Colophon of Dharmottara’s Pramāṇaviniścayaṭīkā.” Wiener Zeitschrift für die Kunde Südasiens 50: 199–205. Ui, H., M. Suzuki, Y. Kanakura, and T. Tada, eds. 1934. Chibetto-Daizō-kyō Sōmokuroku (A Complete Catalogue of the Tibetan Buddhist Canons). Sendai: Tōhoku Imperial University. Vasubandhu. 1975. Abhidharmakośabhāṣyam of Vasubandhu. Edited by P. Pradhan. Patna: K. P. Jayaswal Research Institute (first ed. 1967). Yaśomitra. 1936. Sphuṭārthā Abhidharmakośavyākhyā by Yaśomitra. Edited by Unrai Wogihara (1932– 1936), 2 vols. Tokyo: The Publishing Association of Abhidharmakośavyākhyā.

438

26 DHARMOTTARA Systematic and Innovative Commentator David Nowakowski

Dharmottara, who flourished circa 740–800 CE, is best known as a commentator on Dharmakīrti’s (c. 600–660 CE) epistemological works, though several of his independent (non-commentarial) treatises also survive in Tibetan translation. While biographical information is scant, we have evidence that Dharmottara moved to Kashmir during the reign of King Jayāpīḍa, a great patron of scholars and builder of three large images of the Buddha, who reigned in the latter decades of the eighth century. Roughly a century later, Dharmottara would gain the distinction of being one of the only Buddhist philosophers to appear as a character in classical Sanskrit drama: a Buddhist monk named Dharmottara appears in the first act of a play by the Kashmiri Nyāya philosopher Jayanta Bhāṭṭa, though as we might expect from a self-styled defender of Vedic traditions, Jayanta’s portrayal of Dharmottara is quite far from flattering.1 The present chapter will focus especially on Dharmottara’s commentary on Dharmakīrti’s Epitome of Reason (Nyāyabindu),2 examining the ways in which Dharmottara works out the consequences of key tenets of Dharmakīrti’s philosophy in an integrated and systematic way. In general, Dharmottara’s work as a systematist pushes him to comment in ways that go well beyond the surface meaning of individual passages in Dharmakīrti’s text taken on their own, prompted not by misunderstanding or a desire for novelty but by the need to resolve apparent contradictions or tensions between various aspects of the system. Specifically, Dharmottara’s exposition of Buddhist epistemology and philosophy of mind,3 as seen in his commentary on the Epitome of Reason, seems to blur Dharmakīrti’s bright lines between perception, the internal process of inferential reasoning known as “inference-for-oneself” (svārthānumāna), and the interpersonal dialogue and dialectic known as “inference-for-others” ( parārthānumāna). And in doing so, Dharmottara tends to make all of these much more dependent upon the state and circumstances of particular individual cognizers, as contrasted with a more objective, impersonal account of the cognitive-epistemic processes. Of course, as a good classical Indian commentator, Dharmottara presents all of this as if it were already present – if not on the page of Dharmakīrti’s texts, then at least in his illustrious predecessor’s mind as the author of those texts. Yet at a minimum, we can confidently say, consistently with Dharmottara’s own self-presentation, that Dharmottara’s work constitutes a marked shift in emphasis, such that cognizer-relativity takes center stage in a way that it does not in Dharmakīrti’s root texts. DOI: 10.4324/9781351030908-39

439

David Nowakowski

In what is perhaps that same spirit, the present chapter itself is not particularly innovative. It simply draws together three threads which have emerged separately in the scholarly literature on Dharmottara in order to suggest the interconnections of these three themes and from there to draw some tentative lessons about thinking of Dharmottara as a systematist, as an innovator on Dharmakīrti, and as a stepping-stone to later Buddhist epistemologists such as Jñānaśrīmitra and Ratnakīrti, who will both adopt Dharmottara’s innovations and further innovate in the same areas. The first section will examine what is perhaps Dharmottara’s most significant contribution to the subsequent tradition: the additional complications he brings to Dharmakīrti’s account of the differences between perception and inference, making these two cognitive processes appear much more similar in their structure and operation and (especially for perception) much more dependent on contributions made by the perceiver herself. The second section will consider Dharmottara’s account of “the non-apprehension of the perceptible” in the Buddhist theory of inference, examining how perceptibility is understood as being relative to some particular cognizer’s circumstances. The third section will address Dharmottara’s presentation of the formal conditions which define genuine inferential reasons, where Dharmottara uses resources from his predecessor’s account of debate and dialectic to elucidate an account of inferential reasoning which likewise depends upon the epistemic state of each individual cognizer, where the cognizer’s doubt or certainty about the relations between various properties plays a key role in making it the case that one property constitutes a reason for inferring another. Larger systematic considerations drawn from Dharmakīrti’s philosophy will play a role in motivating Dharmottara’s arguments in each section: universal momentariness, basic causal theory, and the basic account of doubt, in the first through third sections, respectively.

The Processes and Objects of Cognition The issue on which Dharmottara most clearly, and perhaps also most radically, innovates on Dharmakīrti’s work is in defining and distinguishing perception and inference, both in terms of the processes by which these two means of knowledge operate and in terms of the cognitive objects which each of them make known. In the wider Indian philosophical tradition of which the work of Dharmakīrti and Dharmottara is a part, the two questions which organize and structure debates in epistemology are: (1) what are the instruments, or causal processes, by which that subset of awareness-events which count as knowledge are generated? and (2) what sorts of objects can be known by means of each of those instruments or processes? In answer to the first question, the standard claim found throughout the work of Dignāga and Dharmakīrti, and which all of their Buddhist successors feel bound to uphold in some way, is that there are exactly two such instruments of knowledge: perception and inference. Other Indian philosophical schools such as the Nyāya or Mīmāṃsa would countenance a longer list of instruments, such as testimony, analogical reasoning, and the awareness of absence. But for the Buddhist epistemologists, all of these latter candidates will either be subsumed as special cases of perception or inference or else written off as misleading (and thus not invariably reliable means of knowledge at all). The answer to the second question gets more complicated, and these complications point directly to Dharmottara’s elaboration or revision of Dharmakīrti’s account. In Dharmakīrti’s account of cognition, perception and inference each have a single object (i.e., what it is that the cognition is a cognition of), and that single object is of a different kind in perception and in inference: a unique particular is the object of a perceptual awareness, while a universal is the object of an inferential awareness. Thus, perception and inference each range over a distinct 440

Dharmottara

domain of objects – a claim which serves to distinguish Dharmakīrti’s account of knowledge from that of rival, non-Buddhist schools, who argue that all of the instruments of knowledge range over a single, unified domain of objects.4 While Dharmottara will strive to maintain some clear difference between perception and inference, he will argue that, in contrast to Dharmakīrti’s model, each type of veridical awareness has two objects: one which is grasped (grahya) and another which is determined (adhyavaseya). The implications of this shift get at some of the central commitments of the Buddhist epistemological tradition. This section will briefly present the core of Dharmakīrti’s model before examining the motivations for Dharmottara’s revisions and the implications of those changes for Indian Buddhist theories of mind and cognition more generally. Throughout his corpus, Dharmakīrti consistently follows Dignāga in distinguishing two kinds of objects of knowledge: particulars (svalakṣaṇa), which exist ultimately ( paramārthasat), by their nature are utterly distinct from every other particular, and are necessarily momentary (i.e., existing for only a single, smallest instant of time); and universals (sāmānya), which exist only conventionally, are characterized by a common nature shared over multiple instances, and are for those reasons non-momentary (i.e., persisting through time). While this distinction is at work throughout Dharmakīrti’s corpus, an especially clear presentation and defense of the view occurs at the beginning of the chapter on perception in his Pramāṇavārttika (PV 3, trans. Franco and Notake 2014). The chapter begins with the assertion that “the instruments of knowledge are of two kinds, because the objects of knowledge are of two kinds” (PV 3.1a–b).5 The direction of this explanatory relation – the duality of objects giving rise to, and justifying, Dharmakīrti’s postulate of the duality of instruments – is repeated in PV 3.63, which concludes the relevant section of that text: “Therefore, the duality of the instruments of knowledge is accepted because of the duality of the objects of knowledge.” Dharmakīrti follows this arrangement in setting out the argument: first presenting four arguments for the distinction between particulars and universals (PV 3.3–50) and then arguing on the basis of that distinction for a corresponding distinction between perception and inference (3.55–62). The main argument (3.3–6) is that particulars are unlike universals because only the former possess the capacity to bring about effective action (arthakriyā). That is, because only particulars ultimately exist, it is only by acting toward particulars that we can achieve real results in the world, and so it is only when we act toward a particular that the desire which motivated our action will be satisfied. Since universals do not ultimately exist, acting only toward a universal (without a real particular somehow being implicated) will necessarily leave us unsatisfied. Thus, both the distinction between particulars and universals and the corresponding distinction between perception and inference on that basis are central tenets of Dharmakīrti’s epistemology and philosophy of mind. Dharmottara will radically revise his predecessor’s account of both. Commenting on the perception chapter of Dharmakīrti’s Epitome of Reason, Dharmottara explicitly asserts that perception has not one but two objects: the grasped object, defined as “that whose image is produced (in awareness),” and the determined object, defined as that which can be attained. Dharmottara explains: The grasped object is one thing and the determined is something else, since for perception, what is grasped is a single moment, but what is determined  – through a judgment that arises by the force of perception – can only be a continuum. And only a continuum can be the attainable object of perception, because a moment cannot be attained. (NBṬ on NB 1.12, trans. McCrea and Patil 2006, 325) 441

David Nowakowski

As Dharmottara’s argument clearly shows, his revision to the basic account of perceptual awareness  – itself a fundamental doctrine in the Buddhist epistemological tradition  – is motivated by systematic concerns arising from the closely related and equally fundamental doctrine of momentariness: the thesis that whatever ultimately exists exists only for a single smallest instant of time (i.e., a single “moment,” kṣaṇa), after which it immediately ceases to exist. There are two interconnected parts to Dharmottara’s reasoning. The first part follows directly from the theory of momentariness. Given that any particular exists for only a single instant of time, by the time we act in response to our perception of it, it is no longer there to be attained. If, as the Buddhist epistemologists assert, every purposeful action is motivated by desire, and enabling us to attain the objects of our desire is what distinguishes veridical from non-veridical awarenesses, then the particular entity which we grasp in perception cannot possibly be what we desire, since due to its momentariness, it is inevitably unattainable. In other words, because of momentariness, we can never obtain any particular which we perceive. In some way, then, something which is not a particular must be involved both as a cognitive object of awareness and as an intentional object of action. The second part is based on the distinction between conceptual and non-conceptual awarenesses. As Dreyfus notes (1997, 355), effective action (arthakriyā) requires cognition which is veridical, or non-deceptive. Such a cognition, in turn, necessarily involves accurately identifying the object of thought and action. And this need for accurate identification and categorization straightforwardly means that conceptual awareness must be involved. A cognition which is purely non-conceptual will not suffice. Yet, as Dreyfus then remarks, “such an epistemic coordination between perception and conception, however, is in principle outside Dharmakīrti’s system” (1997, 358). Thankfully for Dharmottara, these two lines of reasoning converge on a single solution, since in the standard Dharmakīrtian system, non-conceptual awarenesses deal always and only with unique, momentary particulars, while conceptual awarenesses deal always and only with repeatable, non-momentary universals. Where Dharmakīrti limited his account of perception to involving only the former, Dharmottara need only find a way to include, within his revised account of perception itself, some role for the latter. Dharmottara thus posits a second object of perception, in addition to the grasped object which is “given” from the world: a continuum (santāna), which the cognizer determines. The continuum is an ultimately unreal entity, composed of the various real momentary entities which exist in a causal series. And this continuum is not something simply “found” in the external world, nor simply “given,” but rather is constructed by the cognizer herself based on her own interests, desires, habits, and latent mental traces. Thus Dharmottara’s act of determination, as McCrea and Patil note, “makes available to us constructed objects which are not directly presented to our awareness” (2006, 331). The very suggestion that the perceptual process itself involves such conceptually constructed, not directly presented objects is a break with Dharmakīrti, the radical nature of which cannot be understated. To give a concrete example, according to the theory of momentariness, what we conventionally but inaccurately take to be a single pot persisting through time is really a series of distinct pot-moments, each of which is among the causes of the subsequent pot-moment. These pot-moments are real particulars, while the continuum comprised of all the sequential pot-moments is an unreal, conceptually constructed universal.6 While it lacks ultimate reality, the pot-continuum does have the virtue of persisting through time. And it is that temporal persistence which enables us to perceive, form the desire for, act toward, and finally obtain it. Without the persisting, determined object, we would never be able to attain what we perceive, 442

Dharmottara

and so we would have no basis for distinguishing veridical from non-veridical cognitions; rather, all cognitions would be non-veridical. Immediately following the passage quoted previously, Dharmottara goes on to explain that inference, too, has both a grasped and a determined object but in a way that is reversed from the perceptual case. In inference, it is “an imposed thing,” “a non-entity” – that is, a universal – which is grasped, while it is a particular that is determined. In both perception and inference, the grasped object is given or presented to us, and the determined object is what we act upon or toward. In order to ward off the obvious questions, Dharmottara accounts for the apparent divergence between himself and his teacher by explaining that in the passage of the Epitome of Reason upon which he is commenting (and, by implication, at every other point in Dharmakīrti’s corpus where the subject arises), Dharmakīrti is concerned merely with explaining the grasped objects of each mode of awareness. Yet this should not suggest that the term “determination” (adhyavasāya) is absent from Dharmakīrti’s corpus. Quite the contrary. This is a term employed by Dharmakīrti but radically re-purposed by Dharmottara. Dharmakīrti uses “determination” to refer to a proper subset of conceptual (i.e., verbal or inferential) awareness; the term plays no role in his account of perception as such. For Dharmottara, as we have seen, “determination” shifts to become one component part in every instance of both perceptual and inferential awareness. Thus, we find two important shifts from Dharmakīrti’s presentation of the basic model of awareness to that of Dharmottara. First, as McCrea and Patil (2006, 332) observe, while Dharmakīrti could distinguish perception from inference not merely by their objects but also by the quite distinct processes involved in each, for Dharmottara, the parallel processes of grasping and determination in both cases mean that perception and inference can be distinguished only by their objects. The sharp opposition between the two epistemic instruments in Dharmakīrti is thus significantly worn down. Second, we see Dharmottara bringing conceptual awareness into the process of perception itself, which Dharmakīrti has presented as purely non-conceptual. The process of concept-formation is highly dependent upon each individual cognizer. Here we see an important case of Dharmottara as innovator and as systematic thinker, for it is on the basis of reconciling momentariness with cognitive theory that he makes these major revisions to Dharmakīrti’s basic account of perception and inference. And while it is spelled out less explicitly in the present passages, we can also see Dharmottara shifting toward a more cognizer-relative account of perception and inference thanks to the constructed nature of the determined object. As the first to articulate such a dual-object model, Dharmottara leaves some challenges to be worked out, a task which is taken up enthusiastically by Jñānaśrīmitra (fl. 975) and his student Ratnakīrti (fl. 1000), who adopt Dharmottara’s dual-object account while continuing to revise the details of exactly how precisely to characterize those two objects.7

Perceptibility and Cognizer-Relativity Dharmottara also functions as an (apparently) innovative commentator, expanding on small hints latent in Dharmakīrti to move Buddhist epistemology in a cognizer-relative direction, in his discussion of the theory of inference. Dharmakīrti famously sets out a classification of inferential reasons (hetu) into exactly three legitimate varieties, or styles of cases, which invariably produce knowledge of what is to be inferred. Of these, the most complicated type, normally discussed last, involves the non-apprehension of what would be, if it were present, perceptible (dṛśyānupalabdhi), such as when someone sees a floor in all its full and vibrant detail but does not see a pot which 443

David Nowakowski

would be perceptible were it to be present and infers the absence of a pot on that floor. Here, the problem of interest to Dharmottara is this: what does it mean for the pot (or whatever other object) to be “perceptible” (dṛśya), especially given that it is not present and perhaps even non-existent? And, correspondingly, what would it mean for something to be imperceptible (adṛśya) in the relevant sense? Just as with “determination” in the preceding section, here again, we find Dharmottara deploying a conceptual (and terminological) resource that is present in a very limited way in Dharmakīrti but is used by Dharmottara much more extensively. This is the notion of being “remote in space, time, or kind” (deśakālasvabhāvaviprakṛṣṭa). While Dharmakīrti mentions remoteness a number of times, the full term with its triple classification appears only once in the Epitome of Reason (NB 2.27), where it is given without elaboration, and only rarely elsewhere. The term rises to greater prominence in Dharmottara’s commentary and takes on a great significance in other contexts for subsequent members of the Buddhist epistemological tradition, including Ratnakīrti, for whom a rigorous account of remoteness in space, time, and kind plays an important role in the refutation of the existence of Īśvara, the omniscient demiurge who figures prominently in later Nyāya philosophy, and of the existence of other minds.8 As we will elaborate on in the following, the term is intended to capture all of the relevant ways in which an object could be too far away to be perceived by some particular cognizer: in spatial distance, temporal distance, or metaphysical distance (for instance, certain demons can be perceived only by their fellow demons or advanced yogins but not by ordinary humans). The Buddhist epistemologists’ basic principle, already found in Dharmakīrti, is that there are no unactualized causal capacities. Any entity is only appropriately called “cause of (some effect) X” if at that very moment it is causing X. In the case of perceiving some object, such as a pot, what are the causes of that perceptual awareness? And, correspondingly, what are the causes of the pot’s (or other object’s) being perceptible? Such causes will include the pot itself, along with the cognizer’s sense-organ(s), sufficient light (or similar conditions, for sense-modalities other than sight), and a variety of other factors. For our purposes, following Dharmottara’s discussion in commenting on NB 2.14 (translated and discussed by Kellner 1999, 199–200), we need only note the presence of an extensive list of causes, where the object itself – the distinct entity which is perceived – can be considered the primary cause, and all the others can be grouped together as the complete set of additional causal factors. Now, in the case of two ordinary material objects which actually are situated side by side or one atop the other like the pot and the floor, the complete group of additional factors which, together with the floor itself, produce a perceptual awareness of the floor will be exactly the same as the complete group of causal factors which, together with the pot itself, produce a perceptual awareness of the pot. Given the principle that there are no unactualized causal capacities, it follows that in this case, it is impossible to perceive the floor without simultaneously perceiving the pot or to perceive the pot without the floor. For if each distinct object (pot and floor) is itself present, the only way for that object not to be perceived is if one (or more) of the additional factors were lacking. But since the additional factors are the same for both, then the absence of any factor required for pot-perception would also preclude floor-perception, and vice versa. By contrast, in the case where the floor is present but the pot is not, a cognizer who is aware of perceiving the floor but not the pot, and who knows that the additional factors for potperception and floor-perception are the same, can readily conclude that since all the additional factors are present (evidenced by their bringing about a perception of the floor) the lack of pot-perception must be due to the absence of the pot itself. Reflecting appropriately on all of this, the cognizer infers the absence of a pot here on the floor. 444

Dharmottara

In this latter example, it is the presence of all the other additional causal factors which in some sense justifies calling the pot “perceptible” (dṛśya). And in his discussion of this issue, Dharmottara explicitly states that the account of perceptibility is “given with regard to one particular cogniser.”9 While the foregoing discussion makes clear the basic insight which the Buddhist epistemologists are attempting to capture in the non-apprehension inference, the complicated subjunctive phrasing – what properties a thing would have if it were to be present in a certain way – goes both far beyond a literal translation of the Sanskrit and beyond what Dharmakīrti and Dharmottara’s principle regarding the necessary actualization of causal capacities will permit. The Sanskrit term dṛśyānupalabdhi means merely “the non-apprehension of the perceptible,” while the causal principle seems to exclude the possibility of something’s simultaneously being both perceptible and not perceived. Moreover, the “perceptible” object (or at least, the cognizer’s conception of such an object), precisely insofar as it is not apprehended (and thus does not generate a perceptual awareness), is the cause of the inferential awareness of the absence of the pot. So there must be something here which has real causal capacity with respect to producing that inferential awareness as its effect. As Dharmottara observes in an earlier passage of the same commentary (on NB 2.12), it is only in some sense that the pot (or whatever absent object) is appropriately called “perceptible.” His hesitation about describing the absent object as perceptible in an absolute sense arises from the same principle which underwrote this class of inferences in the first place: the requirement that all genuine causal capacities be immediately actualized. By that principle, we cannot properly call something “perceptible” unless it is actually perceived – which the absent pot is not. Thus, while the inference depends upon causal capacity, “being perceptible” cannot, strictly speaking, be a genuine causal capacity that really belongs to the object whose absence is inferred. If there were a being, like the Naiyāyikas’ Īśvara, who was actually (and not merely potentially) omniscient and who therefore perceived the absent pot (along with all other beings) at every moment, this might seem to resolve the problem: the pot’s perceptibility would be actualized with respect to that special omniscient cognizer. But even if the Buddhist epistemologists were to accept the existence of such a being (which they do not), that would be insufficient to explain what is at work in the relevant class of inferences. Dharmottara would explain this in terms of the object’s being “remote in kind” (svabhāvaviprakṛṣṭa), analogously with the other two ways of being remote: in space or in time. Among the three ways of being remote, the spatial case is easiest to grasp. A pot that is located far from me on some remote mountain may very well be visible to – and, therefore, actually seen by – someone standing beside it on that mountain, but that does nothing to make it perceptible to me. Temporal remoteness will behave similarly: a pot which once existed in the past may well have been seen by someone at that past time, but this too does nothing to make it visible now. For remoteness in kind, the classic Indian example is of a demon who can be perceived by other demons and also by accomplished yogic practitioners (whose perceptual abilities far outstrip those of ordinary humans) but who is imperceptible to ordinary humans. A more contemporary example would be a microorganism or subatomic particle, present at the current place and time, but too small to be seen by the unaided human eye (though again, perhaps, by the accomplished yogin). In all the cases, what matters is not that the pot is perceptible to someone or other but that for the particular cognizer who is seeing the floor, the pot would also be perceptible to her with no additional supporting factors being required. It is for this reason that Dharmottara explains that the pot “is called ‘perceptible,’ even though it does not exist, because ‘being perceptible’ is superimposed. ‘Being perceptible’ is superimposed on the pot, even though it does not exist 445

David Nowakowski

there, when the assumption is made that, if it existed here, it would have to be perceived.”10 Thus, the inference depends not on a property possessed objectively by the pot but rather on a property which is inaccurately but effectively superimposed, based on the epistemic standing of some particular cognizer with respect to the absent or imagined pot – in other words, a property which is both provided by, and relative to, the cognizer. As Kellner (1999, 201) observes, this doctrine of superimposition is associated with Dharmottara in particular by later writers in the Buddhist epistemological tradition.

Doubt and Certainty in Inference-for-Oneself The two preceding sections examined how Dharmottara connects issues involving the cognitive/epistemic processes of perception and inference in his commentaries on the first two chapters of Dharmakīrti’s Epitome of Reason. This final section will briefly address a bridge Dharmottara builds between the second and third chapters of the Epitome of Reason, which connects inference-for-oneself (the inner cognitive process by which each person attains knowledge) with inference-for-others, which appears in the earlier Buddhist epistemologists primarily as a theory of effective debate. This bridge thus connects the philosophy of mind with argumentation and debate theory, again via a move to the cognizer-relative. In theorizing the cognitive process of inference, Dignāga11 famously originates the “triple conditions” meant to specify the conditions under which awareness of one property  – the inferential reason (hetu) – invariably gives rise to inferential knowledge of another property, the target property (sādhya). The conditions are revised and more precisely formalized by Dharmakīrti both in the second chapter of the Epitome of Reason and at much greater length in the Hetubindu. While the formal analysis of Dharmakīrti’s triple conditions is well beyond the scope of this section,12 we can briefly point to one key aspect of the interpretive shift from Dharmakīrti to Dharmottara in terms of two questions: First, what qualifies as an inferential reason? Second, when does inferential knowledge occur for some cognizer? Dharmakīrti makes clear that the triple conditions state purely formal aspects of how the inferential reason relates to the property to be inferred in various groups of property-bearers. A genuine reason is found only in loci where the target is also found and excluded from every locus in which the target is absent while also being found throughout the locus (or loci) in which the target is to be inferred. The conditions thus provide the answer to the first question but not the second. After stating the three conditions themselves in NB 2.5, Dharmakīrti then goes on to state that the cognizer must be certain (niścita) that the conditions obtain, in order for her to infer. Only then does knowledge result. In terms of the stock example of inferring fire from smoke, in Dharmakīrti’s model, the fact of smoke being present only where fire is also present and absent from every place in which fire is absent makes it the case that smoke is in some fairly objective sense an inferential reason for fire. But any particular cognizer will only have the inferential awareness of fire when she is certain that smoke is an inferential reason for fire, even though smoke is (and has been) such an inferential reason for fire all along. Where Dharmakīrti mentioned the need for certainty only once, following the list of the triple conditions, Dharmottara by contrast adds the requirement for certainty to each of the three conditions themselves. In his account, then, a property’s being an inferential reason at all is dependent upon the cognizer’s being free from doubt regarding its status vis-à-vis the other components of the inference. As applied to the stock example, this means that smoke being an inferential reason for fire at all is dependent upon each particular cognizer. And so because certainty is built in to being an inferential reason, upon having such a reason, the resulting inferential awareness of the property to be inferred will result immediately. 446

Dharmottara

This is in keeping with the principle of the necessary actualization of causal capacities, discussed in ”Perceptibility and Cognizer-Relativity.” The term hetu, which in inferential contexts is standardly translated as “(inferential) reason,” also has the meaning of “cause” in a variety of more mundane senses. And it is widely accepted by classical Indian epistemologists, both Buddhists and their rivals, that the inferential reason is in some sense causally responsible for the resulting awareness. While he is not entirely explicit about this point, by making the triple conditions which define the inferential reason depend upon each cognizer’s certainty in this way, Dharmottara solves a parallel problem to that of perceptibility presented in the previous section. Because being an inferential reason depends upon a particular cognizer’s being certain, and therefore inevitably inferring on its basis, the reason’s causal capacity will always be actualized. Neither in the case of inferential reasons nor in the case of perceptible objects will Dharmottara be confronted with causes which fail to produce their effects. How do we know that Dharmottara is making such a shift? In his commentary on the third chapter of the Epitome of Reason (on “inference-for-others”), Dharmottara works out an exhaustive classification of pseudo-reasons (hetvābhāsa): properties which resemble genuine inferential reasons by virtue of meeting some of the triple conditions but which fall short of actually being such reasons insofar as they fail to satisfy one or two of the conditions. To the extent that Dharmakīrti has the beginnings of such a classificatory system, its presence in the chapter on debate and reasoning for others suggests that for him, it is primarily a tool for identifying and critiquing illegitimate moves in a debate or conversation rather than being primarily about the internal cognitive process of inference-for-oneself. While Dharmottara expands on the classification of pseudo-reasons in the third chapter – he is, after all, a commentator, committed to following the root text insofar as possible – he deploys the pseudo-reasons in his formal analysis of the triple conditions in the second chapter of the Epitome of Reason on inference-for-oneself. A standard technique for analyzing definitions in Indian philosophical texts is to show how each word in the definition excludes a particular group of things which fail to meet the definition and which are thus appropriately and effectively removed from the definition’s scope. When all of the words in a single definition are taken together, we see that having excluded everything which does not belong, we are left with a definition of exactly the right scope. In analyzing the triple conditions according to this method, Dharmottara explains that each term excludes particular ways of being a pseudo-reason. And since his classification is exhaustive, when all the types of pseudo-reasons have been ruled out, we are left with only genuine inferential reasons. Importantly, the term “certain” (niścita) plays such a role just as much as every other term, and it plays a different role, excluding different pseudo-reasons, in each of the three conditions. The need for certainty is thus built into the defining statements of what it is to be an inferential reason. Being such a reason, on Dharmottara’s account, is clearly and explicitly cognizer-relative. Dharmottara’s method of analyzing the triple conditions would survive as at least one accepted account of inference-for-oneself through the end of the Buddhist epistemological tradition in India, as attested by its prominent presence in Mokṣākaragupta’s Tarkabhāṣā,13 albeit with some confusion on the part of the latter thinker.

Conclusion In all three of the topics discussed in this chapter, we have seen Dharmottara as a deeply systematic commentator, expanding and perhaps radically revising Dharmakīrti’s epistemic 447

David Nowakowski

theories in order to resolve tensions between various core positions in the Buddhist epistemological tradition. In so doing, Dharmottara displays a consistent tendency toward defining key positions in terms of specific individual cognitive agents and away from more absolute and impersonal accounts. Through all of these contributions, he sets the stage for later Buddhist epistemologists, including Jñānaśrīmitra, Ratnakīrti, and Mokṣākaragupta, all of whom will adopt and modify important parts of Dharmottara’s work on these topics. With this creative and dynamic approach to philosophical commentary, we see Dharmottara, like so many other Buddhist and non-Buddhist scholars in classical India, inhabiting a thought-world. Dharmottara is not simply telling us what Dharmakīrti himself explicitly said or taught; rather, he is thinking himself fully into this system of thought. Once there, he can explore the implications and interconnections of the various parts of Dharmakīrti’s system and also use the resources of that system to respond to new challenges, problems, and objections. In so doing, the commentator engages with the philosophical system not as a museum piece but as a living tradition.

Notes 1 Jayanta’s play, the Āgamaḍambara, is translated by Dezső (2005); for the historical Dharmottara’s migration to Kashmir, see Dezső (2005, 270) and the references cited therein. 2 Dharmakīrti’s Nyāyabindu, literally “The Drop (or Epitome) of Reasoning,” will be abbreviated NB; Dharmottara’s commentary, the Nyāyabinduṭīkā, will be abbreviated NBṬ. Both survive in Sanskrit; the only complete English translation is the now extremely dated work of Stcherbatsky (1962). 3 The reference to “epistemology and philosophy of mind” is, from the perspective of the Sanskrit traditions, a bit redundant, insofar as the issues raised under the single heading of pramāṇavāda, the study of the instruments of knowledge, include within a single domain many of the topics covered by both these distinct labels in Anglophone traditions. 4 In the technical language of these debates, Dharmakīrti’s view is known as pramāṇaviplava, “the distinction of the instruments of knowledge,” while the rival view is called pramāṇasaṃplava, “the mixing-up, or running-together, of the instruments of knowledge.” 5 Both this and the following translation are slightly modified from Franco and Notake (2014), with my italics. 6 In the language of later Indian scholars, each pot-continuum is a “vertical universal” composed of pot-moments, while the potness common to all pot-continua is a “horizontal universal” composed of vertical universals. 7 Jñānaśrīmitra and Ratnakīrti are particularly (though not exclusively) engaged with this question in their treatises on “the exclusion of others” (anyāpoha). Jñānaśrīmitra’s monograph on exclusion has been translated by McCrea and Patil (2010), along with an admirably clear and accessible introduction that explicitly discusses the evolution of the theory from Dharmakīrti to Dharmottara to Jñānaśrīmitra. For Ratnakīrti’s work on exclusion and the dual objects of cognition, again in explicit comparison with Dharmottara, see McAllister 2014, 2017. 8 For Ratnakīrti’s application of “remoteness in space, time, or kind” in the context of challenging the Naiyāyikas’ Īśvara-inference, see his Īśvarasiddhidūṣaṇa, translated in Patil 2001, 307–83. For his application of this concept regarding other minds, see his Santānāntaradūṣaṇa, translated and discussed in Kajiyama 1965. 9 NBṬ on NB 2.14, translated in Kellner 1999, 199. 10 NBṬ on NB 2.12 (translation heavily modified from Kellner 1999, 201). 11 Dignāga (c. 480–540) was the founder of the Buddhist epistemological tradition. Dharmakīrti’s own masterwork, the Pramāṇavārttika, is presented as a commentary upon Dignāga. 12 One such technical analysis, with reference to the precise differences between various Buddhist and Nyāya epistemologists, is in Nowakowski (2017). 13 The relevant section of Mokṣākaragupta is translated by Kajiyama (1998, 65–70). For the relation of Mokṣākaragupta and Dharmottara on the triple conditions, see Nowakowski (2017, 350–52).

448

Dharmottara

Bibliography Dezső, Csaba, ed. and trans. 2005. Much Ado About Religion by Bhatta Jayanta. New York: New York University Press. Dreyfus, Georges B. J. 1997. Recognizing Reality: Dharmakīrti’s Thought and Its Tibetan Interpretations. Albany: State University of New York Press. Franco, Eli, and Miyako Notake, eds. and trans. 2014. Dharmakīrti on the Duality of the Object: Pramāṇavārttika III 1–63. Berlin and Zurich: LIT Verlag. Kajiyama, Yuichi. 1965. “Buddhist Solipsism: A  Free Translation of Ratnakīrti’s Saṁtānāntaradūṣaṇa [sic].” Journal of Indian and Buddhist Studies (Indogaku Bukkyogaku Kenkyu) 13 (1): 9–24 (= 435– 420). Reprinted in Yuichi Kajiyama, Studies in Buddhist Philosophy: Selected Papers, edited by Katsumi Mimaki et al., 401–16. Kyoto: Rinsen Book Co., 2005. ———, trans. 1998. An Introduction to Buddhist Philosophy: An Annotated Translation of the Tarkabhāṣā of Mokṣākaragupta. Reprint, with corrections in the author’s own hand. Wiener Studien zur Tibetologie und Buddhismuskunde 42. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Vien. Kellner, Birgit. 1999. “Dharmottara’s Views in the Dṛśya in Dṛśyānupalabdhi.” In Dharmakīrti’s Thought and Its Impact on Indian and Tibetan Philosophy: Proceedings of the Third International Dharmakīrti Conference, edited by Shoryu Katsura, 193–208. Vienna: Österreichischen Akademie der Wissenschaften. McAllister, Patrick. 2014. “Ratnakīrti and Dharmottara on the Object of Activity.” Journal of Indian Philosophy 42 (2–3): 309–26. ———. 2017. “Competing Theories of Conceptual Cognition: Dharmottara and Trilocana vs. Ratnakīrti?” In Reading Bhaṭṭa Jayanta on Buddhist Nominalism, edited by Patrick McAllister, 291–321. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. McCrea, Lawrence J., and Parimal G. Patil. 2006. “Traditionalism and Innovation: Philosophy, Exegesis, and Intellectual History in Jñānaśrīmitra’s Apohaprakaraṇa.” Journal of Indian Philosophy 34 (4): 303–66. ———. 2010. Buddhist Philosophy of Language in India: Jñānaśrīmitra on Exclusion. New York: Columbia University Press. Nowakowski, David. 2017. “Reasons and Doubt in Dharmottara and His Critics.” Asian Philosophy 27 (4): 340–68. Patil, Parimal G. 2001. “Necessity, Naming, and the Existence of Īśvara.” PhD diss., University of Chicago. Shastri, Dwarika Das, ed. 1985. Nyāyabindu of Acharya Dharmakīrti. Bauddha Bharati 18. Varanasi: Bauddha Bharati. Stcherbatsky, Theodore. 1962 [c. 1930]. Buddhist Logic. Volume II. New York: Dover Publications.

449

27 JÑĀNAGARBHA Two Truths Theory, Gradualism, and Mādhyamika Philosophy Ryusei Keira

There have been at least three figures with the name Jñānagarbha in the history of Indo-Tibetan Buddhism. The first was the author of the Commentary on the Distinction Between the Two Truths (Satyadvayavibhaṅgavṛtti; hereafter Commentary) in the early eighth century. The second Jñānagarbha translated mainly Madhyamaka texts into Tibetan in the early ninth century. The third taught the Secret Assembly Tantra (Guhyasamājatantra) to the Tibetan translator Marpa (1012–1096) in the eleventh century (Matsumoto 1978, 109–10; Seyfort Ruegg 1981, 69 fn. 224). In this chapter, we will focus on the first Jñānagarbha, the author of the Commentary.1 According to Tibetan sources, Jñānagarbha was originally from Oḍiviśa (contemporary Orissa in eastern India) and was a teacher of Śāntarakṣita (ca. 725–788) (Schiefner 1869, 198–99; Matsumoto 1978, 111; Seyfort Ruegg 1981, 69 fn. 225). Śāntarakṣita and his disciple Kamalaśīla (ca. 740–795) were strongly influenced by Jñānagarbha’s Commentary and adopted his two truths theory to establish their own Mādhyamika philosophies. Jñānagarbha is also important for further developing the ideas of the Mādhyamika scholar Bhāviveka (sixth century), who first adopted three interpretations of the term “ultimate” ( paramārtha). Bhāviveka, relying on Dignāga’s theory of valid cognition ( pramāṇa), employed the method of “autonomous reasoning” (svatantrānumāna), proving the voidness of all dharmas by means of inference (anumāna). Like Bhāviveka, Jñānagarbha is often categorized within Tibetan Buddhism as a member of what was retrospectively termed the Svātantrika school, which adopted autonomous reasoning as the means of proving voidness. In contrast to Bhāviveka, however, who relied upon Dignāga’s thought and thoroughly rejected Asaṅga and Vasubandhu’s Yogācāra philosophy, Jñānagarbha’s work, drawing on Dharmakīrti’s theory of valid cognition, constitutes the first transition to the syncretic thinking of the eighth century when Śāntarakṣita and Kamalaśīla integrated Vaibhāṣika, Sautrāntika, and Yogācāra philosophies into one coherent Mādhyamika philosophical system. Jñanagarbha, Śāntarakṣita, and Kamalaśīla are referred to by Tibetans as “the three teachers from the East” (rang rgyud (gyi) shar gsum), that is, present-day Bengal, and Tibetan scholars regard Jñānagarbha as one of the three most important teachers of later Indian Mādhyamika philosophy.

Jñānagarbha’s Works The Commentary is generally regarded as Jñānagarbha’s main work. The Verses on the Distinction Between the Two Truths (Satyadvayavibhaṅgakārikā; hereafter Verses) is 450

DOI: 10.4324/9781351030908-40

Jñānagarbha

also attributed to Jñānagarbha, but contemporary scholars do not believe that it was actually a distinct text composed by Jñānagarbha himself; instead, they believe it was extracted when the Commentary was edited (Matsumoto 1978, 135 n. 2; Akahane 2013, 93–94 n. 2). The Verses and the Commentary are extant only in Tibetan translations and can be found in the Dergé (sDe dge) edition (D) of the Tibetan canon, though not in the Peking edition (P). Śāntarakṣita wrote a subcommentary ( pañjikā) on the Commentary, which is also extant only in Tibetan. In the Dergé edition, in addition to the Verses (D3881) and the Commentary (D3882), there are six other works attributed to Jñānagarbha. Of these six, however, two are tantric texts – Śmaśānavidhi (D1282, P2404) and Caturdevatāparipṛcchāṭīkā (D1916, P2779) – which were most likely composed by the Jñānagarbha who was Marpa’s teacher and not the author of the Commentary. According to the Cordier catalogue, the tantric commentary Guhyasamājaṭīkā is also attributed to Jñānagarbha. This Jñānagarbha is also most likely the one who was Marpa’s teacher (Lalou 1933, 13 and 146; Akahane 2003, 34 n. 5, 56 n. 6). Of the six additional texts attributed to Jñānagarbha in the Dergé edition of the Tibetan canon, then, four may have been composed by the author of the Commentary. Two of these four texts are included in the tantra section of the Tibetan canon and are extant only in Tibetan translation. One is the Verses on the Explanation of the Sacred Dhāraṇī of the Perfection by the Infinite Gates (Ārya-anantamukhanirhāradhāraṇīvyākhyānakārikā, D2695, P3519). A  second is an autocommentary on these verses (Ārya-anantamukhanirhāradhāra ṇīṭīkā, D2696, P3520). There are several verses in these two texts that have parallels in the Commentary. For this reason, the Verses on the Explanation of the Sacred Dhāraṇī and the autocommentary may very well have been composed by the same author as the Commentary (Eckel 1987, 61 n. 29; Akahane 2003, 34–37). The third text, the Path for the Practice of Yoga (Yogabhāvanāmārga, D3909, D4538, P5305, P5452), presents a Mādhyamika account of the path (mārga), outlining steps on the gradual path to awakening. The Path for the Practice of Yoga is distinguished by its interpretation of the five phrases (tshig lnga po) explained in the Discourse on the Enlightenment of Mahāvairocana (Mahāvairocanasūtra, D494, P126) as showing the Mādhyamika account of a gradual path (Yogabhāvanāmārga D4a6–7; Mahāvairocanasūtra D256b3–4, P221b3–4, Taisho 18, 46b23–25). A  similar type of gradualism is also presented in the Commentary. Although Jñānagarbha’s definition of relative or conventional truth (saṃvṛtisatya) in the Commentary – namely “what is appearing” ( yathābhāsa/yathādarśana) – is absent from the Path for the Practice of Yoga, there is no compelling reason to deny that the same author composed both texts (Eckel 1987, 34). It is difficult to determine whether the fourth text, the Commentary on the Maitreya Chapter of the Saṃdhinirmocanasūtra (Ārya-saṃdhinirmocanasūtre ārya-maitreyakevalaparivartabhāṣya, D4033, P5535), was composed by the author of the Commentary. The sūtra that it comments upon, Clarifying the Hidden Intent (Saṃdhinirmocanasūtra), is an important text primarily associated with thinkers identifying with the Yogācāra school such as Asaṅga and Vasubandhu. These thinkers regard the sūtra as a primary source of their philosophy, particularly its accounts of definitive meaning (nītārtha), interpretative meaning (neyārtha), the three natures (trisvabhāva), and the three kinds of non-nature (trividhā niḥsvabhāvatā). Their understanding of these features inspired the doctrinal conflict between what later became two distinct schools, the Mādhyamika and the Yogācāra. The author of the Commentary on the Maitreya Chapter does not seem to present an interpretation consistent with the Mādhyamika philosophy of the Commentary. Rather, he seems to present a standard Yogācāra interpretation. At this point, there is not sufficient reason for determining whether they were both composed by the same author (Eckel 1987, 31–34). 451

Ryusei Keira

Two Truths Theory While Jñanagarbha addresses a number of questions in Buddhist philosophy, he is most famous for his account of the two truths. In the first two verses of the Commentary, Jñānagarbha grounds his account in the words of the Buddha and Nāgārjuna (ca. 150–250 CE). In the Fundamental Verses on the Middle Way (Mūlamadhyamakakārikā), Nāgārjuna distinguishes between two truths: Relying upon two truths, buddhas teach the Dharma: the conventional truth of the world and the truth from the ultimate perspective. Those who do not understand the distinction between these two truths do not understand what is true in the buddhas’ profound teachings. The ultimate is not taught without depending on the conventional. Without understanding the ultimate, one will not attain nirvāṇa. (XXIV.8–9) According to Jñānagarbha, however, even though the Buddha and Nāgārjuna distinguished the two truths, not only non-Buddhists but also his coreligionists such as Dharmapāla (ca. 530–561), Sthiramati (ca. 510–570), Devendrabuddhi (seventh century), and Śākyabuddhi (ca. 660–720) have misunderstood the distinction. This is why, he says, he is motivated to write the Commentary. For, according to Jñānagarbha, those who understand the distinction do not misunderstand the Buddha’s teachings; can accumulate an assemblage of both merit and wisdom ( puṇyajñānasaṃbhāra); can accomplish what benefits themselves and others (svaparārthasampat); and will achieve the goal, that is, becoming a buddha. From Jñānagarbha’s perspective, understanding the distinction between the two truths is necessary to achieve the Mahāyāna ideal of actually becoming a buddha. In what follows, I will focus on this distinction, which Jñānagarbha held to be so important, between ultimate truth and relative, or conventional, truth.

Ultimate Truth (paramārthasatya) Ineffable Ultimate Truth According to his commentators, Bhāviveka and Candrakīrti (seventh century), Nāgārjuna explains the nature of ultimate reality (*paramārthatattva) or ultimate truth ( paramārthasatya) in his Fundamental Verses on the Middle Way: Not to be attained by means of another, peaceful, not fabricated through conceptual fabrication ( prapañcair aprapañcita), devoid of conceptualization, not having different meanings – this is the nature of reality. (XVIII.9) Nāgārjuna argues that the ultimate is understood individually and directly and is not attained dependently on anyone else. It also does not have many different meanings because ultimately entities have the same nature, that is, voidness. For Nāgārjuna, the most important aspect of the ultimate seems to be that it is free from conceptual diversification or fabrication ( prapañca). Because it is free from conceptualization, it is peaceful and beyond language: it is ineffable. On the level of the ultimate, Nāgārjuna argues, because the domain of objects of consciousness is not established, there is nothing to be expressed – like nirvāṇa, the nature of things (dharmatā) is not produced, nor does it cease (XVIII.7). 452

Jñānagarbha

Corresponding to Nāgārjuna’s account of ultimate truth, Jñānagarbha says that reality (tattva), that is, the ultimate truth, cannot be grasped conceptually and is free of the net of all conceptualizations (11b). He also notes in the additional verse (antaraśloka) 5ab on verse 11: In reality, there is nothing that can be expressed at all. And in the Commentary on verse 7, he quotes the following line from the Akṣayamatinirdeśasūtra: The reality is that in which even minds do not function. Jñānagarbha thus seems to accept Nāgārjuna’s account of ultimate truth as free from conceptual diversification or conceptualization. However, Jñānagarbha also emphasizes that “the ultimate does not appear at all in any cognition” (v. 5cd). According to the commentary on verse 5cd (D4b1; cf. Eckel 1987, 157, 15–17), the ultimate does not appear even to the wisdom of the omniscient one (sarvajñājñāna). Therefore, it is said in the Dharmasaṅgītisūtra, “Not seeing anything, this is the seeing of reality.” Jñānagarbha’s view in verse 5cd, that “the ultimate does not appear at all in any cognition,” is the primary reason for his rejection of the Yogācāra interpretation of this line of the sūtra. The Yogācāra interpretation is undertaken on the basis of their theory of the three natures of entities: the imagined nature ( parikalpitasvabhāva), dependent nature ( paratantrasvabhāva), and thoroughly established nature ( pariniṣpannasvabhāva).2 Yogācārins, such as Asaṅga, Vasubandhu, Sthiramati, and Dharmapāla, interpret “not seeing anything, this is the seeing of reality” to be showing that the nonseeing of the imagined nature is the seeing of the ultimate reality of entities. And this, they argue, is the thoroughly established nature, which is defined in Yogācāra texts as the dependent nature being free from the imagined nature. Yogācārins claim that the appearance of the thoroughly established nature is the seeing of ultimate reality. Against the Yogācāra interpretation, Jñānagarbha holds that the ultimate does not appear at all. According to him, the ultimate reality of entities is the nonappearance of any nature of entities, that is, the nonproduction of all entities. Therefore, the appearance of the thoroughly established nature is precisely not the seeing of the ultimate reality (Keira 2004, 99 fn. 151 and 47–86). In his Subcommentary on the Distinction (Satyadvayavibhaṅgapañjikā, D18b4; hereafter Subcommentary), Śāntarakṣita calls the ultimate reality consisting in the nonappearance “the ineffable ultimate” (aparyāyaparamārtha).

The Ultimate Taken as Wisdom While Jñānagarbha accepts the Nāgārjunian account of the ultimate as free of conceptualization, he also draws on Bhāviveka’s method to understand and have access to the ultimate by means of wisdom arising from reasoning, nonconceptual meditation, and so forth. Jñānagarbha interprets this wisdom as a secondary meaning of the ultimate. Jñānagarbha accepts the three interpretations of paramārtha, a compound consisting of parama and artha, that was first adopted in Mādhyamika thought by Bhāviveka. According to Bhāviveka, paramārtha can be interpreted as: (1) an appositional (karmadhāraya) compound, (2) a case-determined (tatpuruṣa) compound, and (3) a possessive (bahuvrīhi) compound.3 The first interpretation is that parama-artha is ultimate ( parama) and is also the object (artha) to be understood as the truth of selflessness (nairātmya). According to the second interpretation, paramārtha is the object (artha) of the ultimate ( parama), that is, supreme 453

Ryusei Keira

nonconceptual wisdom. According to the third interpretation, it is the wisdom which has the ultimate ( paramārtha) as its object. The third interpretation allows Bhāviveka to say that not only the supreme nonconceptual wisdom which has the ultimate as its object but also the wisdom arising from hearing, reflection, and meditation is regarded as the ultimate because the wisdom conducive to the understanding of the ultimate is consistent with the ultimate, due to its object being ultimate (Prajñāpradīpa, D228a3–6). Although any wisdom itself is regarded as conventional, there is a secondary ultimate that is taken as wisdom because it leads one to understanding and having access to the ineffable ultimate. In the Commentary, commenting on verse 4ab, Jñānagarbha applies the third interpretation. Thus, he argues that correct reasoning (nyāya), that is, inferential cognition of the ultimate truth consisting in the nonproduction of entities, ought to be regarded as paramārtha. In his Subcommentary, Śāntarakṣita writes that the reasoning or inferential cognition which has the ultimate truth as its object is called the ultimate, which is conducive to the understanding of the ultimate (don dam pa dang mthun pa’i don dam pa). With the establishment of the third interpretation as a precondition, commenting on verse 4ab, Jñānagarbha then argues that the object, that is, the concept of the nonproduction of entities and so forth, which is determined by means of correct reasoning, is also regarded as the ultimate. This is because correct reasoning is ultimate ( parama) and can also be the object (artha), just as a perception ( pratyakṣa) can be a cognition and also be an object. This can be regarded as the first of the three previous interpretations, the appositional interpretation of parama-artha, which presupposes the establishment of its possessive interpretation. Moreover, in verse 9ab, Jñānagarbha claims that although the nonproduction, that is, negation of the production, which is determined by means of correct reasoning, is regarded as the ultimate or reality (tattva) because it is conducive to the understanding of the ultimate or reality, that negation itself is only conventional when one examines it by means of reasoning. Indeed, when the object to be negated, that is, the production of entities, is not existent, it is clear that the negation of it is also not existent in reality (v. 9cd), since it is absurd that negation would have no object to be negated. Since the production of entities is completely imagined, the negation of the completely imagined production is also imagined, and therefore, because the negation is not existent in reality, it is just conventional (v. 10abc).

On the Qualifier “Ultimately” Jñānagarbha uses the qualifier “ultimately” ( paramārthatas) or “really” ( yang dag par) in his arguments aiming to prove the Mādhyamika position that all entities do not ultimately have any intrinsic nature. Bhāviveka is regarded as the first scholar who adopted this qualifier in articulating Mādhyamika proof-statements. However, Bhāviveka did not explicitly state which terms in the proof-statements should be qualified by this qualifier. This qualifier was rejected by Candrakīrti, who regarded it as incomprehensible. When Candrakīrti critiqued Bhāviveka, he interpreted the qualifier in Bhāviveka’s reasonings as qualifying the negation, that is, “non-having,” or “being without.” According to Candrakīrti, if the intrinsic nature of all entities is ultimately negated, it would follow absurdly that the conventional could not exist and that therefore the Mādhyamika inferences would either invalidate the non-erroneous perceptions of conventional entities or would themselves be invalidated by the perceptional experiences of such conventional things. Jñānagarbha, however, does not hold that the qualifier should qualify the negation of intrinsic nature. According to Jñānagarbha, the qualifier qualifies “having production” (utpāda) or “having intrinsic nature” (svabhāva). That is, he aims to prove the proposition that all entities do

454

Jñānagarbha

not have ultimately existing intrinsic natures, that is, intrinsic natures superimposed upon conventional things by the opponents as being ultimately existent (Keira 2004, 30–38). He writes: We, i.e., the Mādhyamikas, negate only those things imagined on the basis of philosophical texts and the like. That is, we negate aspects that do not appear to cognitions to which things like visible matters (rūpa) appear, aspects such as real production and mental appearances, primordial matter ( pradhāna) and the transformations ( pariṇāma) of the elements. When we negate those imagined things, we do not incur any invalidation by direct perception and the like at all. (Commentary on v. 29) We negate the primordial matter and so forth which appear to conceptualization and have been imagined by others as being really the cause of entities. When we negate the things imagined by others, the qualifier “really” does not qualify the negation. Therefore, how would the conventional be nonexistent? (additional verses 1–2 on v. 30) In the Mādhyamika proof-statements, the objects to be negated are the opponents’ superimpositions or conceptual imaginations. Therefore, they are not existent in reality. Since the negation of the opponents’ superimpositions is also imagined, the negation is not existent in reality. Therefore, the negation is also just conventional. However, as explained earlier, the wisdom that arises from the reasoning which proves the negation is regarded as a secondary ultimate because it is conducive to the understanding of the ineffable ultimate.

Conventional, or Relative, Truth (sam  vrtisatya) According to Jñanagarbha, only those things appearing ( yathābhāsa/yathādarśana) in consciousness are conventionally true (commenting on verse 3cd). This is his definition of conventional truth. A thing like a pot, as everyone sees, is established as being true in the conventional sense because, corresponding to one’s seeing of a pot, one understands the external pot (Commentary on v. 3cd). Jñānagarbha distinguishes between two kinds of conventional truth. The first is “correct conventional truth” (*tathyasaṃvṛtisatya). The second can be understood as precisely not the correct conventional truth because calling something the “incorrect (atathya) . . . truth” (i.e., the truth which is incorrect) would not make sense (*atathyasaṃvṛtisatya, ad v. 8). This type of distinction between two kinds of conventional truth can also be found in Candrakīrti’s Introduction to Madhyamaka (Madhyamakāvatāra, VI.24–26), Commentary on the Introduction to Madhyamaka (Madhyamakāvatārabhāṣya, La Vallée Poussin 1907–1912, 103, 7–106, 1), and In Clear Words (Prasannapadā, La Vallée Poussin 1903–1913, 493, 1–5). Bhāviveka, on the other hand, does not seem to have distinguished two kinds of conventional truth. Still, he believes that there can be something that is conventionally false in contrast to what is conventionally true. For example, Bhāviveka argues that it is impossible even conventionally that Īśvara, the creator God of some Indian traditions, could be the cause of the production of the whole world (Prajñāpradīpa, D51b2). Jñānagarbha and Śāntarakṣita use “the excluded from the correct convention” ( yang dag pa ma yin pa’i kun rdzob; *atathyasaṃvṛti) in their works (Akahane 2013, 87–89). The term “false convention” (log pa’i kun rdzob; *mithyāsaṃvṛti) does not appear in the Jñānagarbha’s

455

Ryusei Keira

Commentary. Nor does it appear in Śāntarakṣita’s Subcommentary, his Ornament of the Middle Way, or his Auto-Commentary on the Ornament of the Middle Way.4

Correct Conventional Truth vs. Imagined Things Jñānagarbha gives an account of what he calls “correct conventional truth” in verse 8abc. A  mere entity (vastumātra) which is free from imagined things and arises dependently, he writes, should be known as the correct conventional truth. Commenting on verse 8abc, he argues: “Imagined things mean ‘real arising’ and so forth, ‘real appearance in consciousness,’ ‘transformation of the primordial matter’ ( pradhāna), ‘real elements’ (bhūta), and so forth.” That is, “imagined things” – here, primarily the ideas of non-Buddhist philosophers – are a kind of false thing excluded from the correct conventional truth; it may be the case that some groups of philosophers believe that these ideas are true, but they are false. In contrast to these things imagined by other philosophers, which he regards as false even conventionally, Mere entities are free from these imagined things. This is because the mere entities free from imagined things have the capacity for causal efficacy (arthakriyāsamartha) to produce an entity, which capacity corresponds to appearances. Mere entities which arise dependent on causes and conditions should be known as the correct conventional truth. That is to say, it is correct that all entities (artha) which causally appear commonly in the consciousnesses of people from the infantile on up, are the correct conventional truth, because external entities are existent corresponding to the appearances in people’s consciousnesses. (Eckel 1987, 160, 10–16) According to Śāntarakṣita, in the Subcommentary, correct conventional truth must meet the following two conditions (D24a1–2): (1) It must be a thing appearing in consciousness, that is, a thing which appears/arises dependently. (2) It must be a mere entity which has the capacity for causal efficacy. The first condition for correct conventional truth, then, is that a thing appears in consciousness. This is in contrast to the ultimate truth, which is precisely not the appearing of things in consciousness: “Not seeing anything, this is the seeing of reality.” Furthermore, Śāntarakṣita argues in his subcommentary (D38b6) that what has the nature of appearance is acceptable when it is not analyzed (avicāraramaṇīya/avicāraikaramya) from the ultimate perspective. That is, an appearance can be accepted as true, but when it is analyzed from the ultimate perspective, we can realize that it is not as we had thought it to be. According to the second condition, correct conventional truth is established when it meets the first condition of appearing to consciousness and when the external entities which correspond to the appearances in consciousness are established. These entities are established by having a capacity for causal efficacy. They must also be free from the imagined nature. An entity free from the imagined nature – that is, a mere entity – can appear in the direct nonconceptual perceptions of many people. When the common appearance of an entity is not established, people’s common understanding of it cannot be established. One cannot start a debate when nothing appears in common to both parties (Commentary on vv. 18–19; Eckel 1987, 173, 23–174, 1). 456

Jñānagarbha

In contrast to mere entities, imagined things, such as primordial matter (namely things superimposed in reliance upon doctrinal systems), cannot appear in direct nonconceptual cognition and therefore cannot be the basis for a common understanding. If these imagined entities were to appear commonly to the minds of participants in debates, and the proponent and opponent were to share a common perception, there would never be any difference of opinion regarding doctrine and therefore no debate. If the participants had a debate, their debate would be invalidated by means of direct perception and so forth (Commentary on v. 8d; Eckel 1987, 160, 19–24). Jñānagarbha thus concludes in verse 8d: Imagined things are not the correct conventional truth. Imagined things do not meet the first condition of appearing to consciousness, nor do they meet the second condition of being causally efficacious. Imagined things – for Jñānagarbha, typically the things postulated in doctrinal systems by philosophers – are thus not mere entities and therefore cannot be even conventionally true. Moreover, conventional entities, although equal in clear appearance, should be divided into correct and incorrect entities on the basis of their capacity or incapacity for causal efficacy (v. 12). When one sees a mirage (marīci) and believes that one is looking at water, although the mirage is clearly appearing in direct nonconceptual cognition, one will not be able to get to the illusory water because mirages have no capacity for causal efficacy to actually get to water and are thus deceptive (visaṃvāda). By this logic, mirages, double moons in the perception of people suffering from eye disease, and so forth do not meet the second condition of being causally efficacious and are therefore outside of correct conventional truth even though they may meet the first condition of appearing to consciousness. How can one distinguish between what is deceptive and what is non-deceptive (avisaṃvāda) with reference to causal efficacy? Discussing verse 12 in the Commentary, Jñānagarbha argues that we should distinguish between what is deceptive and what is non-deceptive according to what is commonly acknowledged to be deceptive or non-deceptive with reference to causal efficacy because the efficacy is also ultimately without intrinsic nature. Thus, because there is no ultimate grounding of conventional truth, one should follow the conventional consensus of the world, based on empirical observation and reasoning, in determining what is casually efficacious and what is not.

The Conventional (sam  vr ti) as an Erroneous Cognition Candrakīrti was the first Mādhyamika scholar to argue that the conventional (saṃvṛti) is a form of ignorance (ajñāna), a claim which characterizes all possible entities that appear in consciousness (In Clear Words on vv. 24–28; La Vallée Poussin 1903–1913, 492, 10–11). According to the Great Commentary Concerning the Teachings (Abhidharmamahāvibhāṣā), this idea that all conventional cognition was a form of ignorance was developed by Sanskrit grammarians: “The grammarians say that this conventional cognition is the ignorance which covers things. It is just like, for example, the bowls which cover the things inside the bowls. Therefore, it is named saṃvṛti (covering)” (Taisho vol. 27, 548b21–22). The same interpretation can be found in the Commentary. Jñānagarbha writes, for example, in verse 15ab that saṃvṛti is considered the cognition by which or in which reality is covered. In the Commentary (Eckel 1987, 171, 3–4), he argues that the cognition covering reality, which is commonly acknowledged in the world (lokapratīti), is regarded as saṃvṛti (Wogihara 1932, 976, 3–4). Quoting the Sūtra on the Descent into Laṅkā (Laṅkāvatārasūtra) and the Perfection of Wisdom Sūtra (Prajñāpāramitāsūtra), 457

Ryusei Keira

Jñānagarbha characterizes saṃvṛti as an erroneous cognition (bhrānti) or mistaken cognition (viparyāsa) (Keira 2016, 47–48). According to Jñānagarbha, then, it is precisely this erroneous and mistaken cognition that is commonly acknowledged in the world (lokapratīti). However, this does not mean that saṃvṛti is necessarily that which is acknowledged by the majority of the world. Jñānagarbha notes in the commentary on verse 15cd (Eckel 1987, 171, 18–20) that mistaken cognition is lokapratīti, and this lokapratīti means that it is true in the realm where valid reasoning can be properly employed. In this respect, his account differs from Candrakīrti’s; Candrakīrti accepts what is acknowledged by the majority of the world (lokaprasiddha) even though it is unacceptable when examined by means of valid cognition (La Vallée Poussin 1903–1913, 67, 7–69, 5, 1907–1912, 179, 14–181, 7).

Gradualism on the Buddhist Philosophical Path Asaṅga, Vasubandhu, Sthiramati, Dharmapāla, and other Yogācāra philosophers interpret the term “mind-only” (cittamātra) as meaning the non-existence of external objects. Neither Bhāviveka nor Candrakīrti, however, accept the Yogācāra interpretation of this term. In his Verses on the Heart of Madhyamaka (Madhyamakahṛdayakārikā), Bhāviveka argues that cittamātra means that there is no agent (kartṛ) and enjoyer (bhoktṛ) other than mind (citta) (V.28cd). And in the Commentary on the Introduction to Madhyamaka, Candrakīrti states that cittamātra simply means that “mind alone” is preeminent among the dharmas (on VI.87; Eckel 1987, 185, 6–20 and 190, 5–8). Thus, although Bhāviveka and Candrakīrti both reject the standard Yogācāra interpretation of cittamātra, they have different understandings of the process of establishing the wisdom of the selflessness of all dharmas. In the Lamp of Wisdom (Prajñāpradīpa), Bhāviveka argues that one should simultaneously understand the selflessness or nonexistence of both external objects and of cognition or mind (D247a1–3; Prajñāpradīpaṭīkā D353a7–354a8). On the other hand, Candrakīrti presents an account of the path as gradual, involving first understanding the selflessness of external objects and then later understanding the selflessness of mind. According to Candrakīrti, the Yogācāra understanding of cittamātra in the sūtras as meaning the nonexistence of external things depends on the interpretative meaning of the teaching and not on its definitive meaning. In his Introduction to Madhyamaka (V.94cd) and the commentary, Candrakīrti argues that the Buddha taught the doctrine of cittamātra to help people remove their attachment to external objects (La Vallée Poussin 1907–1912, 195, 1–9). The Buddha only gradually (rim gyis; *krameṇa) leads people to an understanding and experience of the absence of intrinsic nature (niḥsvabhāvatā), that is, selflessness of all dharmas. Thus, the Buddha first taught the selflessness of external things and then taught the selflessness of mind or cognition so that those on the path could more easily understand the selflessness of mind (Commentary on the Introduction to Madhyamaka on VI.96; La Vallée Poussin 1907–1912, 199, 1–11). Candrakīrti seems to accept the interpretative or provisional meaning of cittamātra as a helpful means to understand the selflessness of mind. Jñānagarbha presents an account of the path of understanding as gradual that is similar to Candrakīrti’s, although it is not clear whether Jñānagarbha knew of Candrakīrti’s work. For example, in the Commentary, Jñānagarbha writes on verse 32: The Buddha dispels people’s ideas of real entities by means of a gradual teaching of five aggregates (skandha), eighteen realms (dhātu), twelve spheres (āyatana), mindonly, and the selflessness of all dharmas, according to the mental capacities of his listeners, and taught people the liberation from bondage. (Eckel 1987, 183, 18–21) 458

Jñānagarbha

In this gradual philosophical path, cittamātra seems to mean the nonexistence of external objects, although it is not clear whether he uses it in its interpretative meaning. Just after his comments on verse 32, in verse 33ab he quotes the Yogācāra commentator Sthiramati (sixth century), who, in his interpretation of Vasubandhu in the Commentary on the Thirty Verses on Representation (Triṃśikābhāṣya), writes: Imagination (rtog pa) is the mind and mental factors which belong to the three worlds and which function with the superimposed forms of external objects. (Buescher 2007, 108, 8) As Sthiramati notes, this imagination is explained as the imagination of what is unreal (abhūtaparikalpa) in Distinguishing the Middle from the Extremes (Madhyāntavibhāga), one of five texts attributed to Maitreya, the future Buddha, who along with Asaṅga and Vasubandhu was regarded as a founder of the Yogācāra tradition. In the Commentary on the Thirty Verses on Representation, this sentence serves as the reason for stating that everything – all cognizable objects – is the mind representing to itself (vijñaptimātra); this is the doctrine of mind-only. It is significant that Jñānagarbha quotes this sentence from the Commentary on the Thirty Verses on Representation because it implies that he accepts Vasubandhu’s justification for the establishment of the cittamātra, or the nonexistence of external objects. According to Śāntarakṣita in his Subcommentary, this imagination, which is the mind and mental factors in the three worlds, is what is appearing, that is, conventional (D46b7–47a1). Ultimately, then, it is without intrinsic nature. Neither Jñānagarbha’s Commentary nor Śāntarakṣita’s Subcommentary gives a comprehensive explanation of the gradual path itself. We can, however, draw on other texts attributed to Jñānagarbha, especially the Path for the Practice of Yoga, which presents the bodhisattva path in some detail. There, Jñānagarbha claims that first, by relying upon the Buddhist śrāvaka teachings of aggregates, realms, and spheres associated with Vaibhāṣikas and Sautrāntikas, one can transcend the non-Buddhist understanding of permanent entities such as a self (ātman). Next, by relying upon the teaching of the mind-only associated with Yogācāra figures such as Vasubandhu and Asaṅga, one can transcend the teaching of śrāvakas. Finally, relying upon the Mādhyamika teaching of the selflessness of all dharmas, one can transcend the mind-only framework and directly understand the nonproduction and voidness of all external and internal dharmas. In Path for the Practice of Yoga, Jñānagarbha’s account of Buddhist gradualism is not fundamentally different from the system of his successors, Śāntarakṣita and Kamalaśīla, whose work also draws on multiple Buddhist traditions (Kajiyama 1978).

Jñānagarbha’s Philosophical Position and Future Research Tibetan scholars have proposed a wide array of interpretations of Jñānagarbha’s two truths theory, especially his philosophical position on convention (saṃvṛti); as a result, Jñānagarbha is regarded as representative of several different Tibetan doxographical categories. I do not take this variety of interpretations to suggest that these Tibetan accounts are somehow wrong. Rather, I  believe that Jñānagarbha’s text actually lends itself to each of these multiple and competing Tibetan interpretations. One interpretation that can be found in the Commentary, later associated with what was termed the Sautrāntika-Mādhyamika position, conventionally accepts the existence of external objects and the cognition in which the object appears can be established. Many Geluk scholars, such as Jamyang Zhepa (’Jam dbyangs bzhad pa, 1648–1722), count Jñānagarbha 459

Ryusei Keira

as a Sautrāntika-Mādhyamika. Their interpretation is justified by Jñānagarbha’s comments on verses 3cd and 8abc, where he argues that external entities exist because they correspond to appearances in people’s consciousnesses. A second interpretation, the position of the Mādhyamika who proceeds in accordance with what is acknowledged in the world, can also be justified by Jñānagarbha’s text. Commenting on verse 15, which characterizes saṃvṛti as an erroneous cognition, Jñānagarbha argues that erroneous cognition is commonly acknowledged in the world (lokapratīti). The Kadampa scholar Üpa Losel (dBus pa blo gsal; ca. 1265–1355) interpreted Jñānagarbha in this way (Mimaki 1982, 27–54). A third interpretation, associated with what came to be called the Yogcācāra-Mādhyamika position, can also be established on the basis of the Commentary. As Jñānagarbha writes in reference to verse 23, When one examines atoms in terms of their separate parts, one understands that they too do not exist at all. So there can be no real cause for the designation of something like a tree. Therefore, saṃvṛti is nothing other than appearances and does not depend on any real causes. This should be accepted. (Eckel 1987, 176, 17–21) Here, negating the existence of atoms, Jñānagarbha argues that appearances in consciousness do not depend on any external real causes like atoms. And, as we saw earlier, at the conventional level, he accepts the mind-only doctrine explained in verses 32–33ab and the commentary. This interpretation of Jñānagarbha as a Yogācāra-Mādhyamika is defended by the Sakya scholars Gorampa Sonam Senge (Go rams pa dSod nams seng ge; 1429–1489) and Büton Rinchendrup (Bu ston rin chen grub; 1290–1364). Note that Jñānagarbha seems to accept conventionally the establishment of reflexive cognition (sva-saṃvitti, -saṃvid, -saṃvedana), which means that cognition is cognized by itself and does not rely upon another cognition to cognize itself (Akahane 2004, 100, 12–102, 16). These differing Tibetan interpretations of Jñānagarbha’s philosophical position on saṃvṛti can all be justified by passages in the Commentary, even as he very clearly defines conventional truth as that which appears to consciousness. One might be tempted to argue that rather than any of the more elaborate Tibetan interpretations, Jñānagarbha’s insistence on just the “appearance in consciousness” is the most important characteristic of his account of saṃvṛti. His account of lokapratīti also should be understood as the appearance of erroneous cognition covering reality. Therefore we should not regard lokapratīti as a fundamental characteristic of Jñānagarbha’s presentation of saṃvṛti. After proving that there is no valid cognition to establish real entities (antaraśloka 1 on v. 13) and that those entities are invalidated by valid cognition (v. 14), Jñānagarbha argues in verse 15ab and the commentary that saṃvṛti is erroneous cognition. From verse 15ab, his view of external entities seems to shift from the Sautrāntika notion that these external entities exist to the Yogācāra account that these apparently real entities grasped by erroneous cognition are completely nonexistent (v. 15cd). Thus, appearances do not depend on any external real causes like atoms (Commentary on v. 23), physical things like visible matter (rūpa) are the appearances of the cognition free from imagination (rtog pa) (v. 30cd), and external entities are imagined by the mind (vv. 32–33ab). Taking the position that saṃvṛti is appearance, he seems to explain two levels of saṃvṛti: the Sautrāntika realistic level and the Yogācāra idealistic or mind-only level, the latter of which corresponds to the gradually ascending steps to remove attachment to real external entities. When we recognize that Jñānagarbha understands saṃvṛti on two different levels, we can be justified in suggesting that all of the Tibetan interpretations of the two truths in the Commentary may be only partial. 460

Jñānagarbha

It will be up to future research to further clarify Jñānagarbha’s philosophy. Jñānagarbha scholarship will benefit from greater understanding of how he was influenced by Dharmakīrti’s refutation of external objects and what has been called a “sliding scale of analysis” (Kellner 2011), according to which one can have multiple ontological accounts which are hierarchically ordered. Clarifying Jñānagarbha’s philosophical position will also enable us to understand differences and commonalities in the philosophies of Jñānagarbha, Śāntarakṣita, and Kamalaśīla and thereby help us better understand the history of late Indian Mādhyamika.

Notes 1 According to the Nyāyabinduṭīkāṭippaṇī, there was a scholar named Jñānagarbha who developed an interpretation of the theory of mental perception (mānasapratyakṣa), although it is not clear whether this is the same Jñānagarbha as the author of Commentary on the Distinction or not (Mimaki 1988, 248–49). 2 For more on the Yogācāra framework of the three natures, see Jonathan C. Gold’s chapter in this volume. 3 For more on Bhāviveka’s account of the ultimate, see Karl-Stéphan Bouthillette’s chapter in this volume. 4 In his Abhisamayālaṃkārālokā Prajñāpāramitāvyākhyā, the eighth-century Buddhist philosopher Haribhadra does not use mithyāsaṃvṛti. He uses atathyasaṃvṛti, although in his text, one can find log pa as a Tibetan translation of atathya (Wogihara 1932, 594, 21; P267a1). Kamalaśīla may have been the first Buddhist author who started to use the term mithyāsaṃvṛti (Madhyamakāloka D230a4–5; Kano and Li 2017, 154, 9–10 for Munimatālaṃkāra).

Bibliography Akahane, Ritsu (赤羽律). 2003. “Nendai-kakutei no shihyō toshite no avicāraramaṇı̄ ya” (年代確定の指 標としての avicāraramaṇı̄ ya). Nanto Bukkyō (南都仏教) 83: 33–59. ———. 2004. “Kyūkyokuteki-shinri to sezokuteki-shinri: Junyānagaruba no ni-shinri-setsu to chibetto ni okeru shisōteki-tachiba” (究極的真理と世俗の真理: ジュニャーナガルバの二真理説とチベ ットにおける思想的立場). Tetsugaku Kenkyū (哲學研究) 577: 80–114. ———. 2013. “The Two Truth Theory of Jñānagarbha.” Memoirs of the Research Department of the Toyo Bunko 71: 69–107. Buescher, Hartmut. 2007. Sthiramati’s Triṃśikāvijñaptibhāṣya: Critical editions of the Sanskrit text and its Tibetan translation. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Eckel, Malcolm David. 1987. Jñānagarbha’s Commentary on the Distinction Between the Two Truths: An Eighth Century Handbook of Madhyamaka Philosophy. Albany, NY: State University of  New York Press. Inagaki, Hisao. 1987. Anantamukhanirhāra-dhāraṇī Sūtra and Jñānagarbha’s commentary: A study and the Tibetan text. Kyoto: Nagata Bunshōdō. ———. 1999. Amida Dhāraṇī Sūtra and Jñānagarbha’s commentary. Kyoto: Ryukoku Gakkai. Kajiyama, Yūichi. (1978) 1989. “Later Mādhyamikas on Epistemology and Meditation.” In Mahāyāna Buddhist Meditation: Theory and Practice, edited by Minoru Kiyota, 114–43. Honolulu: University of Hawai’i Press. Reprinted in Y. Kajiyama, Studies in Buddhist Philosophy: Selected papers, edited by Katsumi Mimaki, 99–128. Kyoto: Rinsen Book Co. Kanō, Kazuo, and Li, Xuezhu. 2017. “Bonbun-kōtei ‘Muni-ishu-shōgon’ Daiisshō (fol. 59v4–61r5): ‘Chūgan-kōmyō’ sezoku no teigi-kasho-itsubun” (梵文校訂『牟尼意趣荘厳』第一章 (fol. 59v4–61r5): 『中観光明』 世俗の定義箇所佚文). Mikkyō Bunka (密教文化) 239: 141–60 (7–26). Keira, Ryusei. 2004. Mādhyamika and epistemology: A study of Kamalaśīla’s method for proving the voidness of all dharmas. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien. ———. 2016. “ ‘Chūgan-kōmyō-ron’ (Madhyamakāloka) Kōshuchō-daiisshō ‘Seiten ni yoru issaihōmujishōsei no shōmei’ no Kenkyū (1); Wayaku, Chūkai, Chibettogo-koutei-tekisuto” (『中観光明 論』 (Madhyamakāloka) 後主張第1章「聖典による一切法無自性性の証明」の研究 (1)： 和 訳・註解・チベット語校訂テキスト). Acta Tibetica et Buddhica 9: 1–121. Kellner, Birgit. 2011. “Dharmakı̄ rti’s Criticism of External Realism and the Sliding Scale of Analysis.” In Religion and Logic in Buddhist Philosophical Analysis, ed. H. Krasser et al., 291–8. Vienna: Verlag der Österreichischen Akademie der Wissenschaften.

461

Ryusei Keira Lalou, Marcelle. 1933. Répertoir du tanjur d’après le catalogue de P. Cordier. Paris: Bibliothèque Nationale. Matsumoto, Shirō (松本史朗). 1978. “Jñānagarbha no ni-tai-setsu” (Jñānagarbha の二諦説). Bukkyōgaku (佛教學) 5: 109–37. Mimaki, Katsumi. 1982. Blo gsal grub mtha’: Chapitres IX (Vaibhāṣika) et XI (Yogācāra) édités et chapitre XII (Mādhyamika) édité et traduit. Kyoto: Zinbun Kagaku Kenkyūsho. ———. 1988. “Kyōryōbu (経量部).” In Iwanami-kōza Tōyō-shisō 8: Indo-Bukkyō 1 (岩波講座 東洋思 想 8: インド仏教 1), edited by Gadjin M. Nagao et al., 226–60. Tokyo: Iwanami-shoten. La Vallée Poussin, Louis de. (1903–1913) 1970. Mūlamadhyamakakārikās (Madhyamikasūtras) de Nāgārjuna: Avec la Prasannapadā, commentaire de Candrakı̄rti. St. Petersburg: Imprimerie de l’Académie Impériale des Sciences. Reprint, Osnabrück: Biblio Verlag. ———. (1907–1912) 1970. Madhyamakāvatāra par Candrakı̄rti: Traduction tibétaine. St. Petersburg: Imprimerie de l’Académie Impériale des Sciences. Reprint, Osnabrück: Biblio Verlag. Schiefner, Anton. 1869. Tāranātha’s Geschichte des Buddhismus in India aus dem tibetischen Übersetzt. St. Petersburg: Buchdruckerei der Kaiserlichen Akademie der Wissenschaften. Seyfort Ruegg, David. 1981. The Literature of the Madhyamaka School of Philosophy in India. Wiesbaden: O. Harrassowitz. Tauscher, Hermut. 1999. Phya pa chos kyi seng ge, dBu ma shar gsum gyi stong thun. Vienna: ­Arbeitskreis für Tibetische und Buddhistische Studien. Wogihara, Unrai. (1932) 1973. Abhisamayālaṃkārālokā prajñāpāramitāvyākhyā: The work of Haribhadra. Tokyo: Toyo Bunko. Reprinted in 1973, Tokyo: Sankibo.

462

28 ŚĀNTARAKS․ITA

Climbing the Ladder to the Ultimate Truth Allison Aitken

Introduction to Śāntaraksita’s Life and Works The scholar-monk and prolific author Śāntarakṣita (c. 725–788)1 left a lasting and significant impact on both Indian and Tibetan Madhyamaka Buddhist philosophy. He is known for his synthesis of Nāgārjuna’s (c. second century) Madhyamaka with elements of Dignāga (c. 480– 540 CE) and Dharmakīrti’s (c. seventh century) tradition of logic and epistemology as well as Yogācāra idealist ontology. Śāntarakṣita’s works are characterized by an emphasis on the indispensable role of rational analysis on the Buddhist path as well as serious and systematic engagement with competing Buddhist and non-Buddhist schools of thought. Śāntarakṣita is said to have been abbot of the great monastic university of Nālandā in Magadha (present-day Bihar, India) and counted Kamalaśīla (c. 740–795) and Haribhadra (late eighth century) among his most prominent students. Yet aside from this, few biographical details about Śāntarakṣita’s life in India remain. He did, however, play a central role in the early transmission (snga dar) of Buddhism to Tibet, and numerous semi-legendary reports of his activities there survive. The earliest accounts agree that, upon receiving an imperial invitation to Tibet from King Trisong Detsen (khri srong lde btsan) (742–797?), Śāntarakṣita oversaw the establishment of the first Tibetan Buddhist monastery at Samyé (bsam yas),2 serving as its abbot and ordaining the first Tibetan Buddhist monks into the Mūlasarvāstivāda monastic order (c. 779), whereupon he became known in Tibet as the “Khenpo (mkhan po)/ Ācārya Bodhisattva,” or “Abbot Bodhisattva.” According to Tibetan sources, Śāntarakṣita’s own ordination lineage proceeds as follows: Śāriputra → Rāhula → Nāgārjuna → Bhāviveka (c. sixth century) → Śrīgupta (c. seventh century) → Jñānagarbha (early eighth century) → Śāntarakṣita.3 This lineage also reflects philosophical affinities among these authors, with the later figures influenced by Bhāviveka’s interpretation of Nāgārjuna’s Madhyamaka (which would later come to be known as *Svātantrika-Madhyamaka).4 A sizeable corpus is attributed to Śāntarakṣita, spanning a range of genres and subject matters, including Madhyamaka metaphysics, logic and epistemology, Buddhist path literature, tantra, as well as several praises. His two most important independent treatises are the Compendium of True Principles (Tattvasaṃgraha, hereafter Compendium = TS) and the Ornament of the Middle Way (Madhyamakālaṃkāra, hereafter Ornament = MA) together with an autocommentary (Madhyamakālaṃkāravṛtti = MAV). Kamalaśīla authored lengthy commentaries ( pañjikā-s) DOI: 10.4324/9781351030908-41

463

Allison Aitken

on both the Compendium (TSP) and the Ornament (MAP). In practice, Śāntarakṣita’s basic texts (particularly the Compendium) are standardly read with the aid – and thus through the lens – of Kamalaśīla’s commentaries, such that Śāntarakṣita’s thought and that of his principal student are often inextricable. Several notable commentaries are also attributed to Śāntarakṣita, including one on Jñānagarbha’s The Distinction Between the Two Truths and its Autocommentary (Satyadvayavibhaṅga and Satyadvayavibhaṅgavṛtti = SDV and SDVV, Satyadvayavibhaṅgap añjikā = SDVP), as well as a commentary on Dharmakīrti’s The Logic of Debate (Vādanyāya) titled Commentary on the Logic of Debate: Elucidation of Its Meaning (Vādanyāyaṭīkā Vipañcitārthā).5 Of all the works attributed to Śāntarakṣita, only the Compendium, his commentary on Dharmakīrti’s The Logic of Debate, and a tantric-cum-epistemological work, Establishing the Truth (Tattvasiddhi),6 survive in Sanskrit. Śāntarakṣita’s references to his own works yield the following chronology of composition: Ascertainment of the Ultimate, Compendium, Commentary on the Logic of Debate, Commentary on the Ornament. This suggests that the Commentary on the Ornament represents his most mature thought and is his definitive work on Madhyamaka.7 As noted previously, Śāntarakṣita’s place in the history of Madhyamaka philosophy is perhaps most remarkable for his synthesis of Nāgārjuna’s Madhyamaka with elements from the Dignāga-Dharmakīrti tradition of logic and epistemology together with Yogācāra idealist onto­ logy. Though not the first Mādhyamika to be influenced by Dignāga and Dharmakīrti, Śāntarakṣita looks to be the first Mādhyamika to author a commentary on one of Dharmakīrti’s works. The influence of the Dignāga-Dharmakīrti tradition of epistemology, according to which testimony as a source of knowledge is reduced to a form of inference, is reflected in Śāntarakṣita’s emphasis on the central role of rational analysis in the gradual progression toward a correct metaphysical view. As Śāntarakṣita repeatedly suggests, the ideal reader of his works and the ideal trainee on the Buddhist path is a discerning person ( prekṣāvat), that is, a rational epistemic agent.8 Prior to Śāntarakṣita, Śrīgupta is noteworthy for integrating the Dignāga-Dharmakīrti tradition of logic and epistemology into his presentation of Madhyamaka, but Śrīgupta rejects Yogācāra ontology without qualification.9 And while Jñānagarbha subsequently alludes to Yogācāra conceptual frameworks in his presentation of the Madhyamaka theory of two truths (satyadvaya), viz. the conventional truth (saṃvṛtisatya) and the ultimate truth ( paramārthasatya) (see, e.g., SDVV ad SDV 30), it is Śāntarakṣita who explicitly formalizes the incorporation of Yogācāra ontology into his presentation of Madhyamaka, though relegated to the domain of conventional truth. Śāntarakṣita accepts as conventionally true not only the Yogācāra doctrine that apparently external objects are merely mental in nature (cittamātra) (MA 91–93) but also the Yogācāra claim that cognition is reflexively aware (svasaṃvitti/ svasaṃvedana) (MAV ad MA 91). Following Śāntarakṣita, Kamalaśīla and Haribhadra likewise adopt the Dignāga-Dharmakīrti tradition of logic and epistemology while also taking a conciliatory approach to Yogācāra. Below, we will return to the question of how best to understand Śāntarakṣita’s Yogācāra-Madhyamaka synthesis. Both of Śāntarakṣita’s main treatises, the Compendium and the Ornament, exemplify his wide-ranging and systematic critical engagement with the philosophical views of competing Buddhist and non-Buddhist schools of thought. The Compendium – over 3,600 stanzas organized in twenty-six chapters – in some ways resembles a critical doxographical survey of the eighth-century Indian religio-philosophical landscape, yet such a description does not adequately reflect its dialogical structure or its in-depth engagement with these competing systems.10 The first twenty-three chapters of the treatise examine and ultimately reject a succession of cosmogonical theories, ontological categories, semantic theories, epistemological theories, and candidate 464

Śāntaraks∙ita

sources of knowledge advanced by Nyāya, Vaiśeṣika, Mīmāṃsā, Sāṃkhya, Jaina, Vedānta, and Lokāyata/Cārvāka traditions, as well as Buddhist traditions such as the Vātsīputrīya.11 The final three final chapters that constitute nearly the second half of the treatise are largely aimed at Mīmāṃsakas, first rejecting their claim that the Vedas lack a human author (apauruṣeya), next critically examining their theory that veridical cognition is self-certified (svataḥprāmāṇya), and finally concluding with an argument in support of the possibility of omniscience. Given the breakdown of the text, the Compendium would seem primarily concerned with competing theories of the Nyāya and Mīmāṃsā, and Śāntarakṣita makes important contributions in formulating Buddhist responses to the Naiyāyika philosopher Uddyotakara (fl. c. 600), as well as the PūrvaMīmāṃsaka Kumārila Bhaṭṭa (fl. c. 660), both of whom were influential critics of Dignāga. The Ornament similarly rejects a range of positions from competing Buddhist and nonBuddhist systems, though it is more metaphysical in its focus than the Compendium, with roughly two-thirds of the ninety-seven stanzas devoted to the neither-one-nor-many argument (ekānekaviyogahetu), which sets out to demonstrate that nothing possesses an intrinsic nature (svabhāva). While the argument takes aim at entities advanced by many of the same nonBuddhist traditions addressed in the Compendium, the Ornament is concerned foremost with competing Buddhist views. In presenting and rejecting the views he addresses in both the Compendium and the Ornament, Śāntarakṣita utilizes a dialectical/pedagogical device of provisionally adopting what he deems to be successively more rational positions in order to reject less rational ones. Sara McClintock has influentially described this method whereby Śāntarakṣita argues from progressively shifting perspectives as a “sliding scale of analysis.”12 To illustrate, in the Ornament, Śāntarakṣita adopts the Sautrāntika representationalist theory of perception in order to reject the Vaibhāṣika direct realist theory of perception. He then assumes the Yogācāra idealist position on which mental representations have no external referents in order to reject Sautrāntika representationalism. Finally, he uses the Madhyamaka account of the nonexistence of fundamentally real cognition to reject Yogācāra theories of the mind and mental content. Although the Compendium includes several allusions to the superiority of the Madhyamaka perspective,13 it might be read as culminating in the Yogācāra perspective. By contrast, the Ornament, in which Śāntarakṣita presents his definitive account of Madhyamaka, devotes more critical attention to Yogācāra than to any other competing system. Yet it is in this same text that Śāntarakṣita presents his provisional endorsement of Yogācāra idealism on the level of conventional truth. The following sections will take up Śāntarakṣita’s contributions to the Madhyamaka theory of two truths in the Ornament.

Ultimate Truth and the Neither-One-Nor-Many Argument The Madhyamaka central commitment, or ultimate truth, is the emptiness of intrinsic nature (svabhāvaśūnyatā), which might be glossed as the universal negation of ontological independence. In other words, according to Mādhyamikas, nothing lays claim to ontological selfsufficiency, which is commonly identified as a necessary condition for fundamentality and substancehood. The Madhyamaka view thus might be described as a kind of thoroughgoing anti-foundationalism as well as a form of substance nihilism. But if there are no ontologically independent or fundamental entities, then whatever there is depends for both its nature and its existence on something else. In his Ornament, Śāntarakṣita makes a major contribution to the Madhyamaka canon of arguments for emptiness with his presentation of the neither-one-nor-many argument. Although he expands on his predecessor Śrīgupta’s more condensed formulation of the argument in the 465

Allison Aitken

Introduction to Reality, it is Śāntarakṣita’s influential Ornament that popularizes the argument in both Buddhist India and Tibet. In his Illumination of the Middle Way (Madhyamakāloka), Kamalaśīla presents the neither-one-nor-many argument among a set of five Madhyamaka arguments for emptiness, which subsequently became known in Tibet as the “five great arguments” for emptiness (gtan tshigs chen po lnga). Nāgārjuna articulates an early precursor to the neither-one-nor-many argument in his Precious Garland (Ratnāvalī), stating: Something is not a unity if it has multiple loci. There is nothing that lacks multiple loci. In the absence of any unity, neither is there a multiplicity. (Precious Garland 1.71) Here, Nāgārjuna argues that whatever is divisible into multiple discrete spatial or temporal loci does not count as a true unity. And everything, he claims, is so divisible. Just as each bit of matter – regardless of how minute – has a right side and a left side, a top and a bottom, and so on; likewise, each moment of time – no matter how brief – has a beginning, a middle, and an end. Otherwise, the existence of the spatially and temporally extended ordinary objects that populate our world – like computers, kangaroos, and cognitions – would be impossible. After all, the thought goes, how could fundamental building blocks that lack spatial/temporal extension ever yield anything that has spatial/temporal extension? And since a plurality presupposes unities as its basic constituents, if there is nothing that is truly one, then neither is there anything that is truly many. Śrīgupta’s expanded formulation of the neither-one-nor-many argument makes explicit the implication that nothing has an intrinsic nature, while also formalizing the argument and defending its soundness according to the Dignāga-Dharmakīrti tradition of logic and epistemology. Closely following Śrīgupta’s Introduction to Reality (TA 1), Śāntarakṣita articulates the central inference of the neither-one-nor-many argument in the opening stanza of the Ornament as follows: In reality, everything that is theorized by our own and other schools of thought is without intrinsic nature, due to lacking an intrinsic nature that is either one or many, like a reflection. (MA 1) The argument poses a destructive dilemma, which says: if anything has an intrinsic nature, then it is either one or many. Śāntarakṣita, in effect, argues that nothing can satisfy either disjunct of the consequent and therefore, by modus tollens, that nothing can satisfy the antecedent. Upon analysis, nothing possesses an intrinsic nature. Three features of the disjunctive property pair are critical for the argument to go through. First, the terms translated as “one” and “many” here (eka and aneka in Sanskrit) are perhaps more precisely rendered as “unity” and “non-unity,” reflecting the fact that they are a mutually exclusive and contradictory pair, conforming to the conceptual, logical, and grammatical structure F and not-F. As Śāntarakṣita makes clear, if anything had an intrinsic nature, then on pain of violating the law of excluded middle, it would have to either be a unity or non-unity: Aside from unity and not-unity, an object’s having some other classification is impossible, since it is established that these two properties are mutually exclusive. (MA 62) 466

Śāntaraks∙ita

Second, the operator, or qualifier, “in reality” (tattvataḥ) in the statement of the central inference clarifies that the target here is a true unity and a true multiplicity. This should be contrasted with a merely conventional status, like the unity of an aggregate such as a flock of sheep or a heap of sand. And, as indicated by Nāgārjuna, a true unity is defined as a mereological simple, that is, something that lacks proper parts, where something has proper parts just in case it is either physically or conceptually divisible. While physical divisibility is more or less straightforward, we can understand that x is conceptually divisible in the mereological sense just in case there are conceptually isolatable proper parts ys that compose x, such that x is the sum of the ys. A true multiplicity, then, is something that has proper parts, the most basic of which are themselves true unities. Finally, a third feature of this property pair is that, unlike most contradictories, unity and non-unity share not only a conceptual priority relation but also a metaphysical priority relation: the existence of a non-unity presupposes the existence of some unities. As Nāgārjuna pointed out in his Precious Garland, a plurality requires singular things as its building blocks. Śāntarakṣita explains, Thus, a “multiplicity” is defined as a composite of unities. If no unity exists, neither does a multiplicity, just like if no trees exist, neither does a forest. (MAV ad MA 61) But if a multiplicity depends for its existence on some unities just like a forest does on some trees, then a multiplicity is not a candidate for ontologically independent being after all. As it turns out, then, true unity is a necessary criterion for ontological independence. And this should not be so surprising: just like ontological independence, true unity is commonly cited as a necessary condition for fundamentality as well as for substantial reality. The neither-onenor-many argument thus reduces to a rejection of true unities, which is to say a rejection of mereological simples.

The Neither-One-Nor-Many Argument Against Material Simples Śāntarakṣita’s rejection of material simples closely follows Vasubandhu’s (c. fourth–fifth century) anti-atomist argument in his Yogācāra work, the Twenty Verses (Viṃśikā 11–15). This section of his neither-one-nor-many argument also features in the “Examination of External Objects” (Bahirarthaparīkṣā) chapter of his Compendium, wherein he assumes the Yogācāra perspective. Śāntarakṣita targets three kinds of views about how atoms aggregate to constitute composites, which recur in debates of this kind in pre-modern Indian philosophy: i. Each atom conjoins with surrounding atoms. ii. Atoms have interceding space between them. iii. Atoms are spatially continuous, neither conjoining with surrounding atoms nor having interceding space between them.14 To each of these views, Śāntarakṣita, in effect, poses the following dilemma: If matter is constituted by fundamental, simple particles, then those particles either face surrounding particles at one and the same locus or at spatially differentiable loci. If, on the one hand, fundamental particles did not have spatially differentiable loci at which to face neighboring particles, and were thus spatially unextended, then they could not compose an extended composite. If, on the other hand, fundamental particles did have spatially differentiable loci at which to face 467

Allison Aitken

surrounding particles (e.g., a right side, a left side, etc.), then they would have spatially discrete parts, which means that they would be composites themselves and could not be fundamental unities after all. As Śāntarakṣita argues in both his Ornament and Compendium, Whether atoms are (i) conjoined, (ii) located at a distance from one another, or (iii) located continuously without interceding space, if the very same part of the central atom in a composite which is facing one atom were also imagined to be facing another atom, then the aggregation of atoms composing mountains, etc., would not be feasible. (TS 1989–90 = MA 11–12) If instead it were accepted that a different part of the central atom faced another atom, then how indeed could an atom like that, i.e., with distinct parts facing different atoms, be truly unitary? (TS 1991 = MA 13) He concludes that there is no account on which matter could be founded in simple particles. And given the metaphysical priority of true unities to a true multiplicity, in the absence of material simples, a material multitude is also precluded.

The Neither-One-Nor-Many Argument Against Mental Simples While the Compendium restricts the subject of the neither-one-nor-many argument to external objects, the Ornament grants the argument a universal scope of application. Having rejected the true unity of any extramental entities within the first fifteen stanzas of the Ornament, Śāntarakṣita devotes stanzas 16–60 to rejecting the true unity of the mind, addressing a variety of Buddhist and non-Buddhist accounts of the mind and mental content, with the argument culminating in a sixteen-stanza section targeting Yogācāra theories. Śāntarakṣita introduces this section, remarking: Even though the Yogācāra view has merit, we shall consider whether such mental entities are to be accepted as real or as satisfactory only when not analyzed. (MA 45)15 The succeeding argument turns on an analysis of the relation between the mind and mental content qua cognition (  jñāna) and mental representations (ākāra). Śāntarakṣita targets two families of views from the Yogācāra tradition on the ontological status of mental representations: i.

Representational realism (*satyākāravāda): representations are real in the same way as cognition is taken to be.16 ii. Representational antirealism (*alīkākāravāda): representations are unreal figments.17 It is important to keep in mind that for Yogācārins, who reject mind-independent material objects, a representation does not actually represent any extramental entity but is simply the intentional object of a cognition. Thus, the question of the ontological status of representations concerns not the represented content (like a desk or a dragon) but rather the representation itself as a feature of the mind.18 Representational realism is commonly associated with a second claim which says that cognition is necessarily and intrinsically endowed with representations 468

Śāntaraks∙ita

(sākāra), while representational antirealism is commonly associated with the claim that cognition is not necessarily endowed with representations, and that invariably veridical enlightened cognition lacks representations (nirākāra).19 In addressing the first view, on which cognition is intrinsically and necessarily endowed with real representations, Śāntarakṣita takes it that, according to this theory, a cognition and its representation are non-distinct, constituting a single subject. Supposing that representations and cognition share a strict identity relation, his reasoning here turns on a version of the law of noncontradiction according to which contradictory properties cannot be predicated of the same subject. Śāntarakṣita observes that a moment of cognition seems to be indisputably simple, and yet the content of cognition looks obviously complex; in any given moment, ordinary experience presents us with a multiplicity of data – a white patch here, a blue patch there, and so on. And, indeed, perhaps the most intuitive representational realist view (the so-called citrādvaita, or “variegated nonduality” theory) says that unitary cognition is non-distinct from its multifaceted representation. But, Śāntarakṣita argues, given the law of noncontradiction, if cognition and representations are non-distinct, it follows that: i. since a representation is manifold, so too is cognition, or else ii. since cognition is truly unitary, so too is its representation. Śāntarakṣita lays out this argument from dilemma as follows: It is difficult to deny that: (i) due to being non-distinct from real representations, cognition must accord with the nature of representations and therefore be manifold. Or else, (ii) due to being non-distinct from unitary cognition, representations would have to be unitary in accord with the nature of cognition. On account of having contradictory properties, ultimately, representations and cognition would have to be distinct. (MAV ad MA 46) Śāntarakṣita takes up the second horn of the dilemma first, arguing that if a representation were truly unitary in accord with cognition, then absurd consequences would follow. For instance, since a simple representation could not be analytically divisible into phenomenal proper parts, we would be unable to conceptually isolate different aspects of our phenomenal field (like the right side and left side of this page). Furthermore, in the absence of phenomenal parts, it could never be the case that one element of our experience was in motion while another was at rest (MA 47–48). On the other hand, Śāntarakṣita reasons, the alternative that cognition is manifold in accord with its complex representation is susceptible to the same kind of argument that he leveled against material atomism: just as an extended material object could not be constituted by unextended material simples, a phenomenally extended representation (like the one you may have of this page) could not be composed of phenomenally unextended building blocks (MA 49).20 And given the metaphysical priority of unity to multiplicity, if there are no simple phenomenal parts, neither can there be a true multiplicity of them. The parts of cognition, then, could not exist in numerical parity with representational parts, since there can be no determinate number of them to which cognition might correspond. He thus concludes that cognition and a real representation could be neither truly one nor truly many. Śāntarakṣita next turns to the representational antirealist view on which cognition is not actually endowed with real representations, which only seem to appear to cognition due to an error (MA 52). This view, Śāntarakṣita argues, is incapable of accounting for ordinary experience, for how could we perceive anything if no percept exists (MA 53–54)? Indeed, he insists 469

Allison Aitken

that there can be no cognition at all in the absence of an intentional object, since cognition is intentional by its very nature; to cognize is to have a cognition of something (MA 55). Moreover, Śāntarakṣita argues, an unreal representation could stand in no relation whatsoever with cognition, whether that be an identity relation or a causal relation (MA 57). If a representation stood in an identity relation with cognition, then given the law of noncontradiction, either: i. since cognition is real, the representation too would be real, or else ii. since the representation is unreal, cognition too would be unreal (MAV ad MA 57). Neither alternative is admissible for the representational antirealist. Furthermore, an unreal representation is no more capable of standing in a causal relation with cognition than an identity relation, since if a representation were caused, then it would be real, but if it had no cause, then there could be no explanation for its appearing with spatiotemporal determinacy or consistency (MA 58). Unable to get the semblance of a defeasible account of cognition up and running on this view, Śāntarakṣita does not even bother to apply the neither-one-nor-many analysis to the representational antirealist position. Having dismissed both the representational realist and antirealist views as untenable, Śāntarakṣita takes himself to have demonstrated that there is no account in which a mental entity could exist as a true unity or a true multitude, and thus concludes that nothing – whether material or mental  – lays claim to ontologically independent being. Instead, the only kind of unity and being that exist are conventional and dependent. To flesh out precisely what Śāntarakṣita means by this, let’s turn to the second of the two truths.

Conventional Truth and Yogācāra Ontology Upon concluding his neither-one-nor-many argument for the ultimate truth, Śāntarakṣita presents his account of conventional truth to clarify that the rejection of ontological independence does not entail an unmitigated nihilism. The term satya translated here as “truth” has a semantic range that is also inclusive of “reality,” and Śāntarakṣita’s account of the conventional is a theory of truth as well as an ontological theory. Yet Mādhyamikas, like Śāntarakṣita, affirm only an ultimate truth and not an ultimate reality qua ontological status. In fact, the ultimate truth as the emptiness of intrinsic nature might be interpreted as the claim that nothing is ultimately real.21 There is thus only one ontological status: conventional reality. According to Śāntarakṣita, whatever is conventionally real (i) has the capacity for causal efficacy (arthakriyāśakti/arthakriyāsamartha), (ii) is dependently originated ( pratītyasamutpanna), and (iii) satisfies our ordinary notions of unity and being only when not subjected to analysis into its final nature (avicāraramaṇīya/avicāramanohara) (MA 64). Conventional truths, then, are pragmatically efficacious claims that concern conventionally real things and which may be verified by our epistemic instruments of perception and inference. With this account, Śāntarakṣita once again follows his predecessor Śrīgupta, who presents the earliest extant formulation of this threefold criterion,22 which was subsequently adopted by Jñānagarbha, Kamalaśīla, Haribhadra, the later Bhāviveka (c. eighth century), Atiśa (982–1054), and others.23 The first criterion for conventional reality – having the capacity for causal efficacy – is a repurposing of Dharmakīrti’s criterion for ultimately reality.24 Though an apparent subversion of Dharmakīrti’s intent, this criterion represents yet another Dharmakīrtian influence on this branch of the Madhyamaka tradition. We will return to the third criterion in treating the role of analysis in Śāntarakṣita’s account of conventional truth below, but it is with the second criterion, being dependently originated, that Śāntarakṣita incorporates Yogācāra ontology into his system. 470

Śāntaraks∙ita

The claim that whatever is conventionally real comes into being in dependence on other things goes back to Nāgārjuna.25 The ontological dependence relation implicated here is inclusive of mereological dependence, mind-dependence, as well as causal dependence, such that every conventionally real thing: (i) comes into being in dependence upon its parts and those parts upon their own parts, ad indefinitum, (ii) is individuated as a conventional unity in dependence upon mental designation, and (iii) is a product of causes and conditions, each of which is in turn a product of its own causes and conditions, ad indefinitum. But, marking a significant departure from prior Madhyamaka accounts, Śāntarakṣita identifies all things involved in causal relations as mental in nature, thereby aligning his presentation of conventional reality with Yogācāra ontology and its central commitment that everything consists in cognition alone (vijñaptimātra):26 Whatever exists as cause and effect is, in fact, merely cognition (  jñāna). Whatever is established by cognition itself exists in cognition. (MA 91) In order to understand how Śāntarakṣita understands Yogācāra ontology to map onto conventional reality, it is necessary to first pin down what precisely he means by affirming the Yogācāra commitment that everything is merely mental (cittamātra). Here are some possible interpretations: i. A phenomenological claim on which the only things relevant to our experience are mental ii. A kind of skepticism which says that we cannot know whether or not there exist any extramental entities iii. An epistemological idealism which says that all objects of knowledge are determined by, or dependent on, the mind and the structure of thought iv. An immaterialism, or metaphysical idealism, on which there are no material things, and the only kinds of things that exist are mental The strongest claim, (iv) immaterialism, is an eliminative idealism insofar as it effectively eliminates, or precludes the existence of, extramental things. The former three options are varieties of non-eliminative idealism insofar as they grant some kind of primacy to the mental but leave open the possibility that extramental things exist. Some have argued for a version of the (i) phenomenological reading of Śāntarakṣita’s Yogācāra by pointing out that he seems to identify conventionally real things with appearances and thus exclusively with what lies within the domain of experience.27 Indeed, in a rather customary Madhyamaka move following an argument for emptiness, Śāntarakṣita insists that the conclusion of his neither-one-nor-many argument does not entail the denial of appearances (MA 78ab). To do so would be tantamount to an implausible thoroughgoing nihilism. But Śāntarakṣita’s nondenial of appearances should not be read as an anti-metaphysical, phenomenological turn; nor does it serve to restrict of the scope of knowledge to the domain of appearances, indicative of skepticism along the lines of view (ii). After all, he rejects as irrational and untenable both direct realist and representationalist theories of perception, which suppose mind-independent external objects to be the direct and indirect objects of perception, respectively. And he does not simply argue that such objects are unknowable or irrelevant to our experience. Rather, he insists that the existence of external objects founded in atoms is incoherent and thus metaphysically impossible. And just as Śāntarakṣita’s arguments rejecting a substantial self and a creator god, for example, are not intended to leave the back door open to 471

Allison Aitken

the existence of such entities lying beyond the scope of our experience, ordinary cognition, or the reach of human knowledge, presumably the same is true of his rejection of material objects. That Śāntarakṣita’s Yogācāra is best read not as (iii) an epistemological idealism (in which all objects of knowledge are determined by, or dependent on, the mind) but as an eliminative (iv) metaphysical idealism is supported by his commentary on MA 91, where he states in no uncertain terms that external material objects do not exist and what we ordinarily take to be external objects simply are one’s own mind. This agrees with the Compendium chapter on the “Examination of External Objects,” wherein Śāntarakṣita assumes the Yogācāra perspective. Here, he lays out two main lines of reasoning against external objects, the first an epistemological argument against the possibility of having knowledge of the existence of external objects founded in atoms and the second an argument against the metaphysical possibility of such objects.28 He transitions from the epistemological to the metaphysical argument with reference to the discerning epistemic agent ( prekṣāvat), once again gesturing to the prominent place of rational analysis in his system: Be as it may that atoms are not established by any source of knowledge, there may nonetheless be doubt. But how should a discerning person come to have certainty about their nonexistence? (TS 1988) Śāntarakṣita aims to pull the rug out from under the external world by rejecting the existence of its purported foundations: material simples.29 That Śāntarakṣita takes the Yogācāra mindonly thesis to negate the existence (and not merely the epistemic accessibility) of real external objects is reiterated in his Yogācāra formulation of the neither-one-nor-many argument: Thus, it is appropriate for discerning individuals to ascertain that atoms are nonexistent, due to being empty of an intrinsic nature that is either one or many, like a lotus in the sky. (TS 1996) According to Śāntarakṣita’s Yogācāra, atoms  – the purported building blocks of material objects – are no more real than lotuses growing in midair.

Śāntaraksita’s Madhyamaka-Yogācāra Synthesis: An Instrumentalist Approach to the Ultimate With Śāntarakṣita’s characterization of Yogācāra ontology in place, let us turn now to the question of how precisely to understand his Yogācāra-Madhyamaka synthesis. That is, does Śāntarakṣita propose a genuine and coherent unification of these two systems, or does he have an instrumentalist story in mind? Given his identification of the conventional truth with Yogācāra ontology, this question is tied up with how best to understand the relation between the two truths, that is, between metaphysical idealism as the conventional truth and the universal negation of ontological independence as the ultimate truth. Are these two truths compatible or contradictory? One Madhyamaka story about the relation between the two truths sees them as contradictory in as much as (i) conventional truths are true according to the non-veridical beliefs and linguistic-conceptual norms of ordinary folks whose understanding is obscured by metaphysical ignorance, while (ii) the ultimate truth reflects the veridical cognition of an ideal epistemic agent. But this is not Śāntarakṣita’s story. The radically unintuitive claim that all things are 472

Śāntaraks∙ita

merely mental in nature is plainly difficult to square with the commonly accepted view of the world.30 The conventional truth of metaphysical idealism can hardly be “read off ” our linguistic-conceptual norms. For Śāntarakṣita, the general consensus is not a guide to what is conventionally true or real. Instead, he insists that a discerning person should seek a correct understanding of conventional reality by employing analysis; rationality is king as a guide to both the conventional and the ultimate. Another Madhyamaka story about the two truths sees them as perfectly consistent and even synonymous. On this view, the ultimate truth as the universal negation of ontological independence is the obverse (viz. negative equivalent) of the conventional truth as the universal affirmation of dependent origination. Yet Śāntarakṣita does not see the two truths as wholly compatible either. This is evident from a comparative analysis of his Yogācāra and Madhyamaka versions of the neither-one-nor-many argument in the Compendium and Ornament. Not only are the subjects of the two arguments different (the Yogācāra iteration takes up atoms and the Madhyamaka argument concerns all things), but the predicate is also different, yielding distinct inference warranting entailment relations (vyāpti):31 Yogācāra entailment relation: Whatever is neither-one-nor-many does not exist. Madhyamaka entailment relation: Whatever is neither-one-nor-many lacks an intrinsic nature. How can Śāntarakṣita consistently maintain that this same neither-one-nor-many reason establishes the non-existence of atoms in his Yogācāra iteration of the argument and the absence of an intrinsic nature of all things in his Madhyamaka formulation? If, on the one hand, the Yogācāra entailment relation holds, then his Madhyamaka neither-one-nor-many argument – which applies this same reason to an unrestricted domain – commits him to a thoroughgoing nihilism. If, on the other hand, the Yogācāra entailment relation does not obtain, then Śāntarakṣita advances a fallacious argument in support of his account of conventional truth. Śāntarakṣita must surely reject the first alternative; nihilism is not an option. He thus looks committed to the second horn of the dilemma, that his argument in support of his view of conventional truth is fallacious. Yet this dilemma stands only if Śāntarakṣita intends to hold the Yogācāra and Madhyamaka perspectives simultaneously.32 But, as he points out, once one has realized the Madhyamaka ultimate truth that all things are equally devoid of ontological independence, one necessarily relinquishes the Yogācāra ontology that grants a privileged position to the mental: Those whose intellectual capacity is not slight and particularly those who are highly industrious will, upon analyzing whether the mind has a unitary or manifold nature, ultimately perceive no such entity. Thus, in reality the mind-only view is not accepted. (MAV ad MA 92) But why bother reasoning our way to metaphysical idealism if it is not ultimately true? According to Śāntarakṣita’s sliding scale of analysis, there are better and worse conventional truths, with more rational stories supplanting the less rational. But there is no definitive conventionally true story. In setting up his characterization of cause and effect as mental in nature in MA 91, Śāntarakṣita states: Whoever accepts the conventional reality of those things that stand in causal relations should analyze what those accepted conventional things are in order to respond 473

Allison Aitken

to fallacious arguments: Are they merely in the nature of the mind and mental constituents or is their nature also extramental? He goes on to explain that among Mādhyamikas, there are two opinions on the matter. The first camp, exemplified by Bhāviveka, accepts conventionally real material and mental entities alike. He introduces the second view – that cause and effect are merely mental – as simply “the opinion of others.” Significantly, he does not reject the first opinion. He does, however, explain the pedagogical utility of the latter. Śāntarakṣita thus sets out an instrumentalist account of the relation between the two truths on which Yogācāra metaphysical idealism is provisionally accepted as conventionally true as an expedient means to arrive at an understanding of the Madhyamaka ultimate truth. Śāntarakṣita’s conventional truth is not determinate but contextual, as indicated by the third criterion for conventional truth that it does not withstand analysis. Importantly, Śāntarakṣita does not follow Candrakīrti (Introduction to the Middle Way, Madhyamakāvatāra 6.35) or Jñānagarbha (SDV 21) in articulating this criterion as the normative claim that one ought not to analyze conventional truths. This would be incompatible with Śāntarakṣita’s progressive path, which the discerning person traverses precisely by analyzing conventional truths. Instead, following Śrīgupta, Śāntarakṣita simply claims that conventional truths satisfy when not analyzed; a conventionally real thing satisfies our notions of reality, independence, and unity when its ultimate nature is not subjected to analysis. We may arrive at progressively more rational conventional truths through analysis, but there is no final conventionally true theory; metaphysical inquiry into the nature of things has no termination point. Still, if there is no determinate conventional truth, what makes one conventionally true theory more rational than another? Śāntarakṣita implies that since nothing withstands analysis, a theory that posits fewer ontologically independent entities is more rational than one that posits more insofar as it is closer to the ultimate truth. In both his Compendium and Ornament, Śāntarakṣita uses analysis to gradually eliminate ontological categories, with the Yogācāra sparse ontology of “mindonly” being, as it were, the last man standing. But at the end of the day, this category too does not withstand analysis. The primary utility of Yogācāra isn’t in what it affirms but in what it denies. From an idealist, rather than a dualist view of conventional reality, it simply takes fewer steps to arrive at an understanding of the ultimate truth, which undermines the fundamentality of anything – whether material or mental. Śāntarakṣita introduces his Yogācāra-Madhyamaka synthesis, stating: Based on the perspective of the mind-only system (cittamātra), one should understand that there are no real external objects. Based on this Madhyamaka system, one should understand that the mind too is utterly selfless. (MA 92) Those who hold the reins of rationality while riding the chariot of these two systems will thereby achieve the state of a genuine proponent of the Mahāyāna. (MA 93) Śāntarakṣita recommends approaching the ultimate truth via the Yogācāra view, but it is just that – an approach to the ultimate, not a definitive or unrevisable claim about the final nature of things. Thus, on this picture, one cannot definitively claim that the two truths are either compatible or contradictory. The conventional truth is a moving target on shifting sands. That Yogācāra idealism is just one among other instructively efficacious conventional stories by 474

Śāntaraks∙ita

way of which one might arrive at the ultimate is supported by the following remark from Kamalaśīla’s Illumination of the Middle Way: Thus, one who is unable to instantaneously realize the fact that all things without exception lack an intrinsic nature should, by temporarily relying on the mind-only system, proceed in stages, beginning with understanding that external objects lack an intrinsic nature. (D 3887, 157a) In this same instrumentalist spirit, Śāntarakṣita (MAV ad MA 70) cites Bhāviveka’s famed metaphor of conventional truth as a ladder to the ultimate truth: Without the ladder of conventional truth, it would not be possible for the learned to ascend to the pinnacle of the palace of reality. (Verses on the Heart of the Middle Way, Madhyamakahṛdayakārikā 3.12) In Śāntarakṣita’s presentation of the two truths, the Yogācāra idealist ontology – though not the exclusive, determinate, or preeminent conventional truth – is nonetheless a highly efficacious penultimate steppingstone to understanding the Madhyamaka ultimate truth.

Conclusion Śāntarakṣita remained a Madhyamaka authority to be reckoned with well into the final period of Buddhism in India. His direct students, Kamalaśīla and Haribhadra, adopted and developed his Madhyamaka-Yogācāra synthesis, and he was recognized as a principal source on the Madhyamaka neither-one-nor-many argument throughout the succeeding centuries by authors such as Dharmamitra (fl. ca. 800), Jitāri (late tenth century), Bodhibhadra (fl. c. 1000), Prajñākaramati (ca. 950–1030), Atiśa, Prajñāmokṣa (ca. eleventh century), Abhayākaragupta (late eleventh–early twelfth century), and so on. Later prominent Yogācāra philosophers including Ratnākaraśānti (ca. eleventh century) and Jñānaśrīmitra (ca. eleventh century) also cited and responded to Śāntarakṣita’s neither-one-nor-many argument, indicating that they considered his attack on Yogācāra theories of the mind and mental content a serious enough threat to necessitate critical engagement several hundred years later. In Tibet, Śāntarakṣita’s Ornament together with Jñānagarbha’s Differentiation of the Two Truths and Kamalaśīla’s Illumination of the Middle Way came to be known as the major works of the so-called “three Mādhyamikas of the East” (dbu ma shar gsum), with commentaries composed on the Ornament by such philosophically and temporally diverse luminaries as Chapa Chökyi Senggé (Phya pa Chos kyi seng ge, 1109–1169), Tsongkhapa Lobsang Drakpa (1357–1419), and Jamgön Ju Mipham Gyatso (1846–1912). Śāntarakṣita’s emphasis on the role of rationality and a progressive path, as well as his synthesis of Madhyamaka with the Dignāga-Dharmakīrti tradition of logic and epistemology, left a lasting and definite impact on Indian and Tibetan Buddhist philosophy.33

Notes 1 See Frauwallner 1961, 141–43. 2 See, for example, the Records of the Ba Clan (sba/dba’ bzhed) for one of the earliest sources (Wangdu and Diemberger 2000). For a much later account from a compilation of sources, see Butön Rinchen Drup’s (bu ston rin chen grub, 1290–1364) History of Buddhism (chos ’byung gsung rab rin po che’i

475

Allison Aitken gter mdzod), translated in Obermiller (1932, 187–92), and Gö Lotsawa Zhönnu Pel’s (’gos lo tsā ba gzhon nu dpal, 1392–1481) Blue Annals (deb gter sngon po), translated in Roerich (1949, 41–44). For a recent compilation of sources on the life and work of Śāntarakṣita, see Eltschinger 2019. 3 Portraits of the members of this lineage were painted on the walls of Samyé; see, for instance, Obermiller 1932, 190; Roerich 1949, 34. 4 Tibetan doxographies commonly present Śāntarakṣita as an exemplar of the so-called *YogācāraSvātantrika-Madhyamaka school of thought and Bhāviveka as the representative of *SautrāntikaSvātantrika-Madhyamaka, while Buddhapālita and Candrakīrti are standardly cited as paradigmatic proponents of *Prāsaṅgika-Madhyamaka. “Yogācāra” in Śāntarakṣita’s doxographical designation indicates his synthesis of Yogācāra ontology into his account of conventional truth, while the *Svātantrika label signifies (in part) his style of argumentation which, following Bhāviveka, deploys independent inferential arguments (svatantrānumāna) to establish his theses rather than exclusively utilizing reductio ad absurdum arguments ( prasaṅga) to undermine the theses of his opponents. On the Tibetan doxographical assignments of Śāntarakṣita, see Seyfort Ruegg 1981, 87–100; Blumenthal 2004, 41–470, and for a critical analysis of this doxographical assignment, see McClintock 2003. For discussions on the historical development of the *Svātantrika and *Prāsaṅgika categories, see Dreyfus and McClintock 2003 and Seyfort Ruegg 2006. 5 Śāntarakṣita’s authorship of this work has been called into question by Tsongkhapa Lobsang Drakpa (Tsong kha pa Blo bzang grags pa, 1357–1419) and Tāranātha (1575–1634) owing to the fact that Kamalaśīla appears to argue against the author of the SDVP in the TSP; see Eckel 1987, 27–31; Blumenthal 2004, 29; Seyfort Ruegg 1981, 68 n. 224. 6 The attribution of Tattvasiddhi to Śāntarakṣita has been called into question by Steinkellner (1999, 356–57). 7 In addition to the texts already mentioned, the other texts attributed to Śāntarakṣita in the Tengyur are: two short praises, the Praise of the Eight Tathāgatas (*Aṣṭatathāgatastotra) and Praise of the Bhagavan: Song of Śrī Vajradhara (*Śrīvajradharasaṃgītibhagavatstotra) together with autocommentary (ṭīkā); two tantric works, Five Great Instructions on Kurukulla Arisen from Hevajra (Hevajrodbhavakurukullāyāḥ Pañcamahopadeśa) and Ritual for the Extensive Recitation of the Previous Aspirations of the Seven Tathāgatas Collected from Sūtras (Saptatathāgatapūrvapraṇidhānaviśeṣavistārakalpavacanavidhi-sūtrāntasaṃkṣepa); and a commentary (vṛtti) on Candragomin’s (seventh century) Twenty Verses on the Bodhisattva Vows (Bodhisattvasaṃvaravṃśaka). In his Commentary on the Ornament, Śāntarakṣita appears to allude to another text he authored titled Ascertainment of the Ultimate (*Paramārthaviniścaya) (Ichigō 1985, 330). As Eltschinger (2019, 384–85) points out, Śāntarakṣita also seems to allude to this work in TS 2083. 8 See McClintock 2010, 58–62, 2013; Tillemans 2011, 153–54, 2016, 143–44 on Kamalaśīla’s elaboration on this concept, particularly in the context of discerning the correct understanding of conventional truth; on the term, prekṣāvat, see also Eltschinger 2007, 137–50, 2014, 195 n. 17, 219–34. 9 On Śrīgupta’s Introduction to Reality (Tattvāvatāra = TA) and accompanying autocommentary (-vṛtti = TAV), see Aitken (Forthcoming). 10 For a short synopsis of the Compendium, see Seyfort Ruegg 1981, 89–90, and for a compilation of secondary scholarship on this work, see Steinkellner and Much 1995, 56–63; Eltschinger 2019, 385–86. 11 Śāntarakṣita presents a summary of the topics of the text in the first six stanzas of the Compendium, with the main points of the first twenty-three chapters glossed as modifying dependent origination and those of the final three chapters modifying the Buddha who taught dependent origination. On the two-part structure of the Compendium, see McClintock 2010, 97–98. 12 On the application of this framework for describing Śāntarakṣita’s method, see McClintock 2003, 2010, 85–91; on this same concept applied to the work of Dharmakīrti, see Dreyfus 1997 and Dunne 2004, 53–79. Blumenthal (2004, 43, 44, 46) articulates this same approach of Śāntarakṣita’s in terms of a “graded ascent of philosophical views,” “multiple levels or stages of provisionality,” and an “ascent through provisional views.” 13 McClintock (2003, 68–76) points to TS 1916–17 as instances where Śāntarakṣita acknowledges the superiority of Madhyamaka. Śāntarakṣita additionally nods to the Madhyamaka tradition in the framing of the text, mirroring Nāgārjuna’s opening to his Fundamental Verses on the Middle Way (Mūlamadhyamakakārikā) in laying out the purpose and structure of the Compendium in its opening stanzas. 14 (i) Vaiśeṣikas defend the conjoined view, (ii) Vaibhāṣikas such as Saṅghabhadra (c. fourth–fifth century) defend the interceding space view, and (iii) Vasubandhu defends the spatially continuous view in the Commentary on the Treasury of Abhidharma (Abhidharmakośabhāṣya) ad 1.43d2, where

476

Śāntaraks∙ita he attributes the position to the Kaśmīri Vaibhāṣikas such as Vasumitra (c. second century CE); Śubhagupta (c. 720–780) also defends this view in his Proof of External Objects (Bāhyārthasiddhi) 50, 52, 53, 56, although Śāntarakṣita cites Śubhagupta in connection to the second view. 15 In his autocommentary, Śāntarakṣita goes on to praise Yogācāra for its theory being confirmed and known by a source of knowledge ( pramāṇa), for serving as a corrective for mistaken views, for its rejection of material simples, and for its agreement with scriptural sources (MAV ad MA 45; Ichigō 1985, 124). 16 MA 46–51; see Dharmakīrti’s Explanation of the Sources of Knowledge (Pramāṇavārttika) 3.209–22 and Śrīgupta’s TA 4. 17 MA 52–60; see TA 5–6, TAV transitional stanzas (antaraśloka) 1–4. 18 In Cartesian terms, the question here concerns the formal reality of thoughts themselves rather than the objective reality of whatever might be represented in thought. 19 *Satyākāravāda (rnam pa bden par smra ba) and *Alīkākāravāda (rnam pa brdzun par smra ba) (literally “theory/proponent of real representations” and “theory/proponent of unreal representations”) are not attested in extant Indic doxographies, where we instead find the Sākāravāda-Nirākāravāda distinction. While these labels were imposed onto diverse sets of thinkers in contriving subschools of Yogācāra, they are nevertheless useful for clarifying the structure of Śāntarakṣita’s argument and the dialectical lay of the land as he understood it. One should be careful to distinguish the use of the terms sākāra vs. nirākāra in the Yogācāra context from the use of this same pair of terms to designate, respectively, representationalist vs. direct realist accounts of ordinary perception among realists about external objects. 20 To the contrary, Berkeley (Principles in Works vol. 2, 98) and Hume (Treatise 1.2.4), for instance, both argue for theories of a minima sensibilia, a kind of phenomenal atomism on which a perception is reducible to indivisible, unextended simples. 21 This account can be traced back to Nāgārjuna, who argues that nothing lays claim to the Abhidharma ontological category of ultimate reality ( paramārthasat), or substantially reality (dravyasat). 22 TA 11; see Aitken (Forthcoming) and (2021a) for an interpretation of Śrīgupta’s version of this threefold criterion. For a comparison of Śrīgupta and Śāntarakṣita on the two truths, see Aitken (2021b). 23 See, for example, Kamalaśīla’s MAP ad MA 64, Haribhadra’s Illuminating the Ornament of Realization (Abhisamayālaṃkārālokā) (Wogihara 1932–35, 594.18–25), Bhāviveka’s Compendium on the Meaning of the Middle Way (Madhyamakārthasaṃgraha) 9–11 and Jeweled Lamp for the Middle Way (Madhyamakaratnapradīpa) 1.4, and Atiśa’s Introduction to the Two Truths (Satyadvayāvatāra) 4. Jñānagarbha also sets out versions of these three criteria (SDV 8, 12, and 21). 24 Explanation of the Sources of Knowledge 3.3ab and Essence of Reasoning (Nyāyabindu) 1.15. 25 Śāntarakṣita cites Nāgārjuna’s Fundamental Verses on the Middle Way 24.18 in support of this criterion. 26 Importantly, Śāntarakṣita does not regard this as an innovation, citing Nāgārjuna’s Sixty Stanzas on Reasoning (Yuktiṣaṣṭika) 21 and 34 (among other sources) in support of this presentation. 27 For example, Garfield (2016) argues that Jamgön Ju Mipham Gyatso (’Jam mgon ’Ju Mi pham rgya mtsho, 1846–1912) reads Śāntarakṣita’s Yogācāra phenomenologically. 28 See Saccone’s 2018 edition, translation, and analysis of this chapter. It is this second, metaphysical argument that follows Vasubandhu’s previously discussed anti-atomist argument from his Twenty Verses 11–15. 29 It is worth pointing out that Śāntarakṣita’s rejection of atoms does not in principle preclude the possibility of conventionally real external objects that are not founded in atoms, though he does not explicitly consider such a scenario, presumably owing to the fact that all external world realists in his intellectual milieu were atomists. 30 To the contrary, Berkeley famously makes a case for subjective idealism as part of his project to restore commonsense. 31 This difference is also reflected in the different examples cited in the two inferences (in which the entailment relation is instantiated), with the Madhyamaka example being a reflection, which lacks an intrinsic nature, and the Yogācāra example being a lotus growing in the sky, which is a nonexistent thing. 32 Since atoms are stipulated by their proponents as partless, and therefore true unities by definition (TS 1992ab), the rejection of the unity of atoms would necessarily preclude their existence. The problem here lies in the sweeping nihilistic implications of the generalized Yogācāra entailment relation when applied to the Madhyamaka all-inclusive subject.

477

Allison Aitken 33 Although Śāntarakṣita’s conciliatory approach to Yogācāra was rejected by many later Tibetan Mādhyamikas in favor of Candrakīrti’s account of conventional truth, his synthesis of Madhyamaka with Dignāga and Dharmakīrti’s logic and epistemology remains influential in Tibetan Madhyamaka cutting across traditions.

Bibliography Aitken, Allison. Forthcoming. Introduction to Reality: Śrīgupta’s Tattvāvatāravṛtti. Harvard Oriental Series. Cambridge, MA: Harvard University Press. ———. 2021a. “No Unity, No Problem: Madhyamaka Metaphysical Indefinitism.” Philosophers’ Imprint 21 (31): 1–24. ———. 2021b. “The Truth About Śrīgupta’s Two Truths: Longchen Rabjampa’s ‘Lower Svātantrikas’ and the Making of a New Philosophical School.” Journal of South Asian Intellectual History 3 (2): 185–225. Blumenthal, James. 2004. The Ornament of the Middle Way: A Study of the Madhyamaka Thought of Śāntarakṣita. Ithaca: Snow Lion Publications. ———. 2009a. “Dynamic and Syncretic Dimensions to Śāntarakṣita’s Presentation of the Two Truths.” Asian Philosophy 19 (1): 51–62. ———. 2009b. “The ‘Neither-One-Nor-Many’ Argument of Śāntarakṣita: A Classical Buddhist Argument on the Ontological Status of Phenomena.” In Buddhist Philosophy: Essential Writings, edited by William Edelglass and Jay L. Garfield, 46–60. New York: Oxford University Press. ———. 2015. “Two Topics Concerning Consciousness in Śāntarakṣita’s Yogācāra-Madhyamaka Syncretism.” In Madhyamaka and Yogācāra: Allies or Rivals?, edited by Jay L. Garfield and Jan Westerhoff, 242–51. New York: Oxford University Press. Dreyfus, Georges. 1997. Recognizing Reality: Dharmakīrti’s Philosophy and Its Tibetan Interpretations. Albany: State University of New York Press. Dreyfus, Georges, and Sara McClintock. 2003. The Svātantrika-Prāsaṅgika Distinction: What Difference Does a Difference Make? Boston, MA: Wisdom Publications. Dunne, John. 2004. Foundations of Dharmakīrti’s Philosophy. Boston: Wisdom Publications. Eckel, Malcolm David. 1987. Jñānagarbha on the Two Truths: An Eighth Century Handbook of Madhyamaka Philosophy. Albany, NY: State University of New York Press. Eltschinger, Vincent. 2007. “On 7th and 8th Century Buddhist Accounts of Human Action, Practical Rationality and Soteriology.” In Pramāṇakīrtiḥ: Papers Dedicated to Ernst Steinkellner on the Occasion of His 70th Birthday, Part 1, edited by Birgit Kellner, Horst Lasic, M. T. Much, and Helmut Tauscher, 135–62. Wien: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. ———. 2019. “Śāntarakṣita.” In Brill’s Encyclopedia of Buddhism, Volume II: Lives, edited by Jonathan A. Silk, Richard Bowring, Vincent Eltschinger, and Michael Radich, 383–90. Leiden and Boston: Brill. Frauwallner, Erich. 1961. “Landmarks in the History of Indian Logic.” Wiener Zeitschrift für die Kunde Süd- und Ostasiens 5: 125–48. Garfield, Jay L. 2016. “Cittamātra as Conventional Truth from Śāntarakṣita to Mipham.” Journal of Buddhist Philosophy 2: 263–80. Ichigō, Masamichi, ed. 1985. Madhyamakālaṃkāra of Śāntarakṣita with His Own Commentary or Vṛtti and with the Subcommentary or Pañjikā of Kamalaśīla. Kyoto: Buneido. ———. 1989. “Śāntarakṣita’s Madhyamakālaṃkāra.” In Studies in the Literature of the Great Vehicle: Three Mahāyāna Buddhist Texts, edited by Luis O. Goméz and Jonathan A. Silk, 141–240. Ann Arbor: Center for South and Southeast Asian Studies, University of Michigan. ———. 2016. “Śāntarakṣita on Satyadvaya.” Journal of Indian and Tibetan Studies 20: 166–94. Jha, Ganganatha. 1937–1939. The Tattvasaṅgraha of Śāntarakṣita with the Commentary of Kamalaśīla. 2 vols. Baroda: Oriental Institute. Kapstein, Matthew. 2001. Reason’s Traces: Identity and Interpretation in Indian and Tibetan Buddhist Thought. Boston: Wisdom Publications. ———. 2009. “Śāntarakṣita’s Tattvasaṃgraha. A Buddhist Critique of the Nyāya View of the Self.” In Buddhist Philosophy. Essential Readings, edited by William Edelglass and Jay L. Garfield, 320–33. Oxford: Oxford University Press. Keira, Ryusei. 2004. Mādhyamika and Epistemology: A Study of Kamalaśīla’s Method for Proving the Voidness of All Dharmas: Introduction, Annotated Translations and Tibetan Texts of Selected Sections of the Second Chapter of the Madhyamakāloka. Wiener Studien zur Tibetologie und Buddhismuskunde, Heft 59. Wien: Arbei­tskreis für Tibetische und Buddhistische Studien Universität Wien.

478

Śāntaraks∙ita McClintock, Sara L. 2003. “The Role of the ‘Given’ in the Classification of Śāntarakṣita and Kamalaśīla as Svātantrika-Mādhyamikas.” In The Svātantrika-Prāsaṅgika Distinction: What Difference Does a Difference Make? edited by Georges B. J. Dreyfus and Sara L. McClintock, 125–71. Boston: Wisdom Publications. ———. 2010. Omniscience and the Rhetoric of Reason: Śāntarakṣita and Kamalaśīla on Rationality, Argumentation, and Religious Authority. Boston: Wisdom Publications. ———. 2013. “Kamalaśīla and Śāntarakṣita on Scripture and Reason – The Limits and Extent of ‘Practical Rationality’ in the Tattvasaṃgraha and Pañjikā.” In Scriptural Authority, Reason, and Action: Proceedings of a Panel at the 14th World Sanskrit Conference, Kyoto, September 1st-5th 2009, edited by Vincent Eltschinger and Helmut Krasser, 209–38. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Obermiller, Eugene. 1932. The History of Buddhism in India and Tibet by Bu Ston: Part II. Heidelberg: Harrassowitz. Roerich, George. 1949. The Blue Annals, Part I. Calcutta: Royal Asiatic Society of Bengal. Saccone, Margherita Serena. 2018. On the Nature of Things: A Buddhist Debate on Cognitions and Their Object. Vienna: Arbeitskreis für tibetische und buddhistische Studien Universität Wien. Sāṅkṛtyāyana, Rāhula. 1935/1936. “Dharmakīrti’s Vādanyāya. With the Commentary of Śāntarakṣita.” Journal of the Bihar and Orissa Research Society: 21–22. Seyfort Ruegg, David. 1981. The Literature of the Madhyamaka School of Philosophy in India. Wiesbaden: Otto Harrassowitz. ———. 2006. “The Svātantrika-Prāsaṅgika Distinction in the History of Madhyamaka Thought.” IndoIranian Journal 49 (3/4): 319–46. Sferra, Francesco. 2004. “Śāntarakṣita: Vātsīputrīyaparikalpitātmaparīkṣa, Disamina del se immaginato dai seguaci di Vātsīputra (Tattvasaṃgraha, 336–49 con il commento [Pañjikā] di Kamalaśīla).” In La Rivelazione del Buddha, vol. II, edited by Raniero Gnoli, 1155–74. Milan: Arnoldo Mondadori. Shastri, Dwarka Das. 1968. Tattvasaṅgraha of Ācārya Shāntarakṣita with the Commentary Pañjikā of Shri Kamalashīla. 2 vols. Varanasi: Bauddha Bharati. Steinkellner, Ernst. 1999. “Yogic Cognition, Tantric Goal, and Other Methodological Applications of Dharmakīrti’s kāryānumāna Theorem.” In Dharmakīrti’s Thought and Its Impact on Indian and Tibetan Philosophy, Proceedings of the Third International Dharmakīrti Conference, Hiroshima, November 4–6, 1997, edited by Shoryu Katsura, 349–62. Vienna: Austrian Academy of Sciences Press. 2001. “Is the Ultimate Cognition of the Yogin Conceptual or Non-Conceptual? Part I: A Critical Edition of the Tantric Tattvasiddhi, Final Section.” In Le Parole e i marmi: Studi in onore di Raniero Gnoli nel suo 70° compleanno, edited by Raffaele Torella, 835–52. Rome: Istituto Italiano per l’Africa e l’Oriente. ———. 2014. The Edition of Śāntarakṣita’s Vādanyāyaṭīkā Collated with the Kundeling Manuscript. Vienna: Arbeitskreis für tibetische und buddhistische Studien Universität Wien. Steinkellner, Ernst, and M. T. Much. 1995. Texte der erkennt-nistheoretischen Schule des Buddhismus: systematische Übersicht über die buddhistische Sanskrit-Literatur II. Göttingen: Vandenhoeck and Ruprecht. Tillemans, Tom T. F. 1982. “The ‘Neither One Nor Many’ Argument for śūnyatā and Its Tibetan Interpretations: Background Information and Source Materials.” Études de Lettres (University of Lausanne) 3: 103–28. ———. 1983. “The ‘Neither One Nor Many’ Argument for śūnyatā and Its Tibetan Interpretations.” In Contributions on Tibetan and Buddhist Religion and Philosophy, edited by Ernst Steinkellner and Helmut Tauscher, 305–20. Vienna: Arbeitskreis für Tibetische and Buddhistische Studien. ———. 1984. “Two Tibetan Texts on the ‘Neither One Nor Many’ Argument for Śūnyatā.” Journal of Indian Philosophy 12: 357–88. ———. 2011. “How Far Can a Mādhyamika Reform Conventional Truth? Dismal Relativism, Fictionalism, Easy-Easy Truth, and the Alternatives.” In Moonshadows: Conventional Truth in Buddhist Philosophy, edited by the Cowherds, 151–66. New York: Oxford University Press. ———. 2016. How do Mādhyamikas Think? And Other Essays on the Buddhist Philosophy of the Middle. Somerville, MA: Wisdom Publications. Wangdu, Pasang, and Hildegard Diemberger. 2000. dBa’ bzhed: The Royal Narrative concerning the bringing of the Buddha’s Doctrine to Tibet. Vienna: Austrian Academy of Sciences Press. Westerhoff, Jan. 2015. “Nāgārjuna’s Yogācāra.” In Madhyamaka and Yogācāra: Allies or Rivals? edited by Jay L. Garfield and Jan Westerhoff, 165–83. New York: Oxford University Press. Wogihara, Unrai, ed. 1932–1935. Abhisamayālaṃkārāloka Prajñāpāramitāvyākhyā (Commentary on Aṣṭasāhasrikā-Prajñāpāramitā) by Haribhadra Together with the Text Commented On. Tokyo: The Toyo Bunko.

479

29 KAMALAŚĪLA Mādhyamika Champion of Magical Reason Sara McClintock

Kamalaśīla’s Tantalizing Traces Unlike the majority of figures in this volume, the monastic scholar Kamalaśīla (c. 740–795) leaves tantalizing historical traces of his embodied existence beyond the many treatises credited to his name and the usual hagiographies. These traces mainly take the form of later Tibetan historiographical writings documenting what has come to be known as either the Great Debate at Samyé or the Council of Lhasa (c. 792–794). This legendary intellectual contest of unknown duration features Kamalaśīla and Heshang Moheyan (Héshang Móhēyǎn), a Chinese rival monk representing the Chan Buddhist tradition, both of whom are said to have been invited to Tibet by the powerful king Trisong Detsen (Khri srong lde btsan, c. 742–797). The earliest and most important of the Tibetan works documenting this event – the historicity of which remains an open question – is the Testament of Ba (dBa’ bzhed), an account of the arrival of Buddhism in Tibet which survives in diverse forms (Wangdu and Diemberger 2000) and upon which later Tibetan historiographical writings appear to rely (Tucci 1958, 10). Other early sources preserved at Dunhuang in both Tibetan and Chinese likewise hint at Kamalaśīla’s existence without, however, naming him – though one text, at least, the Ratification of the True Principle of the Mahāyāna Teaching of Sudden Awakening (Dunwu dacheng zhengli jue) of Wang Xi, refers to the Indian opponent of Moheyan as a “brahman monk” (Demiéville 1952; Seyfort Ruegg 2013), a moniker that fits what we learn of Kamalaśīla from hagiographers. Thus, while much remains uncertain, all sources point to a conflict between Indian proponents of a so-called gradual path and Chinese advocates for a practice allowing for a sudden or “subitist” breakthrough to enlightenment. Not surprisingly, there is no agreement about who actually won the debate, assuming that it actually took place or in any way resembled the traditional accounts. While Chinese documents from Dunhuang identify the Chinese side as the winner (Eltschinger and Marks 2019, 273), Tibetan Buddhist historiographers are unanimous in naming Kamalaśīla the victor, thus ensuring that Tibetan Buddhism would hence forward rely primarily on Indian (and not Chinese) sources for the transmission of Buddhism to Tibet. With this narrative as a backdrop, Kamalaśīla becomes something of a hero of the Tibetan Buddhist philosophical tradition, especially in its monastic and scholastic dimensions. While the facts may be murky, however, there is still quite a bit in the traditional accounts that finds support in Kamalaśīla’s oeuvre. First, Kamalaśīla’s commitment to gradualism is a 480

DOI: 10.4324/9781351030908-42

Kamalaśīla

persistent theme in his writings, as we shall soon see. Second, Kamalaśīla’s faith and expertise in reasoned debate and analysis as a means to resolve doubts is well attested in his numerous philosophical treatises, especially his epistemological works but extending to his Madh­ yamaka compositions as well. And third, his three interrelated tracts, Stages of Cultivation (Bhāvanākrama), take frequent and direct aim at a form of subitism later interpreters identify with the position of the Chinese monk Moheyan. Indeed, it is widely supposed that these three texts on the methods and goals of meditation were composed in Tibet specifically in preparation for the Great Debate, and there is speculation that this fact explains their lack of renown among later Buddhists in India. Sadly, their absence from the scene in India, along with the absence as well of Kamalaśīla’s most important Madhyamaka text, Light of the Middle (Madhyamakāloka), may also corroborate another part of Kamalaśīla’s legend: several sources tell of Kamalaśīla’s demise a few years after the famous debate at the hands of assassins who are said to have “squeezed” his kidneys, thus preventing the monk from returning to India. Whether these murderers were sent by a jealous Moheyan or by non-Buddhist supporters of Tibetan indigenous traditions opposed to the importation of Buddhism to Tibet is a matter on which even the Tibetan sources do not agree (Keira 2004, 2). What is clear, however, is that Kamalaśīla and his teacher Śāntarakṣita (c. 725–788), who is widely credited with having established Samyé Monastery where he ordained the first seven Buddhist monks in Tibet, have both emerged as cultural heroes for Tibetan Buddhists. Many, like Tenzin Gyatso, the current Dalai Lama, see them as key representatives of what the he calls “the Nālandā tradition” in reference to the great North Indian monastery and center of learning where both Indian pandits are said to have taught and to have served as abbot. Our knowledge of Kamalaśīla’s life in India and role at Nālandā, however, is even more sketchy than is our understanding of his activities in Tibet. He is clearly associated with Śāntarakṣita through his writings – having composed two important commentaries on Śāntarakṣita’s chief works – and through the historical legends maintaining that he was called to Tibet upon Śāntarakṣita’s deathbed advice to the Tibetan king. He is also connected, though more remotely, with Śāntarakṣita’s own purported teacher, Jñānagarbha. All three are thought to have hailed from Bengal, with later Tibetans sometimes speaking of them as a group using variants on the phrase the “three SvātantrikaMādhyamikas from the east” (dbu ma rang rgyud gyi shar gsum) (Tauscher 1999, 387). We shall return to the problematic doxographical classification of Svātantrika-Mādhyamika in what follows. For now, we note simply that all that remains beyond these hints about Kamalaśīla’s historical personage are the surviving works attributed to him and the later reception of these works in Tibet. Ironically, despite Tenzin Gyatso’s lauding of the Nālandā tradition and the foundational role of these Indian thinkers in the establishment of Buddhist monastic scholasticism in Tibet, Śāntarakṣita and Kamalaśīla’s lineage appears not to have remained strong in the Land of Snows, where the interruption of their oral lineage combined with the later ascendancy of the so-called Prāsaṅgika-Madhyamaka of Candrakīrti starting in the eleventh century seems to have lessened their influence. This may be particularly true for their epistemological writings, concerning which the fifteenth-century Sakya scholar Gorampa Sönam Sengé (Go ram pa bSod nams seng ge, 1429–1489) said, “Although the abbot Śāntarakṣita and the supreme scholar Kamalaśīla appear to have founded an oral tradition (bshad-sgrol), nowadays, apart from the mere book, there does not appear a continuity of its teaching” (van der Kuijp 1983, 4–5). This statement intriguingly suggests something more about how figures like Śāntarakṣita and Kamalaśīla function as authors. That is, in addition to composing treatises, their philosophical activity is understood to require an embodied dialogical dimension. Reading the mere book ( yig cha tsam) without an authoritative guide empowered to transmit its contents is unlikely to produce the insights toward which these philosophers would like us to progress. 481

Sara McClintock

We are, nevertheless, products of our own time, and this emboldens us to try to understand these works even in the absence of an unbroken transmission lineage. A review of the writings attributed to Kamalaśīla reveals him to be an extraordinary scholar, whose depth of learning and fluency of expression yield a rare combination among philosophers: a thinker who can communicate dense, elliptical arguments with a refreshing mix of clarity and precision. As an author, he has the further virtue of offering frequent, explicit signposts concerning his purposes, presuppositions, and audience. These signposts serve as invaluable instructions to readers while simultaneously reflecting his understanding of the philosophical enterprise: by attending to the signposts, we understand that Kamalaśīla sees the philosophical project as inherently embedded in social and linguistic contexts, thus rendering its value entirely at the conventional (saṃvṛti) or transactional (vyāvahārika) level of reality. This is so even though philosophy itself is also an important tool on the path to the realization of the ultimate ( pāramārthika) reality or the truth of the highest meaning ( paramārthasatya). We shall delve into this topic in greater depth subsequently, as it is crucial to understanding Kamalaśīla’s character as an author committed to a syncretic and gradualist approach.

Stages of Cultivation and Light of the Middle First, however, to get a better picture of Kamalaśīla as an author, we need to give a fuller account of his oeuvre. We have already encountered Kamalaśīla’s two most famous independent treatises: the three interrelated texts known as Stages of Cultivation and the extensive late treatise on Madhyamaka, Light of the Middle. As previously indicated, these works are widely considered to have been composed in Tibet (Keira 2004, 7), and this fact is understood to account for the near total lack of their citation among Indian Buddhists. In the case of Light of the Middle, we do find parallel passages in the twelfth-century Indian scholar Abhayākaragupta’s Ornament of the Sage’s Thought (Munimatālaṃkāra), lending credence to the report that when Atiśa Dīpaṃkaraśrījñāna (c. 982–1054) visited Samyé in the eleventh century, he found the Sanskrit manuscript of Light of the Middle, copied it, and had it sent back to India (Keira 2004, 8). As for Stages of Cultivation, there is some doubt about the first of the three texts in the series, since we find numerous parallel passages in the contemporary Indian scholar Haribhadra’s seminal Light for the Ornament of Realization (Abhisamayālaṃkārāloka); we likewise find parallels in the Instructions on the Stages of Cultivation of the Perfection Vehicle (Pha rol tu phyin pa’i theg pa’i sgom pa’i rim pa’i man ngag; *Pāramitāyānabhāvanākramopadeśa), a work preserved in Tibetan attributed to the relatively unknown Indian author Jñānakīrti (c. 900 CE) and the Sanskrit title for which is unattested (Taniguchi 1992). So it is also possible that Kamalaśīla wrote the work in India or that he had copies sent back, even as the text also appears to address some of the core issues at stake in the legendary Great Debate. Such situations again raise the intriguing problem of the nature of authorship in this period, causing us to reflect that our attributions of authors to particular texts are indeed conventional designations arising from myriad factors, most if not all of which cannot be ascertained with any true certainty. In any case, although both Stages of Cultivation and Light of the Middle are widely revered in Tibet, neither received much direct commentary there. According to the fourteenth Dalai Lama, Tenzin Gyatso, even though Tsongkhapa, the fifteenth-century progenitor of the Gelugpa school of Tibetan Buddhism, quotes extensively from Stages of Cultivation in his own Great Stages of the Path (lam rim chen mo), the transmission lineage for Kamalaśīla’s trio of meditation texts was “fairly rare” in central and southwestern Tibet (Dalai Lama 2001, 23). The Dalai Lama relates how he himself had to go to some effort to obtain the teachings, thus reinforcing the idea that notwithstanding Kamalaśīla’s reputation among Tibetans, the study of 482

Kamalaśīla

his texts had been largely eclipsed. As for Light of the Middle, the only known commentaries in Tibetan are a recently discovered treatise by Chapa Chökyi Sengé (Phya pa chos kyi seng ge, 1109–1169) and a late partial treatment by the Mongolian scholar Ngawang Tendar (Ngag dbang bstan dar, 1835–1915).1 Although Stages of Cultivation has received some sustained attention (Adam 2002; Schmid 2020), perhaps in part due to the fact that two of the three texts survive in Sanskrit (Tucci 1958, 1971), Light of the Middle is extant only in Tibetan and is still relatively unknown. Ryusei Keira’s excellent Mādhyamika and Epistemology (2004), a study and translation of Kamalaśīla’s arguments for emptiness in Light of the Middle, remains the only book-length treatment of this work in English. Still, Keira treats only a relatively small part of this long and masterful text. Philosophically, both Stages of Cultivation and Light of the Middle are extraordinarily rich. Here we touch only on a few of the most salient contributions. Later, after we have reviewed Kamalaśīla’s oeuvre more thoroughly, we shall consider the thorny question of Kamalaśīla’s doxographical affiliation. Here, however, we shall attempt to encounter his philosophy on its own terms. The texts making up Stages of Cultivation contain considerable overlap in their themes, and it is not clear why Kamalaśīla composed three different works.2 For our purposes, we can isolate four distinct themes running through the three works that also characterize Kamalaśīla’s philosophical positions more generally. These themes are (1) that the bodhisattva path is a gradual affair with distinct stages (krama) involving the perfection and integration of method (upāya) and wisdom ( prajñā); (2) that conceptual philosophical activity, including rational analysis (vicāra) and examination ( pratyavekṣā), is integral to the bodhisattva path resulting in buddhahood; (3) that the ultimate nature of reality is to be empty (śūnya) of any essential nature (svabhāva), such that the final realization of that emptiness (śūnyatā) in the perfection of wisdom ( prajñāpāramitā) is necessarily nondual (advaya) and nonconceptual (nirvikalpa) in nature; and (4) that the root of buddhahood is compassion, which when combined with insight into reality (bhūtārtha) results in a state characterized as non-located nirvāṇa (apratiṣṭhitanirvāṇa). This last state is that of the omniscient buddhas, located in neither saṃsāra nor nirvāṇa and thus able to serve the needs of suffering beings by teaching them while remaining undefiled by their unpurified ignorance.3 Kamalaśīla’s gradualism is on full display here, as he insists on the necessity of first using rational analysis to verify the emptiness  – known otherwise as the naturelessness (niḥsvabhāvatā) – of all things before attaining direct nonconceptual realization of that reality in yogic perception ( yogipratyakṣa).4 But how does one verify naturelessness? While Kamalaśīla touches on this problem in the Stages of Cultivation, it is only in his specifically Madhyamaka works, especially Light of the Middle but also his commentary on his teacher Śāntarakṣita’s well-known Ornament of the Middle (Madhyamakālaṃkāra), that he provides a developed set of answers to this question. In these texts, Kamalaśīla sets out the proof statements ( prayoga) that establish emptiness as the incontrovertible ultimate truth ( paramārthasatya) concerning all entities. Developing arguments inherited from the larger Madhyamaka tradition and expressed in the language of the Buddhist epistemological tradition, he details in Light of the Middle five key proof statements for naturelessness.5 The five proof statements – each of which is referenced by way of its reason (hetu) establishing emptiness – are as follows: (1) the “diamond splinter” (vajrakaṇa) reason, which refutes that a thing is produced either by itself, by another, by both itself and another, or without a cause; (2) the reason that refutes the production of either existent or nonexistent things; (3) the reason that refutes that one can produce many, many can produce one, many can produce many, or one can produce one; (4) the reason by virtue of the “dependent arising” ( pratītyasamutpāda) of all things; and (5) the reason of having “neither one nor many” (ekānekaviyoga) essential nature. 483

Sara McClintock

Light of the Middle is structured around two basic divisions: an earlier segment consisting of a series of 83 so-called prior positions ( pūrvapakṣa) raising objections to Kamalaśīla’s Madhyamaka system and a later segment consisting of the so-called subsequent positions (uttarapakṣa) rebutting these objections. The second segment is traditionally understood to be organized into four chapters: a chapter on the proof of emptiness via scripture, a chapter on the proof of emptiness by reasoning, a chapter on the two truths, and a chapter establishing that there is ultimately only one single vehicle (ekayāna) or path to awakening. Not surprisingly, Kamalaśīla addresses the objections with extreme rigor, harnessing the full arsenal of argumentation techniques from the Buddhist epistemological tradition to beat back his Buddhist opponents, some of whom have been identified as generic Vaibhāṣikas and Yogācāras, others tentatively as Dharmakīrti’s early commentators Devendrabuddhi and Śākyabuddhi, one as Dharmapāla, and one even as Dharmakīrti himself.6 Noteworthy here and elsewhere is Kamalaśīla’s embrace of flagship theories and positions associated with the Yogācāra stream of Buddhist thought in India. For example, in Light of the Middle, we see Kamalaśīla accept the Yogācāra theory of the three natures (trisvabhāva) and their attendant three kinds of naturelessness, but we see him doing so in such a way as to interpret all these in light of the Madhyamaka theory of the two truths (satyadvaya). As is well known, the three natures are a Yogācāra scheme to explain our experience of reality consisting of the imagined nature ( parikalpitasvabhāva), the dependent nature ( paratantrasvabhāva), and the perfected nature ( pariniṣpannasvabhāva). Each of these natures is also said to have its own distinct kind of naturelessness: character naturelessness (lakṣaṇa-niḥsvabhāvatā), production naturelessness ( utpatti-niḥsvabhāvatā), and ultimate naturelessness ( paramārthaniḥsvabhāvatā). Kamalaśīla accepts all these categories explicitly. Yet his final position differs from that of the typical Yogācāra thinker in that he wants to emphasize the ultimate emptiness of all three natures and not just the lack of identity in the case of the imagined nature and the lack of production in the case of the dependent nature.7 To take just one example, Kamalaśīla states that “the dependent nature is an entity that is accepted so long as it is not analyzed, that is just as it appears, and that is dependently arisen, like an illusion.”8 By this statement, Kamalaśīla alludes to key characteristics of the conventional advanced by his direct predecessors in Madhyamaka thinking, Jñānagarbha and Śāntarakṣita. He shows that in his view, the Yogācāra category of the dependent nature can be accepted, but that status is nevertheless thoroughly conventional even if it lacks conceptuality. The conventional nature of appearances, even without the presence of language and concepts, is a hallmark of Jñānagarbha and Śāntarakṣita’s Madhyamaka thought.9

The Extensive Commentary on Śāntaraksita’s Compendium of Reality Beyond Stages of Cultivation and Light of the Middle, Kamalaśīla is best known for his massive Extensive Commentary (Pañjikā) on his teacher Śāntarakṣita’s Compendium of Realities (Tattvasaṃgraha). Unlike the previous two works, Kamalaśīla most certainly composed this one in India, before he went to Tibet. The Extensive Commentary, along with the Compendium, which are rarely, if ever, read apart from one another, are staggering in their comprehensiveness, with Śāntarakṣita’s root text coming in at more than 3,600 verses in length and Kamalaśīla’s prose commentary at over 1,000 pages in modern printed editions. Both works are famed not only for their sophisticated argumentation utilizing the epistemological and logical theories of Vasubandhu and Dharmakīrti but also for their extensive engagement with and citation of a wide range of Buddhist and non-Buddhist authors representing a plethora of contemporary Indian philosophical positions. Since both Śāntarakṣita and Kamalaśīla quote opponents directly and often quite faithfully, the works have become valuable resources for the study not 484

Kamalaśīla

only of Indian Buddhism but also of non-Buddhist Indian philosophical traditions whose own root texts have been lost. Kamalaśīla’s unusual practice of directly naming opponents adds extra value. We are extremely fortunate that two medieval manuscripts were preserved in Jain libraries. While the two complete modern editions (Krishnamacharya 1926; Shastri 1968) and the single complete English translation (Jha 1937/1939) all leave significant room for improvement, their publication has been an enormous boon to the field of Indian philosophical studies.10 Some controversy has emerged over how best to read the texts, with a key question emerging regarding how to imagine the original audience for the works. The Compendium of Realities itself is divided into 26 chapters or “investigations” ( parīkṣā) covering major topics in Indian metaphysics, epistemology, and argumentation, and, as is not uncommon during this period, the investigations mostly proceed through the presentation of an opponent’s views, the so-called prior position, followed by section designed to refute those views, the so-called subsequent position. Often within a single investigation we find multiple variants of the prior position, and in some cases, the investigation is itself further subdivided to address variant views. For example, the seventh chapter, the “Investigation of the Self ” (the ātmaparīkṣā), includes six subsections treating the theories of the Self as propounded by proponents of the Nyāya, Mīmāṃsā, Sāṃkhya, Digambara Jaina, Advaita Vedānta, and Buddhist Vātsīputrīya traditions. In each case, Kamalaśīla identifies specific thinkers and also identifies schools of thought by name. His Extensive Commentary is frequently offered as the site of the first appearance of the school name Advaita in a surviving Sanskrit composition. The question of the audience is important not only in terms of getting the intellectual history right, but, perhaps even more so, it is also an important question philosophically. Most likely we will never settle whether the Compendium and its Extensive Commentary were written primarily to win over opponents, to prepare Buddhist scholars for oral debate, or as part of the spiritual training of Buddhists who are “inclined toward wisdom” ( prajñānusārin), as Kamalaśīla puts in his opening essay to the Extensive Commentary (Eltschinger 2014; McClintock 2013). Most likely, all three of these purposes come into play (Marks 2019). What is beyond doubt, however, is that within these works, we see the exemplification of a complex theory of the transactional nature of reason itself according to which all human reasoning, whether undertaken for oneself or another, must start with the lived realities, the embodied experiences, in short, the shared world (loka), of the participants in the debate. To seek an uncontaminated starting point free from the distortions of primordial ignorance would be sheer folly. Yet to eschew what appears in common on the grounds that it is a distortion would be to abandon the commitment to rational inquiry for the sake of the reduction or elimination of the suffering of sentient beings. Kamalaśīla shows a strong commitment to reason throughout the Extensive Commentary, robustly defending the epistemological instruments ( pramāṇa) as developed by Dharmakīrti and his earliest commentators, Devendrabuddhi and Śākyabuddhi. But he shows an equally strong commitment to working within the world of ordinary experience, in one place pointing out that the elaborate philosophical theories of inference of those who espouse formal reasoning (nyāyavādin) are really no different from the day-today forms of reasoning that are undertaken by ordinary persons in the world (McClintock 2019). His argument seems designed to save philosophical inquiry from irrelevance not by elevating it above conventional truths and realities but by confirming its embeddedness with them. Another hermeneutical conundrum concerns the doxographical classification of the Extensive Commentary. While it is generally counted as a work of epistemology (tshad ma = pramāṇa), Tibetan commentators have nonetheless held that it, like the epistemological works of Dignāga and Dharmakīrti, presents a Yogācāra position for its final view. This assessment is made despite the fact that Kamalaśīla clearly indicates much of the argumentation to be offered at the so-called Sautrāntika level of analysis. At this point in the scholastic 485

Sara McClintock

development of Buddhist philosophy, the term Sautrāntika generally refers to a system in which objects of knowledge external to the mind are held to exist, even if they are also understood to be fleeting and infinitesimal. Yet despite this preponderance of Sautrāntika arguments, we also find Kamalaśīla designating certain passages as being from the perspective of the Yogācāra – which he likewise terms Vijñānavāda (“doctrine of cognition”), Vijñaptimātra (“representation only”), or Cittamātra (“mind-only”) – a position in which no objects of knowledge external to the mind are held to exist and which Kamalaśīla clearly judges to be superior to the Sautrāntika. At the same time, a close reading of the commentary also reveals subtle references to Madhyamaka (McClintock 2010, 2019), a position to which the work might be said to gesture as its logical culmination. Whatever the final classification, there is no question that the “sliding scale of analysis” at play in the work is yet another manifestation of Kamalaśīla’s commitment to a vision of rational inquiry as thoroughly conditioned and context dependent.11 An important example of a Yogācāra doctrine defended in the Extensive Commentary is reflexive awareness (svasaṃvedana, svasaṃvitti). This is the idea that awareness has the nature of illumination ( prabhāsvara) and for this reason is always inherently self-aware.12 The doctrine of reflexive awareness is a hallmark of Dignāga and Dharmakīrti’s epistemology, and it is not surprising to see it turn up here as well. For Dharmakīrti, reflexive awareness is the ultimate trustworthy awareness ( pāramārthika-pramāṇa), which makes sense in Yogācāra terms since a mind knowing the real must by definition be knowing the only thing that is real, mind itself.13 Yet this kind of knowing is also radically nondual since knower and known are ultimately one and the same. Kamalaśīla asserts that this reflexive awareness, which amounts in the end to a vision of reality (tattvadarśana), is the primary form of the mind.14 Other forms of trustworthy awareness in which the mind appears to know an object external to itself are accepted in merely conventional (saṃvṛti) or transactional (sāṃvyavahārika) terms.15 In keeping with Kamalaśīla’s gradualism and Madhyamaka leanings, however, we must remember that the acceptance of reflexive awareness as the ultimate trustworthy awareness, as well as the Yogācāra tenet that everything is mind only, are themselves valid only in conventional terms.16 Alternatively, we can take a page from Jay L. Garfield (2016) and recognize that for Kamalaśīla, as for Śāntarakṣita, the Cittamātra position obtains phenomenologically, while the Madhyamaka position obtains on an ontological level.17

Other Works on Madhyamaka and Epistemology in Kamalaśīla’s Oeuvre So far we have considered five of Kamalaśīla’s most famous works: the three texts that all bear the same name and are collectively called Stages of Cultivation; the single long, independent treatise that sets out to establish the Madhyamaka view of emptiness, Light of the Middle; and the detailed, polemical work covering a host of philosophical topics, the Extensive Commentary on Śāntarakṣita’s Compendium of Realities. Of these, the Stages of Cultivation and the Light of the Middle are both classified as belonging to the Madhyamaka (dbu ma) section, and the Extensive Commentary is contained in the Epistemology (tshad ma = pramāṇa) section of the Treatises, or Tengyur (bstan ’gyur) portion of the Tibetan canon. Kamalaśīla is credited with many more texts than these, up to 30 in the Derge edition of the Tengyur. Unfortunately, current scholarship does not allow us to make a clear determination as to whether all these texts should be attributed to a single hand. A number of these texts, however, can be confidently attributed to the same Kamalaśīla. These include several more works in the Madhyamaka section, including the Extensive Commentary on the Ornament of the Middle (Madhyamakālaṃkārapañjikā), a 486

Kamalaśīla

substantive prose subcommentary on his teacher Śāntarakṣita’s well-known Ornament of the Middle (Madhyamakālaṃkāra). Also in this section are the two shorter works Light of Reality (Tattvāloka) and Demonstration of the Naturelessness of All Things (Sarvadharmaniḥsvabhāvasiddhi), both of which again utilize the tools of Buddhist epistemology to demonstrate the correctness of the doctrine of emptiness. Finally, in the Madhyamaka section, we find two short works highlighting the centrality of meditation and spiritual practice, the Introduction to the Cultivation of Yoga (Yogabhāvanāvatāra) and the Cultivation of the Mind of Awakening (Bodhicittabhāvanā). These again point to Kamalaśīla’s project of integrating Madhyamaka’s nonfoundational insights into a gradual path of yogic cultivation that insists on the necessity of adhering to conventionally correct ethical and epistemological procedures along the way. The Cultivation of the Mind of Awakening discusses a passage from the Secret Communion Tantra (Guhyasamājatantra), a point which raises an important question about Kamalaśīla’s familiarity and involvement with the Buddhist esoteric stream and prompts Marks and Eltschinger (2019, 274) to note that “any pronouncement concerning this work’s authenticity seems premature.” More broadly, one might say that any pronouncement concerning the engagement or lack thereof with tantric activities on the part of Kamalaśīla and his Indian confrères is premature. There remains one more work by Kamalaśīla in the Epistemology section of the canon, namely the Summary of Prior Positions in the Drop of Reason (Nyāyabindupūrvapakṣa­ saṃkṣipta). This interesting tract, which sadly does not survive in Sanskrit, seeks to illumine Dharmakīrti’s basic guide to reasoning, the Drop of Reason (Nyāyabindu), not through direct commentary but rather through providing an account of the opponents to whom Kamalaśīla understands Dharmakīrti’s treatise to respond. Marks (2019, 263) wonders about the purpose of this little-studied work. Was it “an assignment given to a close disciple, a young teacher’s lecture notes, the work of an overeager student, or perhaps even preliminary research toward the Pañjikā [i.e., the Extensive Commentary]?” While we may not be able to answer such questions, Kamalaśīla’s interest in the dialogical nature of reasoning is again evident here. For Kamalaśīla, it appears, stating the rules for constructing a proper inference, for example, makes sense only in the context of an actual argument with someone who holds a different position. Reasoning, in other words, is always a thoroughly contextual and conditioned affair. The Summary of Prior Positions in the Drop of Reason is also notable for including a short exposition on a famous scriptural quotation that appears in both the Compendium and in its Extensive Commentary. The quotation epitomizes Śāntarakṣita and Kamalaśīla’s attitude toward reason and religious authority and is therefore of great interest for understanding their perspective on philosophy as a whole. Presented as spoken by the Buddha himself, the verse reads: Just as wise persons accept gold, having first tested it through heating it, cutting it, and rubbing it on a touchstone, so too, O monks, should you accept my words only after testing them, and not out of respect for me.18 Kamalaśīla’s Summary of Prior Positions makes clear that the three tests referenced here are metaphors for three kinds of examinations that wise persons make in relation to three kinds of epistemic objects.19 First, there are objects that can be known directly ( pratyakṣa) through perception. This direct knowing is similar to testing gold by heating or melting it. Next are objects that can be reliably known only indirectly ( parokṣa) through inference. This indirect knowing is similar to testing gold by cutting it. Finally, there are objects that are epistemically remote (atyantaparokṣa) or unknowable for ordinary persons and which can therefore only be known through the testimony of an awakened authority like the Buddha. This extremely indirect kind of knowing is similar to testing gold by rubbing it on a touchstone. As Kamalaśīla makes 487

Sara McClintock

clear, following Dharmakīrti, Buddhists should rely primarily on perception ( pratyakṣa) and inference (anumāna), specifically inference that functions through the force of a real entity (vastubalapravṛttānumāna), to obtain valid knowledge. Only in cases where it is truly necessary for one’s spiritual progress to have knowledge of some supersensible matter should wise people rely on scriptural testimony through a mechanism known as scripturally based inference (āgamāśritānumāna), a kind of pseudo-inference whose justification is entirely cast in pragmatic terms.20 It is not hard to see why the current Dalai Lama is fond of citing this verse, given his self-proclaimed commitment to the methods of modern science.

Sūtra Commentaries An aspect of Kamalaśīla’s contribution to Indian Buddhist philosophy that is often overlooked (and where there remains considerable room for future research) can be found in his substantive Mahāyāna sūtra commentaries preserved in the Tibetan canon. The lack of attention is not so surprising, in part because these texts do not survive in Sanskrit but just as importantly because the philosophical contribution of sūtra literature is only recently gaining the attention it deserves. One difficulty in reading and interpreting the sūtras is their dialogical nature. Unlike other philosophical works in the Buddhist tradition, including the verse and prose treatises we have considered thus far, sūtras (and their Pāli counterparts, suttas) are presented not as works penned by an individual author but rather as the recordings of oral teachings given by the Buddha or his representative to some particular audience or addressee. As a genre, therefore, sūtras require a different set of hermeneutical tools to read and interpret than do the systematic treatises, or śāstras, that we more readily recognize as forms of philosophy. The Tibetan canon credits Kamalaśīla with five sūtra commentaries. These can be useful for refining our understanding of Kamalaśīla as a philosopher, although space allows only a brief hint of how this may be so. We begin with Kamalaśīla’s Commentary on the Rice Seedling Sūtra (Śālistambaṭīkā).21 This sūtra is an extended meditation on a pithy, enigmatic teaching – itself also termed a sūtra within the larger text – that the Buddha is said to have uttered upon spying a young rice seedling. The teaching goes: “Who sees dependent arising sees the Dharma. Who sees the Dharma sees the Buddha.” In the Rice Seedling Sūtra, we encounter this curious teaching through the report of the monk Śāriputra, who further claims that after uttering the sūtra, the Buddha remains silent. Śāriputra reports the teaching to the bodhisattva Maitreya, who subsequently gives a detailed explication, including many core elements of the technicalities of dependent arising. This explication makes up the bulk of the larger sūtra. But despite his thoroughness, Maitreya refrains from commenting on the Buddha’s silence, a fact that Richard Nance (2012, 126) helpfully points out. Kamalaśīla, on the other hand, does remark on it.22 As Nance unpacks this passage (126–128), we learn that Kamalaśīla understands the Buddha’s silence as fulfilling two distinct functions keyed to the Buddha’s two audiences within the sūtra: learned monks, like Śāriputra, and bodhisattvas, like Maitreya. For the monks, the silence is designed to destroy pride, since the monks will be forced to recognize that they do not understand this enigmatic teaching. For the bodhisattvas, the silence conveys the teaching’s profundity. Thus, in this short commentarial passage, we encounter in a different way from what we find in his epistemological treatises Kamalaśīla’s central idea that meaning is always determined at least in part by the audience to which is it directed. This sociolinguistic aspect of language runs through Kamalaśīla’s work as a philosopher, insofar as he understands all philosophy to be transactionally produced. The Commentary on the Dhāraṇī of Entering the Nonconceptual (Avikalpapraveśadhāraṇīṭīkā, known also as the Nirvikalpapraveśadhāraṇīṭīkā)23 is important in relation to another topic in 488

Kamalaśīla

Kamalaśīla’s philosophical thought: namely the issue of the relationship between conceptual (savikalpa) and nonconceptual (nirvikalpa) knowledge or awareness ( jñāna) and how a practitioner moves from the former to the latter. In the first of his three Stages of Cultivation, Kamalaśīla cites a passage from the Dhāraṇī for Entering the Nonconceptual that counsels “non-mentation” (amanasikāra) as a method for turning away from signs such as form and so on.24 But lest one imagine that this practice of non-mentation in any way invalidates or contradicts conceptual cognition and rational analysis, Kamalaśīla is quick to point out, in the words of David Seyfort Ruegg, “that what is here intended by the term amanasikāra is not simple absence of mentation (manasikārābhāvamātra) but, rather, that non-objectifying or non-apprehension which belongs to him who analyses through discriminative knowledge.”25 What does this mean? In his Commentary on the Dhāraṇī for Entering the Nonconceptual Kamalaśīla lays out two explanations for how to construe the scriptural quote.26 First, non-mentation may be understood as the defining characteristic (lakṣaṇa) of the proper examination of the real (bhūtapratyavekṣā). In this view, proper examination of the real is the necessary antidote to our primordial, unexamined, misguided, and habitual disposition to conceptualize signs in appearances. That conceptualization is mentation (manasikāra), and the proper examination of the real is its opposite, non-mentation or non-perception (anupalambha). Alternatively, non-mentation may be understood to be the result of proper examination of the real, which by way of metonymy is then called non-mentation. What really happens is that through the proper examination of the real, which still operates under the influence of primordial ignorance (avidyā), objectification of objects is removed and one enters the space of signlessness (one of three doorways to freedom, along with emptiness and wishlessness). Even though the proper examination of the real has a conceptual nature, it operates in such a way as to ultimately undermine and dismantle its objects of knowledge, much as the fire sticks used to start a fire are destroyed in the ensuing blaze – an image found in the Questions of Kāśyapa (Kāśyapaparivarta) and other scriptures and in various Madhyamaka treatises, including Kamalaśīla’s third of the three Stages of Cultivation.27 Kamalaśīla’s remaining three sūtra commentaries all fall in the Perfection of Wisdom ­section of the Tibetan canon. These are the Commentary on the Perfection of Wisdom in 700 Lines (Saptaśatikāprajñāpāramitāṭīkā), the Commentary on the Diamond Cutter Perfection of Wisdom (Prajñāpāramitāvajra­cchedikāṭīkā), and the Commentary on the Heart Sūtra Perfection of Wisdom (Prajñāpāramitāhṛdayanāmaṭīkā). Here, as in his other sūtra commentaries, Kamalaśīla employs the techniques recommended by his predecessor Vasubandhu in his Logic of Explication (Vyākhyāyukti).28 Scattered references to and treatments of these commentaries may be found in contemporary scholarship in English, but for the most part, these texts remain only partially studied and understood.29 M. David Eckel helpfully highlights that Kamalaśīla’s primary aim in his Heart Sūtra commentary was to emphasize the necessity of inference (anumāna) as the means to generate certainty (niścaya) about the ultimate truth – a position perhaps surprising given the Heart Sūtra’s repeated apophatic discourse but entirely in keeping with Kamalaśīla’s epistemological stance and commitment to gradualism.30

Tantra Commentaries and Other Writings As with other authors from this era, there exists a sizeable collection of works (nine or more) attributed to Kamalaśīla in the Tantra (rgyud) or Tantra Commentary (rgyud ’grel) sections of the Tibetan canon. The general current consensus now is that these tantric texts are unlikely to be the product of the same Kamalaśīla whose philosophical works we have been discussing, even though, as we have noted, Kamalaśīla does comment on one tantric verse in a text attributed to him in the Madhyamaka section of the Tibetan canon.31 Here again we come up against a difficult problem about authorship: we have an idea about who Kamalaśīla must be based on not only the knowledge 489

Sara McClintock

we receive from our own reading of the texts that we ascribe to him but also on our reception of hagiography, commentary, and contemporary scholarship. This image and the power it exerts over us is what Michel Foucault (2003) has called the “author function,” a set of conditions through which the idea we construct of an author serves to help explain the ideas we find in the texts we attribute to that same author. The author function prompts us to seek and find unity in the author’s oeuvre and also provides a key by which differences can be resolved. Similarly, the author function allows us to recognize a certain style or voice across genres of texts attributed to a given author. In the case of Kamalaśīla, the author function with which we currently operate colors how we imagine him and tends to make the ascription of tantric commentaries feel dubious. Yet archeological evidence at Nālandā, as well as Tibetan records, suggest that tantric practice was widespread among Indian monastics of that period. Given that tantra was then an esoteric practice, is it surprising that allusions to it are missing from non-tantric works? Of course, other means to determine authenticity can be applied,32 but sadly the state of our field does not allow us to say more at this point. The Epistles (spring yig) section of the Tibetan canon contains two works attributed to Kamalaśīla. The first, Lamp for the Generation of Faith (Śraddhotpādapradīpa), draws on prior Indian teachings about the preciousness of human life and the importance of developing the altruistic mind of awakening (bodhicitta). In this, it sets forth many of the ethical and religious themes that would later be developed in the Tibetan tradition known as Stages of the Path (lam rim). Although it is in the Epistles section, this text does not mention any particular addressee. In contrast, the second work in this section, the Teaching on the Varieties of Suffering (Duḥkhaviśeṣanirdeśa), does mention an addressee, who intriguingly appears likely to have been a woman.33 The fact that the colophon contains no mention of a translator further suggests that Kamalaśīla may have written the work directly in Tibetan and in response to the questions of a Tibetan disciple. Whatever the case, the work lays out in poetic detail eight forms of suffering – starting with the cycle of rebirth (saṃsāra) and moving through birth, old age, sickness-and-death, seeking the nonexistent, seeking what is unobtainable, meeting an enemy, and separation from friends.34 These are then equated with the five psychophysical aggregates (upadāna-skandha) that both constitute and fuel the cycle of rebirth itself. Kamalaśīla punctuates the verses describing the varieties of suffering – which he describes as a cesspool (’dam rdzab) – using iterative laments modeled after the first interjection: “There is no way to bear the suffering of saṃsāra here in this world!” Suffering, of course, is the first of the Buddha’s four noble truths. That there is no way to bear this suffering leaves only one option: to seek to obtain the freedom of nirvāṇa. Another text ascribed to Kamalaśīla is Remembering the Words of the Fifty Verses for Novices (lnga bcu pa’i tshig gi brjed byang, *Śramaṇapañcāśatkārikāpādābhismaraṇa), a commentary on Nāgārjuna’s training manual for novice monks in verse. Here we find Kamalaśīla in full monastic register, providing summaries of some of the most important narratives surrounding the root downfall ( pārājikā) rules for monks as drawn from the Mūlasarvāstivāda Vinaya lineage of monastic discipline. Although the text has not appeared in English translation, it continues to be read and studied among Tibetan monastics to this day. The lack of a recorded Sanskrit title suggests that the text may have been composed in Tibet, a possibility that makes sense given what must have been a need for pedagogical materials for the newly budding monastic community at Samyé during the period of Kamalaśīla’s reputed stay.

Doxographical Classification Many articles about Kamalaśīla begin with his doxographical affiliation as this has been assigned by later Tibetans and continued by contemporary scholars. Such an approach paints Kamalaśīla as a representative of a school and not as an individual, and this is why we have 490

Kamalaśīla

avoided starting with such classifications. Until now, we have spoken only in terms of school names that Kamalaśīla himself would have recognized: Madhyamaka, Yogācāra, Sautrāntika, and so on. We’ve also spoken about the Buddhist epistemological tradition, a stream of Indian Buddhist thinking that Kamalaśīla likely viewed not so much in terms of doctrinal distinctions but rather in terms of a commitment to particular interscholastic forms of rational inquiry. Indeed, in his Extensive Commentary on the Compendium of Realities, Kamalaśīla praises and identifies with the designation of Nyāyavādin, the “propounder of formal reasoning.”35 It is thus abundantly clear that Kamalaśīla’s commitment to reason and rational inquiry is a through line in nearly all of his writings, as we have seen repeatedly previously. Kamalaśīla’s commitment to the procedures of formal reasoning is also in large part behind what was eventually to become the most prevalent doxographic label attached to his name, that of the so-called Yogācāra-Svātantrika-Mādhyamika (rnal ’byor spyod pa’i dbu ma rang rgyud pa). This label has its own complicated history, combining as it does two separate doxographical divisions devised by Tibetans.36 The first can likely be credited to the eighth-century Tibetan scholar Yeshé Dé (Ye shes sde), who spoke of two kinds of Mādhyamikas: the SautrāntikaMādhyamika (mdo sde pa’i dbu ma pa), who accepts the existence of objects of knowledge external to the mind, and the Yogācāra-Mādhyamika (rnal ’byor spyod pa’i dbu ma pa), who does not. Both Śāntarakṣita and Kamalaśīla fall into the later camp, while Bhāviveka is seen to represent the former. A second doxographical division was promulgated later, from the eleventh century onward, originating perhaps with the translator Patshab Nyimadrak (Pa tshab nyi ma grags, b. 1055), namely the division between the so-called Svātantrika-Mādhyamika (rang rgyud pa’i dbu ma pa) and the Prāsaṅgika-Mādhyamika (thal ’gyur ba’i dbu ma pa). At base, this division depends on whether the Mādhyamika in question accepts the use of so-called autonomous proof statements (svatantra-prayoga = rang rgyud kyi sbyor ba). On this view, Śāntarakṣita and Kamalaśīla, as enthusiastic adopters of Dignāga and Dharmakīrti’s epistemological apparatus with its fulsome reliance on proof statements, must be classified as Svātantrika-Mādhyamikas. Others, such as the seventh-century Mādhyamika Candrakīrti, reject the use of the technical tools of Buddhist epistemology as inappropriate for a Mādhyamika. Such thinkers are then classified by later Tibetans as Prāsaṅgika-Mādhyamikas in a gesture toward their preference for arguments relying on the urging of unwanted consequences ( prasaṅga = thal ’gyur) on their opponents. A detailed examination of these later Tibetan doxographical schemas is beyond the scope of this chapter. Yet because these doxographical systems have exerted a strong influence on contemporary scholars, it is necessary to have a basic understanding of the issues at hand. Key in all of this is the problem of natures (svabhāva), since the inferential reasoning of the Buddhist epistemologists depends on identifying a natural relation (svabhāva-pratibandha) between the elements in an inference (i.e., between the evidence and that which it proves). Those Tibetans who maintain, as nearly all Tibetans eventually do, that the so-called Prāsaṅgika-Mādhyamika is superior to the so-called Svātantrika-Mādhyamika do so on the basis that the Svātantrika puts inappropriate stock in natures, a move held to be antithetical to Madhyamaka thought. Also key is that such natures would operate “autonomously” (rang rgyud kyis) or “by their own nature” (rang bzhin gyis). But such an idea is far less likely to have been acceptable to Kamalaśīla. As a Mādhyamika, Kamalaśīla does not accept that natures are established autonomously or intrinsically. His commitment to natures is pragmatic and operates fully on the conventional or transactional level – which, we must remember, includes even nonconceptual objects of awareness such as the objects known in direct perception (see McClintock 2019). The distinction, therefore, is dubious, though not without philosophical interest. Recently, Kevin Vose (2020) has argued that the development of the Svātantrika-Prāsaṅgika distinction should indeed be traced to Patshab Nyimadrak, specifically to his close reading 491

Sara McClintock

of Śāntarakṣita and Kamalaśīla’s epistemological writings. Vose explains how for Patshab, Śāntarakṣita and Kamalaśīla’s reliance on autonomously established natures qua appearances “makes ultimate conventions in the world” (707). In the conclusion of his article, Vose concedes that the development of the distinction “may be unfair” and involves “several interpretive leaps” that rely on the “stretching of terminology” (745). The upshot, then, is that the label of Svātantrika-Mādhyamika is probably best avoided when discussing Śāntarakṣita and Kamalaśīla except insofar as it offers an excellent opportunity to further investigate and refine our understanding of how these thinkers square their commitment to the structures of reason with their equally vehement commitment to a Madhyamaka vision of naturelessness. How to resolve these two elements of Kamalaśīla’s thought, the rational side which relies on natures to undertake a proper examination of the real and the anti-foundationalist side which deconstructs natures at every turn? Gradualism is one part of the answer, as we have seen. Kamalaśīla asks us to utilize the natures that appear to us in order to examine them and discover their ultimate unreality. In doing so, we are gradually weaned from grasping at objects external to the mind (at the Sautrāntika level of analysis), until eventually we are weaned as well from grasping the mind itself (at the Yogācāra level of analysis). Finally, the philosopher is freed when the Madhyamaka nonconceptual vision of naturelessness dawns. But arriving there takes some work. For Kamalaśīla, it is necessary to engage with natures, despite their lack of ultimate reality, in order to obtain freedom from seeing them as ultimately real. Walking away from natures without doing the work is not an option. Likewise, it is equally necessary to engage with natures if we are to be of benefit to others. Again, this is true even though natures may be ultimately unreal. In Light of the Middle,37 Kamalaśīla states that “even though an illusion is false, it is relied upon when there is a purpose, as in the case of magicians.” Comparing buddhas and bodhisattvas to magicians, Kamalaśīla insists that these compassionate ones “do not turn away from anything whatsoever that is of use to others.” Natures, which form the basis for investigation, are ultimately unreal. Yet like other unreal appearances, they arise in a way that is dependent on causes and conditions. Thus, “just as illusions and so forth arise in dependence on the collection of their own causes, likewise the ‘illusion’ which is the wisdom of the yogis and so forth, which is acceptable as long as it is not analyzed, also arises.” Even the pinnacle of philosophical knowledge, wisdom itself, is an illusion. But it is a kind of necessary illusion, especially for those motivated by compassion. Reason is thus a form of magic for this Mādhyamika champion of reason. Although an illusion, it nevertheless serves a great purpose – namely to convey sentient beings to freedom from the grips of illusion.

Abbreviations BhK MAV MAP MĀ NBPS TS TSP

Bhāvanākrama, Stages of Cultivation Madhyamakālaṃkāravṛtti, Ornament of the Middle Madhyamakālaṃkārapañjikā, Extensive Commentary on Ornament of the Middle Madhyamakāloka, Light of the Middle Nyāyabindupūrvapakṣasaṃkṣipta, Summary of Prior Positions in the Drop of Reason Tattvasaṃgraha, Compendium of Realities Tattvasaṃgrahapañjikā, Extensive Commentary on Compendium of Realities

Notes 1 Chapa’s commentary, entitled Extensive Explanation Commentary on Light of the Middle (dBu ma snang ba’i ’grel pa rgya cher bshad pa or sNang bshad), is published in vol. 6 of the Kadam

492

Kamalaśīla Collected Works (bKa’ gdams gsung ’bum) and is currently being edited and translated by Pascale Hugon and Jongbok Yi. It is also discussed briefly in Hugon 2015. As for the partial commentary by Ngawang Tendar, entitled Notes on Light of the Middle (dBu ma snang ba’i brjed tho), although it covers only the prior positions ( pūrvapakṣa), it is nonetheless quite useful. See also Ye, Li, and Kano 2013, 37–38 for discussion of a partial Sanskrit manuscript of an apparent commentary on the MĀ. 2 Seyfort Ruegg 1981, 96–99 summarizes the contents of all three texts. Arnold 2017 provides a philosophical summary. 3 Kellner 2020 nicely summarizes these interlocking themes in the BhK texts. 4 Funayama 2011 discusses Kamalaśīla’s theory of yogic perception. 5 Keira 2004, 10–13 n. 32 locates the five arguments in the MĀ and discusses other lists of proofs in Indian texts as well as the reception of Kamalaśīla’s list in Tibet. 6 Ichigō 1992, 199–202 reviews various identifications made especially by Japanese scholars while also providing a synopsis of the prior positions according to the commentary of Ngawang Tendar. Keira 2004 includes English translations of the subsequent positions answering to prior positions 4–14 and 33–42 according to Ngawang Tendar’s numbering. 7 Keira 2009 translates relevant passages from MĀ. 8 See Keira 2009, 14–15 for translation and analysis of the larger passage. 9 See Eckel 1987b for a translation of Jñānagarbha’s Distinction Between the Two Truths (Satyadvayavibhaṅgavrtti) with excerpts from Śāntarakṣita’s commentary in the endnotes. 10 Note that Krishnamacharya mistakenly believed a verse to have been elided from the manuscript, so his verse numbers are off by one digit starting at verse 527. We follow Shastri’s numbering throughout. 11 On the topic of the sliding scale of analysis, see my treatment in McClintock 2010, 85–91, in which I also reference the discussion in Dreyfus 1997, 83–105; Dunne 2004, 53–79. Kellner 2011 critiques the idea. 12 See TSP ad TS 3434. 13 Dharmakīrti invokes the term pāramārthika-pramāṇa at Pramāṇaviniścaya I.58. For a recent in-depth exploration of Dharmakīrti’s theory of perception, including reflexive awareness, see Yiannopoulos 2020. 14 See TSP ad TS 3435–3436. See also TS 2000ff. with Kamalaśīla’s commentary. 15 See TSP ad TS 2980–2981. 16 See, for example, Kamalaśīla’s commentary on Śāntarakṣita’s MAV 16 (Ichigō 1985, 71), where he indicates that the characterization of reflexive awareness as independently luminous (gsal ba gzhan la mi bltos pa) is performed according to “the conventional path” (tha snyad kyi lam). For Kamalaśīla’s understanding of Yogācāra philosophy as both crucial and provisional on the path of meditation, see Arnold 2017. 17 The idea is that the Yogācāra philosophical stream is not to be read as arguing for a strict idealism in which nothing but mind is real; rather, it is best understood as providing tools for the analysis of phenomenal experience. These tools reveal that phenomenologically speaking – that is, in terms of how we experience the world – there can be no ultimately real division between subject and object. Instead, the mind, whose nature is to be illuminating, is aware of itself in an endless variety of forms. But this observation concerns our experience alone and says nothing about what does and does not exist in the world. Such ontological questions are best left to the Madhyamaka philosophical stream, which recognizes all things, whether mental or physical, to be empty of essential nature. 18 TS 3587. Kamalaśīla also cites this verse in his commentary on the opening verses of the TS and includes it as well in his NBPS 1.2.1. See also TSP ad TS 3343. 19 See Eltschinger 2013, 99–101; McClintock 2010, 61. 20 See Keira 2009; McClintock 2013 for Kamalaśīla’s understanding of scripturally based inference. For Dharmakīrti’s position, see Tillemans 1999. 21 The sūtra and Kamalaśīla’s commentary have been critically edited, studied, and translated in Schoening 1995. 22 Nance notes that the two audiences may overlap in some instances. 23 Kamalaśīla’s commentary has been partially translated in Robertson 2006. 24 BhK I, Tucci 1958, 212. For a study of amanasikāra, including its reception in Tibet, see Higgins 2006. 25 Seyfort Ruegg 1989, 94. See Tucci 1958, 212. 26 The explanation that follows relies on Seyfort Ruegg 1989, n. 179. 27 Tucci 1971, 20. See also Kellner 2020, 52. 28 See Schoening 1995, 138–43. See also Nance 2012, 100–22 and 129–52.

493

Sara McClintock 29 Tucci 1956, 129–71 contains a summary of the Commentary on the Diamond Cutter Perfection of Wisdom. 30 Eckel 1987a, 76. 31 Eltschinger and Marks (2019, 274) speculate these works may have been authored by the tantric master Padampa Sangyé (Pha dam pa Sangs rgyas, d. 1117). 32 Steinkellner (1999, 355–60) exemplifies the kind of required close reading when he analyzes the tantric text Establishing Reality (Tattvasiddhi) attributed to Kamalaśīla’s teacher Śāntarakṣita. Steinkellner concludes that while the text does show evidence of training in the Buddhist epistemological tradition, the author “commits at least one major logical blunder,” thus effectively nullifying the possibility of the text being authored by the Śāntarakṣita we know. Steinkellner does not discuss, however, the possibility of scribal error or other factors that could explain this blunder. While there may be reasons to doubt the attributions of authorship of tantric texts to famous Indian philosophers, there are reasons to keep this question open for future research. 33 Dietz 1984, 75–76 provides a summary of the work and discusses the identity of the addressee, whose name is given as Lho Zamo Tshangpé Yang (Lho za mo tshangs pa’i dbyang) and variants thereof. Dietz 1984, 340–57 also includes an edition of the Tibetan text and an annotated German translation. 34 This list of eight forms of suffering differs from lists found in other texts. Dietz 1984, 343 n. 3 notes the unusual way Kamalaśīla considers death to fall within the category of sickness. 35 See McClintock 2019 for details. 36 See Mimaki 1983 for details. 37 Citations are from the Derge edition, vol. sa, D 222b–223b.

Bibliography Adam, Martin T. 2002. “Meditation and the Concept of Insight in Kamalaśīla’s Bhāvanākramas.” PhD diss., McGill University. Arnold, Dan. 2017. “Pushing Idealism Beyond Its Limits: The Place of Philosophy in Kamalaśīla’s Steps of Cultivation.” In The Oxford Handbook of Indian Philosophy, edited by Jonardon Ganeri, 379–99. Oxford: Oxford University Press. The Dalai Lama. 2001. Stages of Meditation. Root text by Kamalashila. Translated by Venerable Geshe Lobsang Jordhen, Losang Choephel Ganchenpa, and Jeremy Russell. Ithaca, NY: Snow Lion Publications. The Dalai Lama, Khöntön Peljor Lhündrup, and José Ignacio Cabezón. 2011. Meditation on the Nature of Mind. Somerville, MA: Wisdom Publications. Demiéville, Paul. 1952. Le Concile de Lhasa. Une controverse sur le quiétisme entre bouddhistes de l’Inde et de la Chine au VIIIe siècle de l’ère chrétienne. Paris: Presses Universitaires de France. Reprint Paris: Collège de France, Institut des Hautes Études Chinoises, 2006. Dietz, Siglinde. 1984. Die Buddhistische Briefliteratur Indiens: Nach dem tibetischen Tanjur herausgegeben, übersetzt und erläutert. Wiesbaden: Otto Harrassowitz. Dreyfus, Georges J. B. 1997. Recognizing Reality: Dharmakīrti’s Philosophy and Its Tibetan Interpretations. Albany: State University of New York Press. Dreyfus, Georges J. B., and Sara L. McClintock, eds. 2003. The Svātantrika-Prāsaṅgika Distinction: What Difference Does a Difference Make? Boston: Wisdom Publications. Dunne, John D. 2004. Foundations of Dharmakīrti’s Philosophy. Somerville, MA: Wisdom Publications. Eckel, Malcolm David. 1987a. “Indian Commentaries on the Heart Sūtra: The Politics of Interpretation.” Journal of the International Association of Buddhist Studies 10 (2): 69–79. ———. 1987b. Jñānagarbha’s Commentary on the Distinction Between the Two Truths: An EighthCentury Handbook of Madhyamaka Philosophy. Albany: State University of New York Press. Eltschinger, Vincent. 2013. “Turning Hermeneutics into Apologetics: Reasoning and Rationality under Changing Historical Circumstances.” In Scriptural Authority, Reason and Action. Proceedings of a Panel at the 14th World Sanskrit Conference, Kyoto, September 1st – 5th 2009, edited by Vincent Eltschinger and Helmut Krasser, 71–145. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. ———. 2014. Buddhist Epistemology as Apologetics: Sources on the History, Self-Understanding and Dogmatic Foundations of Late Indian Buddhist Philosophy. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Eltschinger, Vincent, and James Marks. 2019. “Kamalaśīla.” In Brill’s Encyclopedia of Buddhism. Vol II: Lives, edited by Jonathan A. Silk et al., 273–78. Leiden and Boston: Brill. Foucault, Michel. 2003. “What Is an Author?” In The Essential Foucault: Selections from the Essential Works of Foucault 1954–1984, edited by Paul Rabinow and Nikolas Rose, 377–91. New York: The New Press.

494

Kamalaśīla Funayama, Toru. 2011. “Kamalaśīla’s View on Yogic Perception and the Bodhisattva Path.” In Religion and Logic in Buddhist Philosophical Analysis. Proceedings of the Fourth International Dharmakīrti Conference. Vienna, August  23–27, 2005, edited by Helmut Krasser, Horst Lasic, Eli Franco, and Birgit Kellner, 99–111. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Garfield, Jay L. 2016. “Cittamātra as Conventional Truth from Śāntarakṣita to Mipham.” Journal of Buddhist Philosophy 2: 263–80. Higgins, David. 2006. “On the Development of the Non-Mentation (amanasikāra) Doctrine in IndoTibetan Buddhism.” Journal of the International Association of Buddhist Studies 29 (2): 255–303. Hugon, Pascale. 2015. “Proving Emptiness: The Epistemological Background for the ‘Neither One Nor Many’ Argument and the Nature of Its Probandum in Phya pa Chos kyis senge ge’s Works.” Journal of Buddhist Philosophy 1: 58–94. Ichigō, Masamichi. 1985. Madhyamakālaṃkāra of Śāntarakṣita with His Own Commentary or Vṛtti and with the Subcommentary or Pañjikā of Kamalaśīla. Kyoto: Buneido. ———. 1992. “On the Dbu ma Snaṅ ba’i Brjed Tho.” Asiatische Studien/Études Asiatiques 46: 195–211. Jha, Ganganatha, trans. 1937/1939. The Tattvasaṅgraha of Śāntarakṣita with the Commentary of Kamalaśīla. Baroda: Oriental Institute. Kajiyama, Yuichi. 1978. “Later Mādhyamikas on Epistemology and Meditation.” In Mahāyāna Buddhist Meditation: Theory and Practice, edited by Minoru Kiyota, 114–43. Honolulu: University of Hawai’i Press. Keira, Ryusei. 2004. Mādhyamika and Epistemology: A Study of Kamalaśīla’s Method for Proving the Voidness of All Dharmas. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. ———. 2009. “The Description of Niḥsvabhāvatā and its Intentional Meaning: Kamalaśīla’s Solution for the Doctrinal Conflict Between Mādhyamika and Yogācāra.” Acta Tibetica et Buddhica 2: 1–24. Kellner, Birgit. 2011. “Dharmakīrti’s Criticism of External Realism and the Sliding Scale of Analysis.” In Religion and Logic in Buddhist Philosophical Analysis. Proceedings of the Fourth International Dharmakīrti Conference. Vienna, August  23–27, 2005, edited by Helmut Krasser, Horst Lasic, Eli Franco, and Birgit Kellner, 291–98. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. ———. 2020. “Using Concepts to Eliminate Conceptualization: Kamalaśīla on Non-Conceptual Gnosis (nirvikalpajñāna).” Journal of the International Association of Buddhist Studies 43: 39–80. Krishnamacharya, Embar, trans. 1926. Tattvasaṅgraha of Śāntarakṣita: With the Commentary of Kamalaśīla. 2 vols. Gaekwad’s Oriental Series, no. 30–31. Baroda: Oriental Institute. Marks, James Michael. 2019. “Playfighting: Encountering Aviddhakarṇa and Bhāvivikta in Śāntarakṣita’s Tattvasaṃgraha and Kamalaśīla’s Pañjikā.” PhD diss., University of California, Berkeley. Meinert, Carmen. 2003. “Structural Analysis of the Bsam gtan mig sgron: A Comparison of the Fourfold Correct Practice in the Āryāvikalpapraveśanāmadhāraṇī and the Contents of the Four Main Chapters of the Bsam gtan mig sgron.” Journal of the International Association of Buddhist Studies 26 (1): 175–95. McClintock, Sara. 2010. Omniscience and the Rhetoric of Reason: Śāntarakṣita and Kamalaśīla on Rationality, Argumentation, and Religious Authority. Boston: Wisdom Publications. ———. 2013. “Kamalaśīla and Śāntarakṣita on Scripture and Reason: The Limits and Extent of ‘Practical Rationality’ in the Tattvasaṃgraha and Pañjikā.” In Scriptural Authority, Reason and Action. Proceedings of a Panel at the 14th World Sanskrit Conference, Kyoto, September 1st – 5th 2009, edited by Vincent Eltschinger and Helmut Krasser, 209–37. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. ———. 2019. “How to Do Things with Natures: A Madhyamaka Approach to Arguments and Appearances.” Journal of the International Association of Buddhist Studies 42: 405–447. Mimaki, Katsumi. 1983. “The Blo gsal grub mtha’ and the Mādhyamkika Classification in Tibetan Grub mtha’ Literature.” In Contributions on Tibetan and Buddhist Religion and Philosophy, edited by Ernst Steinkellner and Helmut Tauscher, 161–67. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. Moriyama, Seitetsu. 1985. “An Annotated Translation of Kamalaśīla’s Sarvadharmaniḥsvabhāvasuddhi Part IV.” Bukkyo Daigaku Daigakuin kenkyū kiyō 69: 36–86. Nance, Richard F. 2012. Speaking for Buddhas: Scriptural Commentary in Indian Buddhism. New York: Columbia University Press.

495

Sara McClintock Robertson, Raymond E. 2006. A Study of the “Dharmadharmatāvibhaṅga” (Volume 1): The Root Text and its Scriptural Source (the Avikalpapraveśadhāraṇī), with Excerpts from Kamalaśīla’s Avikalpapraveśadhāraṇīṭīkā. Beijing: China-Tibet Publishing House. Schmid, Karl. 2020. “Kamalaśīla’s Theory of Vipaśyanā: An Exposition and Philosophical Defense.” PhD diss., Emory University. Schoening, Jeffrey D. 1995. The Śālistamba Sūtra and its Indian Commentaries. 2 vols. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. Seyfort Ruegg, David. 1981. The Literature of the Madhyamaka School of Philosophy in India. Wiesbaden: Otto Harrassowitz. ———. 1989. Buddha-Nature, Mind and the Problem of Gradualism in a Comparative Perspective: On the Transmission and Reception of Buddhism in India and Tibet. London: School of Oriental and African Studies. ———. 2013. “On the Tibetan Historiography and Doxography of the ‘Great Debate of Samyé’.” In The Tibetan History Reader, edited by Gray Tuttle and Kurtis R. Schaeffer, 108–20. New York: Columbia University Press. Shastri, Dwarika Das. 1968. Tattvasaṅgraha of Ācārya Shātarakṣita: With the Commentary ‘Pañjikā’ of Shrī Kamalashīla, 2 vols. Varanasi: Bauddha Bharati. Steinkellner, Ernst. 1999. “Yogic Cognition, Tantric Goal, and Other Methodological Applications of Dharmakīrti’s Kāryānumāna Theorem.” In Dharmakīrti’s Thought and its Impact on Indian and Tibetan Philosophy: Proceedings of the Third International Dharmakīrti Conference, Hiroshima, November 4–6, 1997, edited by Shōryū Katsura, 349–62. Wien: Verlag der Österreichischen Akademie der Wissenchaften. Taniguchi, Fujio. 1992. “Quotations from the First Bhāvanākrama of Kamalaśīla Found in Some Indian Texts.” In Tibetan Studies: Proceedings of the 5th Seminar of the International Association for Tibetan Studies, Narita 1989, edited by Ihara Shōren and Yamaguchi Zuihō, 303–7. Tokyo: Naritasan Shinshoji. Tauscher, Helmut. 1999. “Phya pa chos kyi seng ge’s Opinion on prasaṅga in his dBu ma’i shar gsum gyi stong thun.” In Dharmakīrti’s Thought and Its Impact on Indian and Tibetan Philosophy: Proceedings of the Third International Dharmakīrti Conference, Hiroshima, November 4–6, 1997, edited by Shōryū Katsura, 383–93. Wien: Verlag der Österreichischen Akademie der Wissenchaften. Tillemans, Tom J. F. 1999. “How Much of a Proof Is Scripturally Based Inference (āgamāśritānumāna)?” In Dharmakīrti’s Thought and Its Impact on Indian and Tibetan Philosophy: Proceedings of the Third International Dharmakīrti Conference, Hiroshima, November 4–6, 1997, edited by Shōryū Katsura, 395–404. Wien: Verlag der Österreichischen Akademie der Wissenchaften. Tucci, Giuseppe. 1956. Minor Buddhist Texts. Part I. Rome: Istituto Italiano per il Medio ed Estremo Oriente. ———. 1958. Minor Buddhist Texts. Part II. First Bhāvanākrama of Kamalaśīla: Sanskrit and Tibetan Texts with Introduction and English Summary. Rome: Istituto Italiano per il Medio ed Estremo Oriente. ———. 1971. Minor Buddhist Texts. Part III. Third Bhāvanākrama. Rome: Istituto Italiano per il Medio ed Estremo Oriente. Ueyama, Daishun, Kenneth W. Eastman, and Jeffrey L. Broughton. 1983. “The Avikalapa-praveśadhāraṇī: The Dharani of Entering Non-Discrimination.” Bulletin of Institute of Buddhist Cultural Studies 22: 32–42. van der Kuijp, Leonard. 1983. Contributions to the Development of Tibetan Buddhist Epistemology: From the Eleventh to the Thirteenth Century. Wiesbaden: Franz Steiner Verlag. Vose, Kevin. 2020. “When Did Svatantra Inference Gain Its Autonomy? Śāntarakṣita and Kamalaśīla as Sources for a Tibetan Distinction.” Journal of Indian Philosophy 48: 703–50. Wangdu, Pasang, and Hildegard Diemberger, trans. and ed. 2000. dBa’ bzhed: The Royal Narrative Concerning the Bringing of the Buddha’s Doctrine to Tibet. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Ye, Shaoyong, Xuezhu Li, and Kazuo Kano. 2013. “Further Folios from the Set of Miscellaneous Texts in Śāradā Palm-leaves from Zha lu Ri phug.” China Tibetology 1: 30–48. Yiannopoulos, Alexander. 2020. “The Structure of Dharmakīrti’s Philosophy: A Study of Object-Cognition in the Perception Chapter ( pratyakṣapariccheda) of the Pramāṇasamuccaya, the Pramāṇavārttika, and Their Earliest Commentaries.” PhD diss., Emory University.

496

30 HARIBHADRA The Voice of Perfect Wisdom Pierre-Julien Harter

Introduction Haribhadra (c. eighth century CE) is renowned as one of the foremost specialists in the lite­ rature of Perfect Wisdom ( prajñā-pāramitā) in late Indian Buddhism. In Haribhadra, we meet two of the most impressive qualities of Buddhist philosophers: exegetical expertise and speculative ingenuity. As an exegete, he stays faithful to the ideal of the Buddhist intellectual who is able both to master the understanding of authoritative texts (āgama) and to defend and establish their claims through reasoning ( yukti). At the same time, although what is left of him consists essentially in commentarial work, his writings showcase a creativity that elevates Haribhadra as a standard for the Perfect Wisdom tradition. Like so many other groundbreaking philosophers, his contributions forced subsequent authors to take a position either for or against him. He was not to be ignored by either late Indian or Tibetan authors, especially as with the passage of time, his authority grew to become almost undisputed in Tibet. Haribhadra’s philosophical contribution might best be summarized as an attempt to provide the metaphysical and ethical foundations of the Buddhist path (mārga). The exegetical tradition with which he is most closely associated, that of the Ornament of Realizations (Abhisamayālaṃkāra, henceforth Ornament), is dedicated to unveiling the “vast” dimension of the sūtras of Perfect Wisdom ( prajñā-pāramitā-sūtras).1 This is said to be in contrast to the “profound” dimension on which well-known Mādhyamika philosophers like Nāgārjuna, Āryadeva, and Bhāviveka famously focus. While the “profound” dimension concerns the metaphysical conception of the nature of things qualified as empty (śūnya) or insubstantial (niḥsvabhāva), the “vast” dimension is about the path or process of transformation that leads an individual from their ordinary human condition to the perfect state of buddhahood. Haribhadra can thus be identified as a “philosopher of the path,” with the proviso that this implies much more than just ethical or pragmatic concerns, since the path is an all-encompassing concept that involves in addition metaphysical, cognitive, emotional, physical, and affective aspects.

Biographical and Bibliographical Context Given the obscurity surrounding many figures of classical Indian philosophy, we are lucky to have some rather specific information regarding Haribhadra’s life.2 Mano (1972) confirms the DOI: 10.4324/9781351030908-43

497

Pierre-Julien Harter

scholarly consensus by estimating his dates as ca. 735–795. The historical context surrounding Haribhadra’s life is marked by the flourishing period of the Pāla dynasty, especially under the kingship of Dharmapāla (not to be confused with the scholar of the same name), which re­presents a golden age in terms of military expansion and economic prosperity. Such favorable conditions allowed for strong patronage of Buddhist monasteries and communities. The great academic centers of Vikramaśīla, Odantapurī, and Somapur, which attracted so many of the famous later Buddhist philosophers featured in this volume, were established during the eighth to ninth centuries with the support of Pāla rulers. Haribhadra’s intellectual profile can be inferred from a few clues (Harter 2019a). As the colophon of his long and complex commentary indicates, he wrote it at the monastery of Trikaṭuka, considered to be situated in Bengal (Sanderson 2009, 88–89). This suggests that he was probably an “academic monk,” and one can imagine that he was renowned for his mastery of the literature of Perfect Wisdom given the massive commentary he composed and, given the fact that his commentaries often refer to the opinions of unnamed interpreters, that he evolved in a learned and scholastic environment, debating and disagreeing with other scholars. It is unclear how close he was to the intense scholastic activity of the big university-monasteries mentioned previously. He is not remembered to have been part of the “teaching staff ” there, but his works cite such well-known scholars as Śubhagupta, Jñānagarbha, and Śāntarakṣita, which indicates that he was well aware of the debates of the time and even took a position on some of the issues discussed then. The tradition does not remember him as anything other than a specialist of Perfect Wisdom literature, which sets him apart from other late Indian Buddhist philosophers who composed treatises in different branches of Buddhist philosophy. As argued in the following, the reason might be that he saw Perfect Wisdom literature as encompassing all aspects of Buddhist philosophy and doctrine, which is why he did not feel compelled to write independent treatises on epistemology ( pramāṇa) or Abhidharma. Perfect Wisdom literature can be said to present two forms during his time: a sūtra form and a śāstra form. A sūtra is considered by the tradition to be a teaching uttered or “authorized” by the Buddha. Many sūtras of Perfect Wisdom were circulating at his time of varying lengths: in 25,000 verses, in 18,000 verses, in 8,000 verses, or even in one syllable (Conze 1973, 1975, 1978). A śāstra is a treatise providing a theory to be implemented practically (Pollock 1985), ranging from elephant training to medicine and philosophy. The Ornament is one such treatise. Dedicated to the topic of Perfect Wisdom, it emerged around the fourth century CE. Haribhadra is the first known commentator to ascribe its authorship to Maitreya. The treatise established the possibility for a theory of the path based on the sūtras of Perfect Wisdom by providing an interpretation of these sūtras that supposedly uncovers their intention to present the whole Buddhist path with its stages, its prerequisites, the virtues and qualities necessary to its cultivation and the faults that slow it down, and the metaphysics that lays at its foundation (Harter 2015, 4–10). The Ornament is not a manual of the path that lists practical advice about how to practice the Buddhist path. Instead, it takes the form of an abstract treatise that reflects on what constitutes the path and what does not and on the theoretical principles the path requires. This is why it is organized into eight chapters, which do not mirror the stages of the path in a chronological order: it begins at the end, so to speak, with a chapter on the omniscience of the buddha, before backtracking to the omniscience of lower forms of realization (the bodhisattva and the hearer). Haribhadra’s bibliography displays his expertise in both sūtra and śāstra literature. He remains famous for being the editor of the Sūtra on Perfect Wisdom in Twenty-Five Thousand Verses (Pañcaviṃśatisāhasrikā, D 3790, P 5188), which was later transmitted only in Tibet (Conze 1978, 34–39; Makransky 1997, 132). He also wrote three commentaries. The least 498

Haribhadra

well known of his commentaries concerns another important sūtra from the Perfect Wisdom literature, the Versified Summary of the Precious Qualities (Ratnaguṇasaṃcayagāthā), a sort of précis of the texts and ideas of the Perfect Wisdom literature (Yuyama 1976). He also wrote two commentaries on the Ornament: his short Explanation on the Treatise ‘The Versified Ornament of Realizations’ (Abhisamayālaṃkārakārikāśāstravivṛti, Amano 2000; Naughton 1989; hereafter Explanation) and his much longer commentary entitled The Light Illuminating the Ornament of Realizations: A Commentary on Perfect Wisdom (Abhisamayālaṃkārālokā prajñāpāramitāvyakhyā, Wogihara 1932–1935; Sparham 2006–2012; hereafter Light). All three of these works focus on the same subject, Perfect Wisdom, albeit with different emphases. The commentary on the Versified Summary of the Precious Qualities focuses on this sūtra while blending in references to the structure of the Ornament in eight chapters and seventy topics and reading the sūtra through that lens. The Light balances its focus on both the Sūtra on Perfect Wisdom in Eight-Thousand Verses and the Ornament. It refers to both texts constantly, offering a gloss on every word. Finally, the short Explanation focuses exclusively on the Ornament (except for a few quotations from sūtras devoid of commentary), becoming the first autonomous commentary on that text. With this commentary, Haribhadra elevated the Ornament to a new status, that of a theoretical investigation into the concept of the path. These three commentaries thus seem to make a statement that the meaning of Perfect Wisdom is not enclosed in one text or even one form of text (sūtra or treatise) but rather resonates in different corners of Buddhist literature, even when it is not recognized.

Haribhadra and the Voice of Perfect Wisdom Haribhadra’s works suggest that his main philosophical concern can be identified with an attempt to recover the voice of Perfect Wisdom. Speaking of a voice is not only a poetic metaphor: Haribhadra explains that a sūtra is a statement (vacana) and needs to be analyzed as such (Wogihara 1932–1935, 2). It also implies that the voice of Perfect Wisdom be articulated with Haribhadra’s own voice, which raises an important issue: is there a space for Haribhadra’s philosophical individuality? To address this issue, we should not dodge the philological question regarding the authorship of the three commentaries ascribed to Haribhadra. To this question, the answer can be unequivocal: there is absolutely no reason to doubt the attributions of all three works to the same author. The two commentaries on the Ornament are very close in terms of content and style, and entire passages are reproduced from one to the other. As for the Commentary on the Versified Summary of the Precious Qualities, it showcases the same features as the two other commentaries: blended references to the sūtras of Perfect Wisdom and the Ornament, usage of “excursus” to expand on chosen subjects, and the exegetical reflex of adding “in reality” to explain away the seeming nihilism that texts from the Perfect Wisdom literature can be accused of falling into.3 The philological question thus leads us to an interesting literary remark: there is a similar tone that resonates in all of Haribhadra’s commentaries, which echoes the voice of Perfect Wisdom by defending it from nihilist interpretations. Haribhadra does so by reevaluating the status of relative truth (saṃvr․ ti-satya) and expanding the significance of Perfect Wisdom beyond a mere defense of emptiness by providing philosophical essays on various subjects. Removing doubts regarding the authorship of all three works to Haribhadra certainly helps sketching Haribhadra’s philosophical profile, in contrast, for instance, to the uncertainty surrounding Nāgārjuna’s authorship of so many texts. It nonetheless would be simplistic to think that the issue is resolved. Indeed, Haribhadra is and presents himself merely as a commentator. He never wrote independent works. His voice is thus constantly intertwined with the voice he 499

Pierre-Julien Harter

is trying to serve. Does a commentator have an individual voice? This question is not unique to the Indian tradition. The vast majority of philosophical traditions are exegetical traditions, at least until the nineteenth or twentieth century (although, arguably, exegesis is still largely practiced by today’s philosophers, if in a subtler form). Philosophers have long thought that the purpose of philosophical work is not to display originality in the way that, for instance, a poet might want to. Truth and clarity are supposed to trump any other preoccupation. Hegel himself stated that the task of the philosopher was nothing else than to be “the secretary of the universal Spirit,” which tells us something important about such philosophical perspective: what matters is less originality than authenticity as the most faithful way to serve truth, or the Spirit, or Perfect Wisdom, and it is the particular way with which a philosopher renders that service that marks their individuality.4 What is, then, this Perfect Wisdom of which Haribhadra could be said to be the secretary? Haribhadra recognizes the polysemy of the term and articulates its different meanings to give a comprehensive and rational account of Perfect Wisdom. He writes: Perfect Wisdom is the perfection of wisdom, which is defined as the discernment of phenomena. The principal Perfect Wisdom is the blessed buddha, the nondual know­ ledge that is similar to illusion. But, in conformity with its attainment, the secondary Perfect Wisdom is the text, which is a collection of words and sentences, and the path, defined by the different stages like vision.5 Because the Buddhist tradition speaks of the sūtras of Perfect Wisdom, the term Perfect Wisdom is generally understood as designating a specific corpus of texts. But Haribhadra argues that this is Perfect Wisdom only in a secondary sense; that is, it is derived from the original sense of Perfect Wisdom. He defines it as “the discernment of dharmas” (dharma-pravicaya), which is a long-standing definition of wisdom in the Buddhist tradition: wisdom or knowledge is the capacity of telling apart, identifying, (etymologically) analyzing the different factors or constituents (dharmas) that make up reality and our experience of it. For Haribhadra, this analytical power not only results in an ontological classification between different kinds of constituents of reality but also concerns the different “modes of existence” of these factors, that is, the ultimate truth ( paramārtha-satya) and the relative truth (saṃvr․ ti-satya).6 Thus Perfect Wisdom is nothing other than the capacity of seeing reality as it is ( yathābhūta) in its nature and its multiplicity, freed from the delusions of false ideas such as self and other substances (svabhāva). More generally, Haribhadra’s commentaries extend the meaning of Perfect Wisdom so as to comprise any kind of knowledge, including those he considers inferior, like the knowledge of non-Buddhists and of other Buddhist schools and groups (śrāvakas and pratyekabuddhas). Hence, the term Perfect Wisdom arguably includes both the objective aspect of truth (what reality is) and its cognitive counterpart (the apprehension of reality), standing as the ultimate standard of truth undifferentiated into subject and object. By extending the meaning in this way, Haribhadra accomplishes what we can call a “detextualization” of Perfect Wisdom: Perfect Wisdom is first and foremost the actuality of wisdom, before it is even considered a text.7 In the previous quotation, he describes this wisdom in two ways: he calls it “buddha” and “nondual knowledge that is similar to illusion.” It is not absolutely clear what he means by “buddha” in this passage because of the complexity of his theory regarding what a buddha is,8 but it is clear that he does not simply mean the historical figure born sometime in India around the sixth to fifth centuries BCE and known as Śākyamuni Gautama. He means rather the actual embodiment of wisdom, including its embodiment in a human life. The second description suggests also that “buddha” is primarily taken as an 500

Haribhadra

epistemic term, indicating a form of knowing cognition – the actual event of understanding rather than the abstract disposition of knowledge. The expression “nondual knowledge that is similar to illusion” (māyopamaṃ jñānam advayam) summarizes well the metaphysical position Haribhadra holds, which we will explore in the following. But in a secondary sense, Haribhadra argues, “Perfect Wisdom” can also signify texts that explain this primary, actual wisdom, or the path that leads to it, since both texts and path are called “Perfect Wisdom” only in reference to the first, primary sense of the phrase. Engineering is not a textbook on engineering but rather the knowledge used to accomplish bridge construction or road design; yet this primary sense of engineering gives its name to textbooks through which we can learn it. In the same way, Haribhadra has an instrumentalist vision of texts and the path: they are both teleologically oriented toward the realization of Perfect Wisdom, that is, the attainment of the summum bonum that buddhahood or nondual knowledge are understood to be. Thus philosophy’s main accomplishment is outside its texts and even outside the “philosophical practices” that are designed to make us wise. Rather, philosophy’s main accomplishment is the result consisting in wisdom. In this sense, there is much affinity between Haribhadra’s thought and pre-modern forms of philosophy in Greece, Rome, or the Islamic, Jewish, and Latin Middle Ages.9 Even though Perfect Wisdom does not have primarily a textual nature, it does express itself in many texts, in sūtras and in treatises. By relating the multiplicity of sūtras of Perfect Wisdom to a common non-textual source, Haribhadra can provide a commentary on Perfect Wisdom mixing in references to the Sūtra in Eight-Thousand Verses as well as other sūtras and to the Ornament. This plasticity in the expression of Perfect Wisdom suggests that Hari­ bhadra’s conception of a “text” differs from the way modern scholars think about texts.10 It is an ideal text expressing the essence of Perfect Wisdom he is aiming at in his commentaries rather than any specific historical text. It is within this space between the ideal text and the actual texts that Haribhadra’s creativity as an individual commentator can best be seen: his comments sometimes seem rather loosely related to the words he is explaining, and he allows himself to make many digressions or excursuses, which function as mini philosophical essays throughout his Light. In other words, he is not a literalist and is not blindly subservient to the written texts he has in front of his eyes. This broad consideration of what constitutes the “text” of Perfect Wisdom is also the basis for his criticism of those who preceded him in commenting the Ornament. In the opening verses of his Explanation, Haribhadra emphasizes the comprehensiveness of Perfect Wisdom and how the Ornament is a key to the understanding of all sūtras of Perfect Wisdom, not only one of them, as Vimuktisena (ca. fifth to sixth centuries) thought when he explained the Ornament as exclusively referring to the sūtra in 25,000 verses. Others showed partiality in their perspective, like Vasubandhu, who seems to be accused of explaining the text from his Yogācāra bias. In affirming his difference, Haribhadra found his own individual voice as a commentator who echoed, distinctly, that of Perfect Wisdom with its wide compass. Key to Haribhadra’s philosophical perspective is his commitment to such comprehensiveness and the multiplicity of approaches, which reflects his conception of the totality of the Buddhist path.

The Philosopher of the Path as the Great Synthesizer Haribhadra can rightfully be seen as one of the greatest Buddhist philosophers of the path, that is, the process of transformation that leads an individual to the summum bonum of buddhahood. His commentaries give an account of what constitutes this process, the qualities to develop and faults to get rid of, the possibility of different trajectories and stages, and so on. 501

Pierre-Julien Harter

This account is not really descriptive, as if he were to report his own experience or that of others. It rather provides a theory of the path: given a certain conception of buddhahood, arguments can be elaborated about what counts as the path and what does not, what can be deemed the best path, what is required for specific betterment, and so on. This is why his commenta­ ries are not directly practical, since their purpose is not to provide a recipe for awakening but instead to reflect about what the path is (Dreyfus 2003, 172–76). Many Buddhist philosophers have focused on issues of metaphysics, understood as a discourse about the general nature of things. But once we know, for instance, that reality is insubstantial, that it is devoid of any substance or essence, as the Mādhyamika philosophers like Nāgārjuna or Āryadeva argue, what do we do about it? And in what sense do we know that this is the case? The reflection on the status and the different kinds of knowledge is a longstanding approach in the Buddhist tradition. Haribhadra situates himself in a line of thought that privileges forms of knowledge that are “nonconceptual” (avikalpita) in contrast with forms of knowledge that are “conceptual.” What is wrong with a concept? Haribhadra quotes Dharmakīrti’s definition of conceptualization: One talks about conceptualization because something appears that is not a thing, and it is a disturbance because it is discordant.11 The idea is that a concept or conceptualization (vikalpa) is something that is not a particular thing but rather a general entity or universal. General entities have very different characte­ ristics than do particulars: they are not restricted in place and time, they do not “appear” to a subject (at least in the same sense as things can appear in perception), and they are the result of processes of determination and abstraction. In that sense, concepts are not (particular) things, are not real (because reality is impermanent and insubstantial), and are some sort of fictions resulting from a judgment subsuming multiple instances under a unity – hence they are “discordant” in never exactly corresponding to particulars. Nevertheless, these fictions can allow us to understand things and reality if we don’t get mesmerized by them such that we come to believe that they have more ontological weight than they actually do. This is a very important aspect of Haribhadra’s (and other Buddhist thinkers’) thought that cuts through the heart of his theory of the path: fictions and even errors can help us to understand and to progress toward the summum bonum as long as we can see through them, as we shall see in the following. Such definition already shows that epistemological developments of Dignāga and Dharmakīrti are presupposed and put in use in his theory of the path. These epistemological and ontological considerations have epistemic consequences for the path. If a conceptual co­gnition does not conform to reality but can still help to understand it, the process of cognitive progress can reproduce this passage from conceptual to nonconceptual understanding. This is the paradigm he uses as a representation of the path: The science of all aspects, which is the full knowledge of all aspects such as the production of the thought of awakening, is stated first for the bodhisattva longing for buddhahood by indicating that it is the fruit [...] Then in order to have mastery over the three kinds of omniscience that have been acquired like that, again one realizes in every way the awakening of all aspects by cultivating the three kinds of omniscience, which consists in taking together all the sorts of gnoses of aspects, paths, and elements.12 In the first sentence of this typically technical passage, Haribhadra indicates that the first chapter of the Ornament is about the science of all possible aspects of reality  – in other 502

Haribhadra

words, the highest and broadest knowledge possible, that of a buddha. It is presented first, he writes, because it is the end of the path, and the bodhisattva, the ideal practitioner of the Buddhist path, needs to know the destination to which the path leads in order to be willing to engage in it. The next two chapters present two lower kinds of knowledge, which have all paths and all elements as their objects, respectively, and belong to bodhisattvas and hearers. Reading about these three kinds of omniscience might be enough to “acquire” ( prāpta) them (one can understand, through arguments, what reality is and, broadly, what ethical practices are warranted), but this is not enough to “realize” (adhigamya) them. These two cognitive and soteriological states of acquiring and realizing represent the two ends of the path, and in other passages, Haribhadra clearly identifies the first moment of knowledge as a conceptual one and the second as a nonconceptual one. The path is this passage from a certain inferior kind of knowledge about reality in all its dimensions (i.e., elements, paths, and aspects) to a superior form of knowledge, a knowledge that is transformative because it is actual and no longer merely abstract. The passage from the first to the second form of knowledge is ascribed to cultivation (bhāvanā), that is, a form of engagement that is not only cognitive, but also practical, emotional, bodily, and so forth, that consists in habituating the practitioner to a certain understanding so that it is appropriated, implemented, and becomes truly the way she sees reality – hence the quotation indicating that all aspects, paths, and elements need to be “taken together.” This conception raises questions about the conditions for transformation, specifically about the epistemological characteristics a cognition must bear to be able to be transformative. In more familiar Western philosophical terms, the question is how understanding can determine the will. How is the transformation of an individual based on a philosophical approach possible? Greek philosophers, for instance, considered progress toward the good life an essential aspect of what philosophy is about, but they said little about how to proceed with it. The concept of the path implies that a theoretical proposal cannot by itself trigger the change necessary for the summum bonum to be attained. In other words, listening to the Buddha does not make you become a buddha. There is much work to be done. An active process of transformation is necessary that would guarantee the passage from the first stage of knowledge to the second. This is what the Ornament, and by implication Haribhadra’s commentaries, are about. Hari­bhadra, following a lead from Vimuktisena, addresses the question of the subject matter of Perfect Wisdom in the Ornament at the beginning of his two commentaries. He says there are three options concerning this topic: it can be about the fundamental elements of reality (vastu) – a sort of ontological and metaphysical categorization of what reality is made of. But, he argues, this would just reproduce what has already been done by Abhidharma treatises. Or the Ornament could present all the “remedies” ( pratipakṣa) – a sort of ethical approach that addresses the practices and ideas necessary for eliminating all obstacles that prevent us from attaining the summum bonum such as negative inclinations and affects. Such a treatise, Hari­bhadra explains, would be incomplete: to present “remedies,” you have to account for what they are remedies of. Finally, it could be about the totality of aspects (ākāra), that is, the totality of objects of knowledge. But Haribhadra wonders why this would differ from the first proposal: aspects should be related to the fundamental elements of reality. He finally makes his own proposal of a comprehensive and updated conception of the path as the subject matter of the Ornament and Perfect Wisdom that would comprehend all three options: It is the entire unmistaken path, comprehending the realizations of hearers, indepen­ dent buddhas, bodhisattvas, and buddhas – thus including the three kinds of things mentioned above and avoiding the faults of each option – organized in the stages of 503

Pierre-Julien Harter

the eight realizations starting with the science of all aspects, and leading to goods, whether they are success in existence or the final good.13 This gives us a helpful glimpse into Haribhadra’s wholistic understanding of the path: the path is a total transformation, whether it is cognitive, dispositional, emotional, ethical, and even physical, which is why its theory needs to comprehend all aspects, all paths, and all elements of reality. This is why Haribhadra’s commentary displays such comprehensiveness in its approach as well: within a Mādhyamika framework, he incorporates Abhidharma, epistemology ( pramāṇa), and Yogācāra elements, discussing positions to either appropriate them or refute them but also to articulate them along the dynamic arch of the path (Makransky 1997; Harter 2015).

Philosophical Excurses As noted, Haribhadra’s long commentary is a very rich text not only because it manages to provide a word-by-word exegesis of the long Eight-Thousand Sūtra on Perfect Wisdom together with the Ornament but also because it uses certain passages from these texts as opportunities to launch sustained philosophical developments – a procedure that turned into a norm in Tibet, where commentaries of the Ornament basically became encyclopedias of the Buddhist path in which commentators would address any issue they deemed interesting. These “digressions” or “excursuses” ( prasaṅga), as he calls them, are at times closely tied to the text of the sūtra and are at other times seemingly completely autonomous and quite long. Among many possible examples, we examine two here.

The Path as Beneficial Illusion Haribhadra advocates unfailingly for the Mādhyamika philosophy of universal emptiness: nothing has a substance or inherent essence (svabhāva), and yet things appear as if they do. Such a position is potentially threatening to the whole Buddhist path, since a nihilist interpretation is always possible: if nothing has any substance, what is the point of cultivating virtues, eliminating anger, developing compassion? How would a path with its different stages and practices be consistent? This is the reason one of Haribhadra’s main concerns in his commentaries is to assert the reality of the relative truth (saṃvṛti-satya), which is the ontological dimension where things appear so that they can be differentiated and identified – a necessary ontological dimension for any language and action – in contrast with the ultimate truth, which rejects any real existence of substances or essences. The danger of a nihilist reading of the sūtras of Perfect Wisdom was always looming, since these texts use negations heavily to reject any substantial conception of reality. They even go as far as negating the existence of the Buddha, of wisdom, of the path, of practitioners, of bodhisattvas, of virtues, and so on. One of Haribhadra’s signatures is to add as a gloss “in reality” (tattvena) to restrict such negations to the dimension of ultimate truth and to preserve some sort of ontological status for the relative truth. This strategy is essential for a philosopher of the path, without which the path would lose all significance. In one of his excurses, Haribhadra addresses the question of the status of these appearing realities, with particular reference to the appearance of beings (Wogihara 1932–1935, 159–61; translation in Sparham, vol. 2, 107–10). Often sūtras state that things are “similar to illusions” or “analogous to illusions” (māyopama). In one particular passage, the gods, who are often characters in the stories or interlocutors of the Buddha, wonder if there is a difference between saying that beings are similar to illusions or that they are just plainly illusions – the question possibly pointing at the seeming similarity of the Buddhist position to the Vedāntin idea of the world 504

Haribhadra

being an illusion (māyā). If beings are just similar to illusions, it might mean that they are not really illusions and that behind an illusory veil we can find what they really are. That is a slippery slope for Mādhyamikas. In a nutshell, Haribhadra explains that, even if there is no ontological difference between beings and illusions, there is still a phenomenological one. Indeed, beings are illusions: the Buddhist tradition is known for its rejection of the existence of a self as some sort of a core that would give beings their identities and unities. There is no such inhe­ rent core or soul but rather fleeting bodily, sensuous, conscious, and unconscious processes that function together for some time and to which we attribute identities. Such “consistent functioning” gives the illusion of identity, which is a conceptual construction (vikalpa) subsuming under an imaginary unity multiple different instances in time and place. And so, Haribhadra argues, there is no ontological difference between beings and illusions; otherwise we would have to believe that there is something real “behind” the illusion of individual beings, and we would be back to the same substantialist assumption that the Mādhyamikas fight so hard to uncover. Using the analogy of a magic show, Haribhadra explains that what appears during the show is not real in the sense that it is a pure illusion fabricated by a performer who knows how to use tricks to conjure up images appearing to spectators. And yet, during the magic show, there are still things that appear to the spectators, and these appearances make a difference between what can actually appear and have a relative reality (due to their being produced by the mechanism of the tricks; in other words, due to causal efficiency, the mark of relative truth for Haribhadra following Dharmakīrti) and what cannot appear because it is just purely nonexistent, as illustrated by the well-known examples of the son of a barren woman or the horns of a rabbit, which are outside any possible causal production. This double dimension of reality, Haribhadra continues, does not suffer exceptions: even the Buddha or nirvāṇa, he states in a very Nāgārjunian fashion, are, ontologically speaking, illusions because they are not ultimately real. No single phenomenon can be excused from emptiness; otherwise emptiness would be only partial. But that does not mean that suffering in cyclic existence or being liberated from it are phenomenologically the same. Phenomenologically speaking, they are said to be like illusions because they can still be differentiated one from another and because they still function in ways that really do allow the practitioner to abandon ignorance and to cultivate Perfect Wisdom. A second “excursus” (Wogihara 1932–1935, 205; translation in Sparham, vol. 2, 152) could be usefully connected with this one to get a fuller theory of illusion, one which allows Hari­ bhadra to argue that the liberating process of the path can make use of illusions. In this passage, Haribhadra wonders whether the perfect knowledge of a buddha has any intentional object at all: if all things are insubstantial according to Mādhyamika metaphysics, wouldn’t a consistent epistemological position amount to an absence of any intentional object, since intentional objects always entail a certain nature or substance? This is a common objection raised by non-Buddhist thinkers in India, namely that the Buddhist summum bonum resembles more that of a stone than of an accomplished human being. And after all, the sūtras do repeat that the final cognitive state is “without intentional object” (anālambana). Haribhadra responds by making a distinction between error (bhrānti) and aspect (ākāra). Using the example of the dream, he explains that the objects seen in a dream are indeed ultimately nonexistent (they disappear once we wake up), and yet these objects, which are mental aspects seen by the person who is dreaming, follow certain causal processes: they provoke fear or desire in that person, they can be acted upon, and so on. In the same way, in the waking state, many cognitive states like perception arise with various clear and distinct objects, which do appear as having unity (although there is none) and as being external (although they always appear through our cognitions and we have no way to step outside of them to verify whether they reflect what is actually outside). These aspects of our cognitions follow causal processes 505

Pierre-Julien Harter

that explain their various appearances and modifications, but these aspects themselves are not “false” once we understand how they are the results of causes and conditions. Only our naïve judgment that their appearance indicates ultimate reality is problematic. This is why Haribhadra concludes: Even when error stops, it is impossible for aspects to stop since there is no connection between them.14 Spinoza argued in a similar fashion when he pointed out that even when we understand that the appearance of the sun as being only 200 feet away from us is false and that it is in truth more than 600 diameters of the earth away from us, the false appearance of it being so close does not cease. Why? Because it is created by a complex of causes and conditions (our situation in the universe, the laws of reflection of light, the particular condition of our eyes, etc.) that results in such appearance. The appearance is true when we can understand its causal genealogy, but the judgment we superimpose on it – that the sun is indeed 200 feet away – is erroneous. Such argument is more radical than it seems because it means that all appearances can be transfigured into elements of truth if we know how to consider them: when believed naïvely, they can fool us, but when we can understand how they come to be, they open up opportunities for understanding and liberation from error. And thus, even though they can be deemed ultimately false, when reconsidered within the web of the causal process, they can turn out to be useful for the progress on the path. A fascinating example Haribhadra takes is how dedication, that is, the mental disposition and action to dedicate the benefits of an action to the wellbeing of all beings (a core practice in Mahāyāna Buddhism consisting in orienting oneself to constantly act for the benefit of others instead of pursuing one’s selfish goals), can be used to progress on the path even though it is devoid of any reality.15 To dedicate an object or an action to all beings, one has to use a concept of that object; that is, one has to superimpose onto past, present, and future instances of that object (which can never coexist) a unity that is a pure fabrication. And yet this illusion can be beneficial and contribute to the progress on the path (by cultivating compassion and lessening self-promotion, for instance), as long as the practitioner is aware of the manipulation. It is when this dedication (and everything entailed by it: the practitioner, its addressees, etc.) is understood as being “similar to an illusion” that a gap between the naïve realist apprehension and the nihilist diversion from ethical action can open up for the possibility of knowledge and progress toward liberation.

Haribhadra the Metaphysician: The Great Discussion on the Ultimate Nature of Things The longest excursus in which Haribhadra indulges is an inquiry into the nature of reality or suchness, which he frames as an inquiry into the nature of buddhahood. For Haribhadra, buddhology (understood as the inquiry into what a buddha is) is never separate from metaphysics, and there is a metaphysical reason for that: buddhahood is not a personal condition but rather nothing else than the actual knowledge of reality. This knowledge is nondual in the sense that it does not consider reality an object projected in front of (ob-jectum) a subject (as if the subject could exclude itself from reality). Such a framework of subject–object is rejected as a distortion of reality. Haribhadra, in a somewhat similar fashion to Neoplatonist thinkers, maintains a radical or idealized conception of knowledge free from the duality of subject and object, even though he still accepts a range of inferior kinds of knowledge that maintain a dualist form and can be more or less attuned to it. The long excursus of chapter four (Wogihara 506

Haribhadra

1932–1935, 624–42; translation in Sparham, vol. 3, 238–71) deserves a much longer study, but we can at least focus on the result of the inquiry. After asserting the absence of a diffe­rence between what a buddha is and reality, Haribhadra wonders what account he can give of what is real and whether it has a single or a multiple nature, a line of inquiry common to many thinkers of the time.16 His inquiry takes the shape of a review of hypotheses that he identifies with specific philosophical positions. He first refutes the position that ascribes reality to objects of knowledge ( jñeya), especially the materialist-atomistic position that takes atoms as constituting reality. Next he turns toward the position that considers cognitive instances ( jñāna) the only exis­tent reality, which he attributes to Yogācāra thinkers. Haribhadra seems to accept the idea that such mental events can be described as nondual, impermanent, and endowed with aspects. This last characteristic indicates that our cognitive instances display appearances of distinct objects as well as appearances of subjects, including our own consciousness as aware of intentional objects. The rest of the discussion concerns the ontological status of these aspects, and once again he frames the discussion in terms of two sub-positions in the Yogācāra perspective (although he never actually frames them as being “schools”). The first is the position that takes aspects to be real (sākārajñāna), and the second is the position that takes aspects to be false (nirākārajñāna). He rejects the first position as lending too much credit to aspects. Haribhadra posits that the mind needs to be thought of as a capacity of being a unity, holding a potentially infinite number of changing aspects. Such multiplicity in unity can only be noncontradictory if those aspects are unreal, since a real unity would be contradictory with a real multiplicity.17 He criticizes the second option as lending not enough credit to aspects when they are dismissed as false, because such a perspective misses their causal and relative natures. Thus cognitive instances are no more ultimately existent than objects of knowledge but no more illusory than other relative phenomena. This extended, detailed, and subtle argument can be understood as a summary and reassessment of his commitment to insubstantiality, reformulated in terms of the kind of ideal knowledge (that of a buddha) to which the practitioner of the path should aspire. Haribhadra’s final characterization of such knowledge is that of “a mental continuum that is merely nondual consciousness, which is similar to an illusion.”18 The awareness that knows reality as it is is itself impermanent; that is, it is a continuum or series ( prabandha) of mental events arising due to a complex causal process. This claim therefore means that even the knowledge of a buddha has a relative nature being made up by a causal process – quite a bold statement for some Buddhist thinkers. This mental continuum is not a pure (subjective) idealism à la Fichte or Berkeley but a nondual reality, since objects never appear outside of a conscious event, and awareness is itself not separated from its objects qua mental appearances. But these should not be given an ultimate ontological status either, since the whole structure of nondual awareness is itself “similar to illusion,” an expression, as we have already discussed, that tries to escape the pitfalls of realist and nihilist interpretations. Haribhadra’s position thus avoids the tempting characterization in terms of idealism, since awareness or consciousness is not given a privileged metaphysical status. It is just as insubstantial as other phenomena. He should be understood in his own terms: as a Mādhyamika thinker who strives to maintain a balance between ultimate and relative truths all the way to his interpretation of final truth or buddhahood. Finally, his style of metaphysics does not adopt a purely theoretical approach, as the common contemporary conception of metaphysics would want. It is a metaphysics closer to Neoplatonist or a range of medieval (Latin, Arabic, or Jewish) metaphy­ sical thoughts, which see metaphysics as a practical process of gradual insight bringing about co­gnitive, spiritual, ethical, and emotional progress. He stresses at the end of this passage how 507

Pierre-Julien Harter

this final position should be cultivated through the well-known Buddhist process of the three kinds of understanding: understanding through study (śrutamayī), which is purely receptive; understanding through reflection (cintāmayī), which uses discursive analysis to deconstruct substantial assumptions and arrive at an insight into the nature of “illusion-like” reality; and finally understanding through habituation (bhāvanāmayī), which appropriates in one’s mental continuum the results of the two previous understandings. Metaphysics is thus more than a theoretical scholastic exercise: it is what fuels the very process of the path, and it has direct practical implications.

Conclusion Haribhadra’s thought shows a quite remarkable breadth and force, which explains why he exerted a powerful influence on late Indian and Tibetan Buddhist thinkers (the Ornament was never translated in China before the twentieth century; hence he was completely unknown to pre-modern Chinese and Japanese thinkers). Indian commentators saw his work as a gamechanger in the exegetical tradition of the Ornament, which explains why his small commentary was in turn the object of subcommentaries by Dharmamitra (ninth century) and Dharmakīrtiśrī (ninth to tenth century). But he was also hotly contested by important late Indian figures like Ratnākaraśānti (ca. 970–1045 CE) due his pro-Mādhyamika interpretation and his stance on multiple topics of the Buddhist path. Tibetan commentators of the Ornament almost universally considered him an authority, although rare exceptions like Dol po pa (1292–1361) went as far as reviling him as an inferior commentator with a narrow understanding of Mādhyamika philosophy. As argued previously, Haribhadra’s work offers a double perspective worth exploring: a comprehensive view of Buddhist philosophy and especially of the central concept of the path, which can be extracted from an overall consideration of his various compositions, and more focused reflections on multiple subjects ranging from hyper-specialized questions on the path to general metaphysical and Mādhyamika topics. His thought is complex since it is formulated within the confines of commentaries of one of the most technical texts of Buddhist philosophy, and it is not always easy to distinguish between what is truly original and what is an appropriation of the exegetical tradition. It also represents a challenge for philosophers, as it seems to vacillate between what could be designated at times as philosophy and at other times as theology, depending on the nature and the extent of the assumptions in play. But arguably, the history of theology is also a part of the history of philosophy in the West, and it would be problematic to exclude it, even for the very understanding of the evolution of philosophical thought. Just as problematic is the neglect of Haribhadra’s works, which represent a line of thought that has marked late Indian Buddhist philosophy decisively in shaping a comprehensive and systematic account of the Buddhist path.

Notes 1 The qualifications “vast” and “deep” are Tibetan interpretative tools and are useful heuristic means for presenting the aims of this literature. See Harter 2019b. 2 For details, see Seyfort Ruegg 1981, 101–2; Makransky 1997, 213–14; Sparham 2006, xv; Apple 2008, 29–30; Harter 2015, 11–18; Harter 2018, 204–5. 3 In addition, during the history of Buddhist philosophy, no scholar seems to be have questioned the authorship of these three works, apart from Ngog blo ldan shes rab, who doubted that the Commentary on the Versified Summary of the Precious Qualities was Haribhadra’s (Obermiller 1937, 5).

508

Haribhadra 4 The Hegelian statement seems to be an apocrypha coming from the Hegelian tradition, since no such statement can be found in Hegel’s writings. 5 My translation. Wogihara 1932–1935, 23; Sparham 2006, 199. 6 On the topic, see chapters on Nāgārjuna, Āryadeva, Śāntarakṣita, Candrakīrti in this volume. 7 The move is not unique to Haribhadra, although it is certainly quite forceful in this passage. For another similar “de-textualization” strategy, see the beginning of Vasubandhu’s Commentary on the Treasury of Abhidharma (Abhidharmakośa-bhāṣya) and his definition of abhidharma. 8 This is known as the theory of the buddha-bodies (kāya), which was harshly debated by commentators of the Ornament. See Makransky 1997 for a comprehensive overview of the Indian and Tibetan debates. 9 Some might talk of a common conception of philosophy as a way of life as in Pierre Hadot’s works, although differences should not be underestimated. See Hadot 1995; Kapstein 2013; Harter 2018. 10 Maria Heim develops a similar idea about Buddhaghosa (Heim 2018, 3–4). 11 Wogihara 1932–1935, 158: vikalpo ’vastu-nirbhāsād visaṃvādād upaplava. The full definition can be found in PVin 1.33ab. 12 Wogihara 1932–1935, 16. 13 Ibid., 4. 14 Ibid., 205. 15 See the whole passage in Wogihara 1932–1935, 332–49; translation in Sparham, vol. 2, 232–40. 16 See, for example, the chapter on Śāntarakṣita and Kamalaśīla in this volume. 17 See also Wogihara 1932–1935, 534; Sparham, vol. 2, 176–78. 18 Wogihara 1932–1935, 641.3: māyopāmādvayavijñānamātraprabandha.

Bibliography Amano, Koei H. 2000. Abhisamayālaṃkāra-kārikā-śāstra-vivṛti: Haribhadra’s commentary on the Abhisamayālaṃkāra-kārikā-śāstra edited for the first time from a Sanskrit manuscript. Kyoto: Heirakuji Shoten. Apple, James. 2008. Stairway to Nirvāṇa: A Study of the Twenty Saṃghas Based on the Works of Tsong kha pa. Albany: State University of New York Press. Conze, Edward. 1973. The Perfection of Wisdom in Eight Thousand Lines and Its Verse Summary. Bolinas: Four Seasons Foundation. Reprint, San Francisco: City Lights, 2006. ———. 1975. The Large Sutra on Perfect Wisdom. Berkeley: University of California Press. ———. 1978. The Prajñāpāramitā Literature. 2nd ed. Tokyo: The Reiyukai. Dreyfus, Georges. 2003. The Sound of Two Hands Clapping: The Education of a Tibetan Buddhist Monk. Berkeley, Los Angeles and London: University of California Press. Hadot, Pierre. 1995. Philosophy as a Way of Life. Malden and Oxford: Blackwell Publishing. Harter, Pierre-Julien. 2015. “Buddhas in the Making: Path, Perfectibility, and Gnosis in the Abhisamayālaṃkāra Literature.” PhD diss., University of Chicago. ———. 2018. “Spiritual Exercises and the Buddhist Path: An Exercise in Thinking with and against Hadot.” In Buddhist Spiritual Practices: Thinking with Pierre Hadot on Buddhism, Philosophy, and the Path, edited by David V. Fiordalis and Luis O. Gómez, 147–79. Berkeley, CA: Mangalam Research Center for Buddhist Languages. ———. 2019a. “Haribhadra,” In Brill’s Encyclopedia of Buddhism. Volume II: Lives, 204–8. Leiden and Boston: Brill. ———. 2019b. “From Exegesis to Hermeneutics: Rongtön and the Hidden Meaning (sbas don) of Perfect Wisdom.” In Reasons and Lives in Buddhist Traditions: Studies in Honor of Matthew Kapstein, edited by Dan Arnold, Cécile Ducher, and Pierre-Julien Harter, 379–91. Somerville, MA: Wisdom Publications. Heim, Maria. 2018. Voice of the Buddha: Buddhaghosa on the Immeasurable Words. Oxford: Oxford University Press. Kapstein, Matthew. 2013. “Stoics and Bodhisattvas Spiritual Exercise and Faith in Two Philosophical Traditions.” In Philosophy as a Way of Life: Ancient and Modern. Essays in Honor of Pierre Hadot, edited by Michael Chase, Stephen R. L. Clark, and Michael McGhee, 99–115. Malden and Oxford: Wiley Blackwell. Makransky, John. 1997. Buddhahood Embodied: Sources of Controversy in India and Tibet. Albany: State University of New York Press. Mano, Ryūkai. 1972. Genkan shōgonron no kenkyū. Tokyo: Sankibōshoten.

509

Pierre-Julien Harter Naughton, Alexander. 1989. “The Buddhist Path to Omniscience.” PhD diss., University of Wisconsin, Madison. Obermiller, Eugene. 1937. Prajñā Pāramitā-Ratna-guṇa-saṃcaya-gāthā: Sanscrit and Tibetan Text. Bibliotheca Buddhica, 29. Leningrad: Akad. Nauk. Pollock, Sheldon. 1985. “The Theory of Practice and the Practice of Theory in Indian Intellectual History.” Journal of the American Oriental Society 105 (3): 499–519. Sanderson, Anderson. 2009. “The Śaiva Age – The Rise and Dominance of Śaivism during the Early Medieval Period.” In Genesis and Development of Tantrism, edited by Shingo Einoo, 41–349. Tokyo: Institute of Oriental Culture, University of Tokyo. Seyfort Ruegg, David. 1981. The Literature of the Madhyamaka School of Philosophy in India. Wiesbaden: Otto Harrassowitz. Sparham, Gareth, trans. 2006–2012. Abhisamayālaṃkāra with Vṛtti and Ālokā, Vṛtti by Vimuktisena, Ālokā by Haribhadra, vols. 1–4. Fremont: Jain Publishing Company. Wogihara, Unrai. 1932–1935. Abhisamayālaṃkār’ālokā Prajñāpāramitāvyākhyā, Commentary on the Aṣṭasāhasrikā-Prajñāpāramitā by Haribhadra together with the text commented on. Tokyo: The Toyo Bunko. Yuyama, Akira. 1976. Prajñā-pāramitā-ratna-guṇa-saṃcaya-gāthā: Sanskrit Recension A, Edited with an Introduction, Bibliographical Notes, and a Tibetan Version from Tunhuang by Akira Yuyama. Cambridge and New York: Cambridge University Press.

510

31 ŚĀNTIDEVA Virtue on the Empty Path of the Bodhisattva Stephen Harris

Śāntideva was an eighth-century CE Mahāyāna Indian Buddhist monk. He was the author of two surviving texts, the Introduction to the Practice of Awakening (Bodhicaryāvatāra; hereafter BCA) and the Compendium of Trainings (Śikṣāsamuccaya; hereafter ŚS). Śāntideva’s philosophical influence is most significant in the field of moral philosophy, with each text making substantial contributions to the study of virtue, well-being and the relation between metaphysics and ethics. The ninth chapter of the BCA also presents an influential exposition and defense of the core Madhyamaka metaphysical tenet of emptiness (śūnyatā).

Life and Major Works Very little is known about Śāntideva’s life. He is generally accepted to have lived between the late seventh and mid-eighth century CE. Traditional Tibetan biographies claim that he spent much of his career at the Indian monastic university of Nālandā. Both the BCA and the ŚS are training manuals for the development of the bodhisattva, the being who takes a vow to achieve full buddhahood in order to liberate all sentient beings. The influence of the shorter text, the BCA, is attested to by the existence of numerous Indian and Tibetan commentaries (Williams 1998, 4–5). Recently, it has undergone a resurgence of popularity, with a number of influential contemporary practitioners of Tibetan Buddhism, including the fourteenth Dalai Lama and Pema Chödrön, writing commentaries on it (Gyatso 1994; Chödrön 2007). The last several decades have also seen an increase in academic attention to the BCA; it is often characterized as the most significant text for the study of Indian Buddhist ethics (Goodman 2009; Garfield 2012). By contrast, until recently, the ŚS received relatively little academic attention and almost no philosophical study. A new translation of the ŚS by Charles Goodman (2016), as well as several recent important studies (Mahoney 2002; Clayton 2006; Harrison 2007; Lele 2007; Mrozik 2007), have increased its academic visibility. The received version of the BCA is composed of ten chapters that total 913 verses. This recension has been the subject of all extant commentaries and has been translated into English recently a number of times (Śāntideva 1997, 2006, 2008). A shorter version of the BCA, composed of nine chapters and 701.5 verses, is preserved in Tibetan translation but has not yet been published (Saito 1993). Most contemporary scholarship takes the longer text as its object of study, as do I in this chapter. DOI: 10.4324/9781351030908-44

511

Stephen Harris

The primary purpose of both the BCA and the ŚS is to instruct the reader in developing the virtuous qualities of the bodhisattva. These include the six perfections ( pāramitās) of generosity (dāna), ethical restraint (śīla), patience (kṣānti), energetic effort (vīrya), concentration (samādhi) and wisdom ( prajñā), as well as closely associated virtues such as compassion (karuṇā), mindfulness (smṛti) and introspection (saṃprajanya). A significant difference between the two texts is that the BCA is presented in Śāntideva’s own voice and is organized around the six perfections ( pāramitās). By contrast, the ŚS is largely composed of quotations from Mahāyāna Buddhist sutras. It presents its insights through a frame of twenty-seven verses written by Śāntideva, articulating the central training points of the bodhisattva path, which are interspersed throughout the text’s body. Each verse is further developed with occasional commentary by Śāntideva and through quotations from Mahāyāna sutras which constitute most of the text’s length. As Susanne Mrozik argues, the ŚS should not be looked on merely as an anthology of Mahāyāna sutras; Śāntideva’s organization of the sutra quotations, as well as his brief verse and prose commentary, allow a distinctive authorial voice to emerge (2007, 10–12). Nevertheless, we cannot assume that every line of the sometimes lengthy sutra quotations expresses a view which Śāntideva would fully endorse. For this introduction, therefore, I focus primarily on the shorter and more influential BCA. Nevertheless, the ŚS treats certain topics in much greater depth than the BCA, making it invaluable in understanding his thought. I draw upon it in my explanation of Śāntideva’s presentation of generosity (dāna), a topic which receives much less attention in the BCA than it does in the ŚS. In this chapter, I focus on four philosophically significant topics treated in the BCA and the ŚS: bodhicitta and the bodhisattva vow; the perfection of generosity; the perfection of patience; and the perfection of wisdom, with emphasis on its role in the development of the other virtues. In terms of this last topic, I give particular attention to an influential argument in the BCA’s eighth chapter, in which Śāntideva links acceptance of the nonexistence of a self to altruistic commitment. In addition, one of the most philosophically stimulating claims made in the BCA is that adopting the apparently demanding bodhisattva path benefits the bodhisattva. Therefore, in this chapter, I also explore how altruism and self-interest converge as a result of Śāntideva’s development of the perfections.1 This in no way exhausts Śāntideva’s philosophical contributions; nevertheless, exploring these themes will serve as an introduction to much of what is most distinctive about his ethical thought.2 As we examine these positions, however, we should keep in mind that the BCA and ŚS are not merely theoretical texts meant to provide a conceptual understanding of virtue and moral motivation. Rather, both are meditation manuals whose purpose is to aid in the development of virtue for those engaged on the bodhisattva path. Most of the characterizations of virtue presented in the texts, as well as the arguments Śāntideva provides to defend his ethical positions, also function as meditations designed to develop virtue and moral commitment. This is the case, for instance, in Śāntideva’s argument linking the metaphysical position of selflessness (ānatman) with a commitment to impartial benevolence, which I  will consider below. It also explains Śāntideva’s predilection for providing numerous distinguishable meditations and arguments for developing a given virtue. Philosophical understanding plays an important role in moral development for Śāntideva, but theoretical knowledge for its own sake is not his primary aim.

Bodhicitta and the Bodhisattva’s Vow A central concept in both texts is the bodhisattva, the being (sattva) who vows to achieve the full awakening (bodhi) of a buddha in order to most effectively work for the benefit of sentient 512

Śāntideva

beings. The motivation to attain this goal is bodhicitta, the mind (citta) dedicated to full awakening (bodhi). The first four chapters of the BCA are focused on bodhicitta: the first praises its virtuous qualities; the second and third present a ritual in which it is strengthened and the bodhisattva vow is taken; the fourth focuses on maintaining it. These early chapters stress the dangerously unstable state we are in prior to entering the bodhisattva path. Śāntideva introduces this theme in a well-known pair of verses at the beginning of the first chapter. Just as lightning illuminates the darkness of a cloudy night for an instant, in the same way, by the power of the Buddha, occasionally people’s minds are momentarily inclined toward merit. Thus, virtue is perpetually ever so feeble, while the power of vice is great and extremely dreadful. If bodhicitta did not exist, what other virtue would overcome it? (BCA 1:5–6, translation modified)3 The verses make two striking and philosophically significant points. First, the minds of those outside the bodhisattva path are depraved, unable to focus on virtuous development with any consistency. Vice ( pāpa) here indicates the mental afflictions (kleśas), in particular craving (lobha), anger (krodha) and delusion (moha). Our dispositions (anuśaya) for their arising are so strong that consistent virtuous activity is impossible. Second, bodhicitta itself is the best, or perhaps the only, solution to our predicament. This is surprising, since developing bodhicitta and undertaking the bodhisattva path requires remaining in saṃsāra for eons, during which time one perfects generosity by sacrificing one’s body (BCA 7:20–22) and willingly taking painful rebirths (BCA 8:107–108) to work for the benefit of beings. Yet this claim, of the self-beneficial power of bodhicitta and the virtues of the bodhisattva, is a frequent and repeated theme within the BCA: Those who long to overcome the abundant miseries of mundane existence, those who wish to dispel the adversities of sentient beings, and those who yearn to experience a myriad of joys should never forsake bodhicitta. (BCA 1:8, translation altered) The second line gives the expected advice that those dedicated to eliminating the suffering of sentient beings should adopt the bodhisattva path. Developing the perfections, like generosity, patience and so on, will enable the bodhisattva to effectively intervene for their aid. Surprisingly, the verse also claims that those pursuing their own well-being, both in terms of liberation from saṃsāra and ordinary saṃsāric happiness, should also develop bodhicitta. Claims like this are repeated forcefully in the eighth chapter, which emphasizes the self-beneficial aspects of compassion: All those who are unhappy in the world are so as a result of their desire for their own happiness. All those who are happy in the world are so as a result of their desire for the happiness of others. (BCA 8:129) The central concepts which make Śāntideva’s claimed convergence between altruism and selfinterest intelligible are the mental afflictions (kleśas) and a set of virtuous mental states called 513

Stephen Harris

the perfections ( pāramitās). The most important mental afflictions, for Śāntideva and many Mahāyāna authors, are craving (lobha), anger (krodha) and delusion (moha). For Śāntideva’s Madhyamaka school, delusion refers most centrally to the superimposition of intrinsic existence (svabhāva), an essence which constitutes non-relational identity, upon impermanent and dependently originated phenomena. Attachment and anger are the affective states of attraction and repulsion, which are generated as a result of this superimposition. Śāntideva also recognizes a number of other mental afflictions  – including laziness (ālasya), jealousy (īrṣyā) and pride (māna) – which contaminate our experience. The mental afflictions are the deep causes of our suffering, inciting harmful actions and generating negative karma, which leads to unfortunate rebirths. The perfections ( pāramitās) are six virtuous mental states which many Mahāyāna authors, including Śāntideva, present as the central qualities developed by the bodhisattva. They are generosity (dāna), ethical restraint (śīla), patience (kṣānti), energetic effort (vīrya), concentration (dhyāna) and wisdom ( prajñā). Their primary role is to eliminate the mental afflictions (kleśas). Energetic effort, for instance, acts as the antidote to laziness by providing the bodhisattva with the motivational force to progress along the path and, at its conclusion, to tirelessly work for the benefit of sentient beings. The perfections, then, have two interrelated functions. First, they eliminate the mental afflictions and therefore remove all deep sources of suffering for their possessor. This is why progressing on the bodhisattva path, which is largely constituted by their development, is in the individual’s own benefit. Second, they enable the bodhisattva to work to liberate sentient beings from suffering. Developing the perfections, therefore, is the most effective way to pursue one’s own well-being, as well as the well-being of others. These early chapters also present a ceremony called “The Ritual of Supreme Worship (anuttara-pūjā),” which constitutes the entirety of BCA chapters two and three. In this ceremony, the bodhisattva makes offerings to the buddhas and bodhisattvas and confesses past wrongdoings as a way of preparing to take the bodhisattva vow. Powerful expressions of compassion are made directly before the vow is taken. May I be a protector for those who are without protectors, a guide for travelers, and a boat, a bridge, and a ship for those who wish to cross over. May I be a lamp for those who seek light, a bed for those who seek rest, and may I be a servant for all beings who desire a servant. (BCA 3:17–18) The Bodhisattva Vow follows shortly afterwards: Just as the Sugatas of old adopted bodhicitta and just as they properly conformed to the practice of the bodhisattvas, So I myself shall generate bodhicitta for the sake of the world; and so I myself shall properly engage in those practices. (BCA 3:22–23, translation altered)

514

Śāntideva

The incorporation of ritual activity in a philosophical text may at first seem incongruous, but it illustrates that Śāntideva’s primary commitment in the BCA is to providing practices which will develop virtue and moral commitment in practitioners on the bodhisattva path. For Śāntideva, ritual activity complements clear intellectual understanding by developing a feeling of joyous confidence ( prasāda) in the efficacy of Buddhist trainings. Moreover, both the ritual and the taking of the vow itself embed the bodhisattva in an infinitely long lineage of buddhas and bodhisattvas who have previously traveled on the bodhisattva path. Both thereby aid the bodhisattva in adopting a particular kind of conventional identity, grounded in the central project of liberating sentient beings from suffering. Philosophical argumentation and ritual activity are each kinds of cultivation (bhāvanā) by which to develop the virtues of bodhisattvahood. The bodhisattva vow, expressing the bodhisattva’s infinite commitment to remain in saṃsāra forever to work for the benefit of beings, also indicates a creative tension running throughout the entire text. With his promise, the bodhisattva binds all his future lives into a shared project of liberating the world from suffering. Nevertheless, Śāntideva later explicitly acknowledges that future lives are not numerically identical to the current one (BCA 8:98), presumably given their lack of psychological and physical continuity (Harris 2018). Moreover, the bodhisattva himself, and all those he liberates, are empty of intrinsic existence. The bodhisattva’s infinite commitment, therefore, is open to charges on multiple fronts that it is inconsistent with Buddhist commitments to selflessness (anātman) and impermanence (anitya). Śāntideva confronts this tension in the ninth chapter, which is dedicated to developing wisdom ( prajñā). [Qualm:] If no sentient being exists, for whom is there compassion? [Mādhyamika:] For one who is imagined through delusion, which is accepted for the sake of the task. [Qualm:] If there is no sentient being, whose is the task? [Mādhyamika:] True. The effort, too, is due to delusion. Nevertheless, in order to alleviate suffering, delusion with regard to one’s task is not averted. (BCA 9:75–76) Here, Śāntideva’s opponent objects that the metaphysical selflessness of the bodhisattva and those he liberates makes his task incomprehensible. Śāntideva’s response is to claim that delusion itself can be embraced to reestablish the existence of persons and provide a provisional ground for compassion. The term translated as delusion is moha, a synonym for avidya, the superimposition of intrinsic existence (svabhāva) upon empty (śūnya) phenomena. It is striking to find Śāntideva invoking it as having at least provisional benefit for the bodhisattva’s progression along the path. These verses suggest that Śāntideva is less interested in resolving the tension between emptiness, impermanence and the bodhisattva’s multi-life commitment than he is in manipulating both the metaphysical fact of emptiness, and our habitual tendencies towards reification and superimposition, for liberative benefit. Here, Śāntideva invokes delusional self-existence to provide a basis for compassion to arise. Elsewhere, he offers meditations upon the nonexistence of the self (anātman) to dissolve selfishness (BCA 8:101–103) and eliminate anger (BCA 6:22–41). We also find him turn anger against itself (BCA 8:41) and invoke terror of death to provide

515

Stephen Harris

motivational energy to enter into the bodhisattva path (BCA 2:28–66) and develop the perfection of energetic effort (BCA 7:4–13) (Harris 2017). I return to Śāntideva’s strategic employment of the realization of selflessness when I discuss the perfection of wisdom subsequently. In the next two sections, which examine Śāntideva’s development of generosity and patience, I explore more systematically another of Śāntideva’s overarching themes in the BCA, the convergence of altruistic commitment and self-interest as one progresses along the bodhisattva path.

Generosity (dāna) Although generosity is not given its own chapter in the BCA, Śāntideva characterizes it in several important verses dispersed throughout the text. A much more extensive treatment is provided in the ŚS, which I  also draw upon in this section. For Śāntideva, generosity is a mental state which fully renounces all items in one’s possessions; it does not primarily refer to the physical distribution of goods. This is stated most clearly in a verse in the BCA, in which Śāntideva replies to a hypothetical opponent who claims that the existence of poverty implies no previous buddhas or bodhisattvas perfected generosity. If the perfection of generosity makes the world free of poverty, how is it possible that the Protectors of the past acquired it, when the world is still impoverished today? The perfection of generosity is interpreted simply as a state of mind due to the intention of giving away everything, together with the fruits of that, to all people. (BCA 5:9–10) Śāntideva’s response is to explain that the perfection of generosity is a mental quality rather than an act of giving which reduces poverty. It is the state of mind which abandons (tyāga) everything to sentient beings. Giving, therefore, is initially and primarily the elimination of craving. Moreover, the perfection of generosity is total; it is the willingness to give everything to others. In the ŚS, the totality of generosity is indicated by listing three kinds of objects which the bodhisattva renounces: his possessions, his body and mental states and his karmic merit (Goodman 2016, 20; see also Mahoney 2002; Clayton 2006). Together, these items constitute the entire experience of the bodhisattva, all of which is offered to sentient beings. Another distinctive feature of Śāntideva’s characterization of generosity is its self-beneficial nature. This is stated forcefully in the BCA, where generosity is equated with liberation. Abandoning everything is nirvāṇa, and my mind seeks nirvāṇa. If I must abandon it, it is better that I give it to sentient beings. (BCA 3:11, translation modified) This striking verse presents a three-fold equation between abandoning (tyāga), giving (dāna) and nirvāṇa. Linking abandoning with nirvāṇa is relatively natural; nirvāṇa is the cessation of the mental afflictions (kleśas), and these arise in relation to possessions, as a result of attachment (tṛṣṇā/rāga) to them, aversion ( pratigha) to losing them, jealously (īrṣyā) of another person possessing them and so on. A practitioner first abandons possessions by becoming a monastic, then abandons mental afflictions through meditational cultivation (bhāvanā) and 516

Śāntideva

ultimately abandons the false sense of self. Abandoning, in all these senses, is in the deepest interest of the individual; its perfection is individual liberation. Remarkably, in 3:11, Śāntideva extends this equivalence between abandoning and nirvāṇa to the first of the Mahāyāna perfections, generosity (dāna). Giving is, first and foremost, abandoning. One can abandon physical possessions by giving them to others and diminish craving in so doing (Lele 2007). More significantly, as we saw previously, perfect generosity is itself the mental attitude which abandons craving towards possessions, even when they remain within our control. Since giving for Śāntideva is total, encompassing all mental and physical experience, perfect generosity is equivalent to the complete abandonment of craving, which is nirvāṇa. Giving is, therefore, a self-protecting and beneficial virtue. This characterization of the perfection of generosity as the elimination of craving is developed more extensively in the ŚS, where one of Śāntideva’s favorite Mahāyāna sutras to quote is the Questions of Ugra (Ugraparipṛcchā-sūtra). This text is ideal for guidance about generosity, since many of its instructions are directed towards household bodhisattvas who remain legal owners of property even while dedicating themselves to awakening. Such a situation provides the ideal opportunity to consider what attitude a bodhisattva should take towards possessions. For thus, leading merchant, a household bodhisattva should not develop a sense of grasping or a sense of “mine” toward any property, nor clinging, nor addiction, nor the propensity to craving. (Goodman 2016, 23, translation slightly altered) Notice that the household bodhisattva is not instructed to physically abandon or give away his possessions but instead eliminates grasping ( parigraha) and his sense of ownership (mamatvam) towards property. A quotation from the Questions of Nārāyaṇa sutra repeats this advice but also brings into view the connection between generosity as the elimination of grasping, with giving in its ordinary sense of transferring items to those in need: Don’t appropriate anything if you can’t arouse the thought of giving it away. Don’t own any property if you can’t arouse the thought of giving it up. . . . A bodhisattva great being should arouse the following thought: “I  have given away and offered this, my body, to all sentient beings; how much more so any external property! If any sentient being needs anything for any reason, whatever it might be, if it exists, I will give it to that being. I will give my hand to whoever needs a hand. I will give my foot to whoever needs a foot. I will give my eye to whoever needs an eye. I will give my flesh to whoever needs flesh. I will give my blood to whoever needs blood. . . . What need is there even to speak of external property, such as money, grain, gold, silver, gems, jewelry. (Goodman 2016, 25) The bodhisattva is instructed to refrain from acquiring material goods until they are able to relate to them as if they are not the owner and therefore can give them up without hesitation or regret. As we saw in the last quotation, this is accomplished through abandonment (tyāga) of any sense of grasping ( parigraha) or ownership (mamatvam) towards possessions. In the current quotation, these insights are also applied to giving in its ordinary sense of the transfer of property to those in need. Based upon our prior abandonment of any sense of ownership, we can effortlessly give items away. Giving, therefore, emerges as a two-step process: first, mental appropriation (upādāta) and all forms of craving are eradicated, and second, physical property 517

Stephen Harris

is transferred. What is essential to understand about Śāntideva’s account of giving is that it is the first, rather than the second, of these steps that is the perfection of generosity. Certainly, the bodhisattva wants to benefit others through giving away possessions, but for this to be deeply effective, perfect giving must have already taken place, privately, by mentally turning over everything to sentient beings in general. Perfect gifts are made to nobody in particular, and this is why it is most natural to think of them initially and primarily as acts of abandonment. In Śāntideva’s characterization of generosity, we find several distinctive features which reoccur in many of his descriptions of the virtues. First, generosity, and indeed all the perfections, are mental states. Śāntideva interprets the perfection of ethical discipline as the elimination of the attitude which wishes to harm others (BCA 5:11) and defines the perfection of patience as the mind that is free of anger (BCA 5:12). The other perfections, of energetic effort, concentration and wisdom, are all naturally understood as mental. In his bodhisattva manuals, Śāntideva’s primary concern is to develop the state of the bodhisattva’s mind; only once mental training is complete can systematic effort be made to intervene in the lives of others. Second, Śāntideva’s account of generosity is structured to be both self-beneficial and altruistic. Generosity benefits oneself, because it is the abandonment of craving, which is one of the deep sources of suffering. Once generosity as a mental attitude is perfected, however, it is then possible for the bodhisattva to intervene in the world skillfully by gifting possessions. It is this two-fold structure of self-benefit and altruism, repeated in his development of many of the perfections, which explains how Śāntideva can characterize the apparently demanding bodhisattva path as being conducive to one’s own well-being. We will find this same structure again in his characterization of patience, to which we now turn.

Patience (ksānti) The sixth chapter of the BCA is devoted to developing two forms of patience (kṣānti). Early in the chapter, Śāntideva offers meditations to develop patience in the form of the ability to endure great pain without mental distress (BCA 6:12–21). The remainder of the chapter presents meditations which allow the bodhisattva to prevent anger from arising when others harm us.4 To understand Śāntideva’s characterization of these two forms of patience, it helps to keep in mind the relation between the perfections ( pāramitās) and the mental afflictions (kleśas). It is the mental afflictions which both cause one’s own suffering and prevent the bodhisattva from working effectively for the well-being of others. Likewise, it is the perfections which eliminate the mental afflictions and in so doing protect the bodhisattva’s well-being and allow him to work for others’ welfare. In developing meditations to develop both forms of patience, Śāntideva recognizes that certain events, such as physical pain and abuse from others, reliably trigger the arising of the mental affliction of anger (krodha/dveṣa). This stimulates further mental afflictions, creates negative karma, causes us to harm others in retaliation and simply hurts. In these ways, anger increases our own suffering and prevents us from working for the well-being of others. Situations in themselves, however, no matter how severe, do not inevitably cause us to react with anger. Chapter six, therefore, offers meditations to change our attitude towards painful and unpleasant situations as a way of preventing anger from arising (Bommarito 2014). Understanding patience as the mental factor which protects against anger by changing our perspective on difficult situations shows that the two kinds of patience are subtypes of a single mental state. Physical pain and abuse by sentient beings are reliable triggers of anger, and patience intervenes by altering our attitude towards these situations so that anger no longer arises.

518

Śāntideva

Patience as Enduring Pain Śāntideva uses two methods to develop the kind of patience which allows one to endure physical pain without distress. Here he combines the strategy of altering our perspective on painful situations with the recognition that over time we become habituated to, and suffer less from, physical pain. Ultimately these are not separate strategies, however; becoming habituated to a painful sensation without adverse mental response is itself a way of taking an equanimous attitude towards the sensation. This type of patience is vital for the bodhisattva’s development. The bodhisattva path, especially at the later stages, requires endurance of vast amounts of pain as the bodhisattva willingly gives up his body and life to benefit sentient beings (BCA 7:20–22) and takes rebirth in hell realms to work for the beings there (BCA 8:107–108). Moreover, Śāntideva has claimed that the bodhisattva path is conducive to one’s own well-being. Patience in its form of enduring pain, therefore, must be developed to an extremely high degree, which will allow the bodhisattva to endure any grade of physical pain without mental distress. In the verses on becoming habituated to pain, Śāntideva points out that we already frequently experience unpleasant physical sensations without psychological distress. There is nothing whatsoever that remains difficult as one gets used to it. Thus, through habituation with slight pain, even great pain becomes bearable. Do you not consider the pain of bugs, gadflies, and mosquitoes, of thirst and hunger, and the irritation of a serious rash and the like as insignificant? Cold, heat, rain, wind, traveling, illness, captivity, and beatings should not induce a sense of fragility. Otherwise, the distress becomes greater. (BCA 6:14–16) Verse 6:14 begins by suggesting that over time we may become used to painful sensations, and supports this position in 6:15, by reference to irritations like bug bites and skin rashes which many of us endure with ease. Śāntideva then suggests developing these same attitudes to more serious hardships, since adding mental distress to physical pain only increases overall suffering. The second part of the technique for developing the ability to endure pain is to alter one’s perspective on it, recognizing that its badness is minor or even that the experience of pain has positive elements. This is dependent on the recognition that the mental affliction of anger, rather than physical pain, is the real cause of suffering. One of the ways Śāntideva encourages us to change our perspective on painful physical sensations is by considering its relation to our overall goals (Bommarito 2014). He sometimes employs everyday examples outside of religious life to illustrate this strategy, as in this example from chapter four. If fishermen, outcasts, farmers, and others, whose minds are fixed merely on their own livelihoods, withstand the adversities of cold and heat, then why do I not endure for the sake of the well-being of the world? (BCA 4:40)

519

Stephen Harris

This verse combines the two strategies that Śāntideva recommends to develop the ability to endure pain. First, fishermen and other kinds of workers experience discomfort related to their occupations on a daily basis and slowly become accustomed to them. More importantly, however, these discomforts are willingly endured because the workers are intent on a valued goal, here the earning of their livelihood. This changes their perspective on pain, since it is seen as a necessary condition of pursuing the valued item. Of course, the bodhisattva’s pain will eventually be much greater, when he offers his life for others, but the goal of liberating all sentient beings is much greater as well, a fact which Śāntideva reminds his listener of in the last line of the verse. Śāntideva repeats this point in chapter six, but also suggests the possibility that patience can eliminate all adverse reaction to pain: Some, seeing their own blood, show extraordinary valor, while some faint even at the sight of others’ blood. That comes from mental fortitude or from timidity. Therefore, one should become invincible to suffering, and surmount pain. Not even in suffering should a wise person disrupt his mental serenity, for the battle is with the mental afflictions; and in battle pain is easily obtained. (BCA 6:17–19) There are several important points being made in these verses. First, Śāntideva again emphasizes that contemplating the value of a goal will alter our attitude towards physical pain which must be endured in its pursuit. If we like, we can substitute contemporary examples like the running of a marathon for the warriors’ endurance of battlefield suffering. Second, verse 6:18 indicates that it is our mental attitudes of fortitude (dṛḍhatva) or timidity (kātaratva) which determine our ability or inability to withstand pain. Over time, these mental responses can be trained, so that we can endure physically painful situations with tranquility. Verses 6:18 and 6:19 also indicate the radicalness of Śāntideva’s position. With the proper training and perseverance, we can become invincible to any grade of physical pain. The approach so far has been to emphasize that pain is insignificant, at least in relation to the goal of buddhahood. In the following verse, Śāntideva takes this strategy a step further, claiming that pain itself may have positive aspects. Suffering has another quality since arrogance diminishes because of despair, and one feels compassion for beings in the cycle of existence, fear of sin, and a yearning for the Jina. (BCA 6:21) Feeling pain has great value in relation to the bodhisattva’s goals, since it aids in developing compassion, by helping us to empathize with others who have experienced similar kinds of discomfort. Likewise, it stimulates effort (vīrya) by motivating us to deeply commit to the bodhisattva path, which will eliminate all suffering. 520

Śāntideva

The results of these meditations is a reevaluation of physical pain, which is recognized as relatively innocuous and in some cases can be conducive to progress on the bodhisattva’s path. This does not eliminate the physical sensations of pain but alters our perspective on them so that we are less averse to experiencing them. Therefore, anger, which is the real source of disvalue, does not arise. It is worth emphasizing that Śāntideva also instructs the bodhisattva to preserve their bodies in ordinary circumstances as instruments for progressing on the bodhisattva’s path (BCA 5:70). Although eventually the bodhisattva must be able to endure great pain without distress, this should not be interpreted as a senseless masochism in which pain is sought out for its own sake.

Patience Towards Others’ Wrongdoings Śāntideva also employs the strategy of altering our perspective on unpleasant situations when he presents meditations designed to develop patience towards those who harm us. In these meditations, Śāntideva argues that anger is always irrational, both in that it conflicts with our wish to be happy and that it arises from mistaken views about reality (Lele 2007). These verses should be read both as arguments for the badness of anger and as meditations by which to dissolve it. Early in the chapter, Śāntideva emphasizes the mental discomfort which anger causes. The mind does not find peace, nor does it enjoy pleasure and joy, nor does it find sleep or fortitude when the thorn of hatred dwells in the heart. (BCA 6:3) The irrationality of anger is also argued for by pointing out that it is always a needless additional suffering, since resolutions of difficult situations can be sought without anger (BCA 6:10), and that it causes social unrest (BCA 6:4–5). Meditations like these do not depend on accepting Buddhist presuppositions. Other meditations, however, rely on the efficacy of karma, such as Śāntideva’s claim that anger can cause one to be reborn in hell: If one is unable to endure even this slight suffering of the present, then why does one not ward off anger which is the cause of pain in hell? Thus, solely due to anger I have brought myself into hells thousands of times, and I have not brought about benefit for myself or others. (BCA 6:73–74) These verses illustrate a reoccurring feature of the BCA. Śāntideva often presents basic psychological or social observations as a way of beginning to develop virtue and dissolve mental afflictions. These initial insights are then magnified through the incorporation of Buddhist cosmological and karmic presuppositions. Anger should be abandoned because it keeps us awake at night (BCA 6:3) and disrupts social relations (BCA 6:4–5) but even more so because it will cause us to be reborn in hell (BCA 6:73–74). When working with Śāntideva’s texts philosophically and interculturally, therefore, it can be helpful to recognize that arguments depending on cosmological and karmic presuppositions will often be expanding on a central insight which may also be presented in a more 521

Stephen Harris

interculturally available fashion. Nevertheless, we lose something vital if we do not pay attention to the karmic and cosmological aspects of Śāntideva’s work. They illustrate the vastness of his project and the extraordinary degree to which he develops the virtues. Generosity and patience are total; the bodhisattva gives away everything, including his life and body, and endures any amount of physical pain without discomfort. Moreover, he develops compassion so strongly that he remains in saṃsāra forever. We have seen that Śāntideva draws attention to positive aspects of physical pain as a way of developing the bodhisattva’s ability to endure it. He employs the same strategy in developing patience towards sentient beings’ wrongdoing: For a supplicant is not a hindrance to generosity at the time of almsgiving; and when a person who bestows an ordination arrives, he is not called a hindrance to the ordination. Beggars are easy to meet in the world, but malefactors are difficult to find, for no one will wrong me when I do no wrong. Therefore, since my adversary assists me in my bodhisattva way of life, I should long for him like a treasure discovered in the house and acquired without effort. (BCA 6:105–107) The verses point out that the bodhisattva has committed to the extraordinarily valuable goal of developing the perfections and achieving full buddhahood for the benefit of all. This requires developing patience, since patience is one of the perfections, and this in turn requires enemies whose actions give the bodhisattva the opportunity to develop it. Therefore, the attitude towards such persons should be gratitude, rather than anger. All the meditations surveyed thus far can be understood as parts of Śāntideva’s extended argument for the irrationality of anger. Perhaps the most obvious defense of this claim, however, comes in a set of verses in which Śāntideva emphasizes the incompatibility between anger and the dependently originated nature of phenomena. I am not angered at bile and the like even though they cause great suffering. Why be angry at sentient beings, who are also provoked to anger by conditions? (BCA 6:22) Śāntideva’s argument continues for several verses, and there has been scholarly disagreement as to how we should understand it. The general point is that accepting that beings are subject to innumerable causes and conditions entails accepting as well that our enemies are influenced by numerous factors beyond their control. Recognizing this will dissolve our anger towards them, since they are not deeply responsible for their actions. Some contemporary commentators have understood Śāntideva in these verses to be endorsing a kind of incompatibilism in which moral responsibility is determined to be incompatible with dependent origination (Goodman 2002). A  weaker reading is to see Śāntideva as merely drawing attention to the relatively powerless position of our enemies who are dominated by the mental afflictions. Both 522

Śāntideva

readings entail that anger towards our enemies is an irrational response. This in turn suggests that Śāntideva, like the Stoics, held at least a weak cognitivism, in claiming that emotions like anger are highly responsive to our judgments. Moreover, it is a distinctive position of Buddhists and Stoics alike that the emotions are highly malleable and with long-term training can wholly be brought under our control (McRae 2015). There is, however, an obvious contrast between Śāntideva’s position and authors like Aristotle who claim that anger is natural and in the right circumstances is an essential component of a flourishing life. Likewise, some philosophers have emphasized that circumstances of injustice may require anger as an appropriate moral response (Nussbaum 1994, cf. Vernezze 2008). By contrast, Śāntideva claims that anger is always harmful and seems to have little concern for justice. This is largely because anger causes suffering, and like virtually all Indian Buddhist ethicists, Śāntideva places the removal of suffering at the center of his moral project. It suggests as well that at least a certain strand of deontic thought will have little traction with Śāntideva. If moral indignation increases the overall suffering of a particular situation, then it must be resisted. Nevertheless, this should not be mistaken for a moral passivity for the suffering of those in need. Rather, it demonstrates another of Śāntideva’s central commitments, impartial benevolence, manifested in an equal-minded concern for all, where this includes both the victims of aggression and the aggressors who will suffer horribly from the negative consequences of their anger.

Wisdom ( prajñā) and the Relation Between Metaphysics and Ethics Chapter nine of the BCA, dedicated to developing wisdom ( prajñā), provides an extended defense of the core Madhyamaka metaphysical tenet of emptiness (śūnyatā). Although Śāntideva does not explicitly define emptiness in the text, he appears to accept the standard Madhyamaka position that wisdom is the realization that all phenomena are empty (śūnya) of intrinsic existence (svabhāva). This means that entities do not bear their own identity conditions but exist only in dependence upon their causes and conditions and upon conceptual imputation. Most of the chapter presents a series of objections brought by philosophical opponents, including Abhidharma Buddhists (BCA 9:38–53), Yogācāra Buddhists (BCA 9:17–35) and other philosophical schools such as Sāṃkhya (9:60–67) and Nyāya (9:68–69). The chapter is a valuable contribution to the earlier expositions and defenses of Madhyamaka metaphysics by authors such as Nāgārjuna and Candrakīrti. Nevertheless, Śāntideva’s arguments largely support positions already established by these earlier authors, and I am not able to survey them here. What is perhaps more distinctive about Śāntideva’s approach is how he integrates various forms of wisdom, including the realization of emptiness (śūnyatā), selflessness (anātman) and dependent origination ( pratītyasamutpāda), into his development of the other perfections ( pāramitās). We saw one example of this in the last section, in which he develops patience by arguing that dependent origination entails the irrationality of anger. A similar argument occurs in the eighth chapter, in which Śāntideva appeals to the core Buddhist tenet of selflessness (anātman) in arguing that prioritizing one’s own well-being is irrational. This argument has attracted considerable contemporary philosophical interest. It opens by drawing attention to the similarity of all beings in their desire to be happy and free from suffering. One should first earnestly meditate on the equality of oneself and others in this way: “All equally experience suffering and happiness, and I must protect them as I do myself.” (BCA 8:90) 523

Stephen Harris

When happiness is equally dear to others and myself, then what is so special about me that I strive after happiness for myself alone? (BCA 8:95) When fear and suffering are equally abhorrent to others and myself, then what is so special about me that I protect myself but not others? (BCA 8:96) As with the passages in the patience chapter, these verses should be understood both as meditations designed to increase our compassion and as premises in Śāntideva’s argument for impartial benevolence. They act as meditations, since reflecting on the fact that others’ wish for happiness and freedom from suffering stimulates compassion for them. But Śāntideva also challenges his egoistic opponent to provide a distinction (viśeṣa) which justifies giving greater priority to his own well-being. Since the wish for freedom from suffering is the same for all, one’s own desire to be free of it cannot provide the needed justification. The remainder of the argument consists of a series of potential distinctions offered by the egoist to justify his self-interest, each of which is in turn dismissed by Śāntideva. The most powerful of these reasons for egoistic behavior is that this particular pain belongs to me. Śāntideva dismisses this response in a trio of verses which conclude his argument. The continuum of consciousness, like a series, and the aggregation of constituents, like an army and such, are unreal. Since one who experiences suffering does not exist, to whom will that suffering belong? All sufferings are without an owner, without exception. They should be warded off simply because they are suffering. Why is any restriction made in this case? Why should suffering be prevented? Because everyone agrees. If it must be warded off, then all of it must be warded off; and if not, then this goes for oneself as it does for everyone else. (BCA 8:101–103, translation altered) These verses respond to an implied objection that I am justified in giving my own well-being greater emphasis because it is mine. Śāntideva responds by pointing out that since the self does not exist (8:101), suffering does not belong to anyone (8:102). Rather, it is only the impersonal badness of suffering which should motivate its removal. Verse 103 responds to an implied follow-up objection. Since there are no persons, the opponent wonders, why is it worth eliminating anyone’s suffering? Śāntideva responds by pointing to the robust universal agreement that suffering is bad. Arguments must end somewhere, and the self-evident badness of suffering is a sufficient reason for committing to its removal. The argument concludes by claiming that since no justifying distinction has been offered, we should commit to removing the suffering of all. 524

Śāntideva

The argument has generated a great deal of philosophical attention, with some commentators being critical of its success (Williams 1998; Harris 2011; Westerhoff 2015) and others defending it (Pettit 1999; Siderits 2000). One of the objections frequently leveled against Śāntideva is that he has not taken seriously enough the possibility that eliminating the belief in a self makes plausible the nihilistic possibility that we no longer need to care about anyone’s pain (Harris 2011). I  cannot survey these contemporary treatments here. We can, however, notice two features of Śāntideva’s argument that reinforce repeating themes in his text. First, it illustrates the close connection between meditation and reasoning in the BCA; even if the argument fails to convince his interlocuter, meditating on the nonexistence of the self, and contemplating others’ suffering, will respectively lessen selfishness and strengthen compassion. Second, we find in these verses once more the creative tension running through the text between the conventional and ultimate levels of reality. At the conventional level, we meditate on the pain of others and generate compassion as a way to motivate ourselves to progress along the bodhisattva path. At the ultimate level, we recognize that selves do not exist and use this insight to dissolve selfishness. But dipping down to the level of ultimate reality threatens the coherence of the entire path, since it casts doubt on the existence of those whom we vow to save. We have already seen that Śāntideva’s solution to this tension is not to resolve it but to float between ontological levels as a way of stimulating compassion and dissolving craving and aversion (Todd 2013). What is most distinctive in his approach to wisdom, therefore, may not be the long defense of it in chapter nine but his manipulation and employment of it, along with even delusion itself, to develop the perfections and help the bodhisattva make progress along the bodhisattva path.

Conclusion In both India and Tibet, Śāntideva’s presentations of the bodhisattva path have exerted great influence on the development of Mahāyāna Buddhism. Although the BCA is far from exhaustive in its treatment of Mahāyāna doctrine, its organization, and most importantly the vivacious style with which Śāntideva presents core Mahāyāna insights, have made the text an icon of Buddhist literature. The ŚS has had somewhat less historical influence but is now being recognized as a valuable complement to the BCA for understanding Śāntideva’s thought. The last twenty years have seen the academic study of Buddhist ethics develop into a significant subfield of Buddhist studies. The BCA has become one of the key texts of this developing movement. In this introduction, I have emphasized some of the text’s core contributions to moral philosophy. This includes Śāntideva’s consideration of the relation between metaphysical positions, in particular not-self, and the normative conclusion of committing to impartial benevolence. Just as striking is his attitude to wisdom itself, which he integrates into meditations designed to develop the perfections. His account of generosity, best understood through consultation of both BCA and ŚS, is striking in its interpretation of giving as initially and primarily a mental attitude of non-clinging. Likewise, his development of patience is remarkable in its radicalness in that it aims to eliminate all anger and enable the bodhisattva to endure any amount of physical pain without mental discomfort. Finally, the bodhisattva vow itself is extraordinary in its infinite magnitude but also in the tensions it contains, in that bodhisattvas without selves make multi-lifetime vows to liberate empty beings. All of these topics also factor into a reoccurring theme of the text as a whole: Śāntideva’s claim that a perfect and limitless commitment to altruism is the best way to eliminate the mental afflictions which are the source of all suffering. For Śāntideva, altruism and self-interest converge perfectly as one progresses along the bodhisattva path. 525

Stephen Harris

Notes 1 For a partly contrasting view, see Edelglass 2017. 2 A natural question to ask is whether Śāntideva’s ethical commitments are best captured by one or another of the ethical theories developed in Western ethics, such as consequentialism and virtue ethics. Śāntideva’s emphasis on the development of virtue throughout his texts does not of itself settle the issue, given that most ethical theories accept the importance of virtue, with consequentialists emphasizing its ability to maximize good outcomes and deontologists theorizing virtue’s role in enabling keeping one’s commitments. Śāntideva’s deepest goals in the texts are the promotion of a particular consequence, the ending of suffering, as well as the development of the virtues of the bodhisattva. It is less clear whether he takes consequences or virtue as the deeper theoretical commitment in terms of theory of right action, however, or whether this was even an interesting question for him. The most systematic treatment of this issue is provided by Goodman 2009. Goodman takes Śāntideva to be a consequentialist whose central ethical commitment is to promote the development of virtue. Goodman’s interpretation is insightful but also controversial. See Barnhart 2012; Davis 2013; Harris 2015 for responses. See Edelglass 2013 for an accessible introduction to similarities and contrasts between Buddhist and Western moral thought. 3 All translations of the BCA are from Śāntideva 1997. All translations of the ŚS are from Goodman 2016. 4 In the ŚS, Śāntideva describes these as the patience which endures pain (duṣkhādhivāsanakṣānti) and the patience towards others wrongdoing ( parāpakāramarṣanakṣānti), respectively (Goodman 2016, 178). Śāntideva also distinguishes a third kind of patience, which arises from reflecting on dharma (dharmanidhyānakṣānti). However, I agree with Lele (2007, 120) that, at least in the BCA, dharmic patience is treated as a variety of the patience towards others’ wrongdoing. Therefore, I do not distinguish them here.

Bibliography Bommarito, Nicolas. 2014. “Patience and Perspective.” Philosophy East and West 64 (2): 269–86. Barnhart, Michael. 2012. “Theory and Comparison in the Discussion of Buddhist Ethics.” Philosophy East and West 62 (1): 16–43. Chödrön, Pema. 2007. No Time to Lose: A Timely Guide to the Way of the Bodhisattva. Translated by the Padmakara Translation Group. Boston: Shambhala. Clayton, Barbra. 2006. Moral Theory in Śāntideva’s Śikṣāsamuccaya: Cultivating the Fruits of Virtue. London and New York: Routledge. Davis, Gordon. 2013. “Traces of Consequentialism and Non-Consequentialism in Bodhisattva Ethics.” Philosophy East and West 63 (2): 275–305. Edelglass, William. 2013. “Buddhist Ethics and Western Moral Philosophy.” In A Companion to Buddhist Philosophy, edited by Steven M. Emmanuel, 476–90. West Sussex: Wiley-Blackwell. ———. 2017. “Buddhism, Happiness and the Science of Meditation.” In Meditation, Buddhism and Science, edited by David L. McMahan and Erik Braun, 62–63. New York: Oxford University Press. Garfield, Jay L. 2012. “Mindfulness and Ethics: Attention, Virtue and Perfection.” Thai International Journal of Buddhist Studies 3: 1–24. Goodman, Charles. 2002. “Resentment and Reality: Buddhism on Moral Responsibility.” American Philosophical Quarterly 39 (4): 359–72. ———. 2009. Consequences of Compassion: An Interpretation and Defense of Buddhist Ethics. Oxford: Oxford University Press. ———. 2016. The Training Anthology of Śāntideva: A  Translation of the Śikṣāsamuccaya. Oxford: Oxford University Press. Gyatso, Tenzin. 1994. A Flash of Lightning in the Dark of Night: A Guide to the Bodhisattva’s Way of Life. Translated by the Padmakara Translation Group. Boston: Shambhala. Harris, Stephen. 2011. “Does Anātman Rationally Entail Altruism? On Bodhicaryāvatāra 8:101–103.” Journal of Buddhist Ethics 18: 92–123. ———. 2015. “On the Classification of Śāntideva’s Ethics in the Bodhicaryāvatāra.” Philosophy East and West 65 (1): 249–75. ———. 2017. “The Skillful Handling of Poison: Bodhicitta and the Kleśas in Śāntideva’s Bodhicaryāvatāra.” Journal of Indian Philosophy 45: 331–48.

526

Śāntideva ———. 2018. “Promising Across Lives to Save Non-Existent Beings: Identity, Rebirth and the Bodhisattva’s Vow.” Philosophy East and West 68 (2): 386–407. Harrison, Paul. 2007. “The Case of the Vanishing Poet: New Light on Śāntideva and the ŚikṣāSamuccaya.” In Indica et Tibetica: Festschrift für Michael Hahn, edited by Konrad. Klaus and JensUwe Hartmann, 215–48. Zum 65. Vienna: Geburtstag von Freunden und Schülern überreicht. Lele, Amod Jayant. 2007. “Ethical Revaluation in the Thought of Śāntideva.” PhD diss., Harvard University. Mahoney, Richard. 2002. “Of the Progresse of the Bodhisattvamārga in the Śikṣāsamuccaya.” M. A. Thesis, University of Canterbury. McRae, Emily. 2015. “Buddhist Therapies of the Emotions and the Psychology of Moral Improvement.” History of Philosophy Quarterly 32 (2): 101–22. Mrozik, Susanne. 2007. Virtuous Bodies: The Physical Dimensions of Morality in Buddhist Ethics. Oxford: Oxford University Press. Nussbaum, Martha. 1994. The Therapy of Desire. Princeton: Princeton University Press. Pettit, John. 1999. “Paul Williams: Altruism and Reality: Studies in the Philosophy of the Bodhicaryāvatāra.” Journal of Buddhist Ethics 6: 120–37. Saito, Akira. 1993. “A  Study of Akṣayamati (=Śāntideva)’s Bodhisattvacaryāvatāra as Found in the Tibetan Manuscripts from Tun-huang.” Report of the Grant-in-Aid for Scientific Research. Miye University, Japan. Śāntideva. 1997. A Guide to the Bodhisattva Way of Life (Bodhicaryāvatāra). Translated by Vesna Wallace and Alan Wallace. Ithaca, NY: Snow Lion Publications. ———. 2006. The Way of the Bodhisattva. Translated by the Padmakara Translation Group. Boston: Shambhala. ———. 2008. The Bodhicaryāvatāra. Translated by Kate Crosby and Andrew Skilton. Oxford: Oxford University Press. Siderits, Mark. 2000. “The Reality of Altruism: Reconstructing Śāntideva.” Philosophy East & West 50 (3): 412–24. Todd, Warren. 2013. The Ethics of Śaṅkara and Śāntideva: A Selfless Response to an Illusory World. London and New York: Routledge. Vernezze, Peter. 2008. “Moderation or the Middle Way: Two Approaches to Anger.” Philosophy East and West 58 (1): 2–16. Westerhoff, Jan. 2015. “The Connection Between Ontology and Ethics in Madhyamaka Thought.” In Moonpaths: Ethics and Emptiness, edited by the Cowherds. New York: Oxford University Press. Williams, Paul. 1998. Altruism and Reality: Studies in the Philosophy of the Bodhicaryāvatāra. Surrey, UK: Curzon Press.

527

32 PRAJÑĀKARAGUPTA Buddhist Epistemology as the Path to the Wisdom of Non-Duality Shinya Moriyama

Introduction In the Buddhist epistemological tradition that was founded by Dignāga (ca. 480–540) and developed by Dharmakīrti (ca. 600–660), a variety of topics concerning the means of valid cognition ( pramāṇa) were explored extensively, mainly through commentaries on the works of these two authors. Through word-by-word glossing, paraphrasing, and analyzing each stanza and each sentence, commentators clarified the intentions behind Dignāga and Dharmakīrti’s subtle arguments. Besides this work of clarification, commentators have sometimes inserted their original philosophical ideas in between the lines. In this manner, Buddhist epistemology developed through extending and updating the founders’ thought in accordance with each new stage of Indian philosophy. Among these commentators, Prajñākaragupta (ca. mid-eighth century–ninth century) stands out due to his extensive commentary on Dharmakīrti’s Commentary on Epistemology (Pramāṇavārttika), The Ornament of the Commentary on Epistemology (Pramāṇavārtti kālaṅkāra).1 Prajñākaragupta’s commentary covers the chapters on the establishment of the means of valid cognition (chapter II), perception (chapter III), and inference for others (chapter IV) of the Commentary on Epistemology and does not address the chapter on inference for oneself (chapter I), on which Dharmakīrti composed his own commentary. Before Prajñākaragupta, Devendra­buddhi (ca. 630–690) had already composed a commentary ( pañjikā) on Dharmakīrti’s work. But Prajñākaragupta’s commentary was particularly influential on later Buddhist and Indian philosophy. This influence is seen in Ravigupta (ca. first half of ninth century), Jayanta (the Buddhist, ca. tenth century), Jñānaśrīmitra (ca. 980–1040), Yamāri (ca. 1000–1060), and the anonymous author of the Secret of Logic (Tarkarahasya). Prajñākaragupta’s philosophy also received severe criticisms from Nyāya and Jaina philosophers such as Bhāsarvajña (ca. first half of tenth century) and Vidyānandin (ca. second half of tenth century). We have little biographical information on Prajñākaragupta. Although the Tibetan monkscholar Tāranātha (1575–1634) writes that Prajñākaragupta was a lay devotee and lived at the time of King Mahāpāla, no clear evidence for the description exists (Schiefner 1869, 230). According to Theodore Stcherbatsky, Dharmakīrti’s followers can be classified into three groups: the “philological school” represented by Devendrabuddhi and Śākyabuddhi, the “philosophical school” represented by Dharmottara, and the “religious school” represented by 528

DOI: 10.4324/9781351030908-45

Prajñākaragupta

Prajñākaragupta. Stcherbatsky’s classification of “religious school” is motivated by the fact that Prajñākaragupta’s commentary puts stress on the explanation of religious issues discussed in the second chapter of Dharmakīrti’s Commentary on Epistemology (Stcherbatsky 1930–32, 39–47). It is also interesting to note that Prajñākaragupta’s school is distinguished from Dharmottara’s school, because we know that Prajñākaragupta criticized several times an opponent’s view that can most probably be attributed to Dharmottara and that the two philosophers stand in opposition concerning interpretations of Dharmakīrti’s philosophy, specifically in terms of the ontological status of mental image in cognition. In the lineage of Prajñākaragupta, three followers’ names are well known: Ravigupta, Jayanta, and Yamāri, the latter two of whom wrote commentaries on the Ornament of the Commentary on Epistemology. In this chapter, I will introduce Prajñākaragupta’s most prominent arguments related to his ontology, epistemology, and philosophy of religion. First, his ontology presents an original view on causation known as “backward causation.” According to this view, a future entity is taken to play the role of a cause of a present entity, against our commonsense view that cause precedes effect. Second, his epistemology is determined by the definition of existence for which he argues. While Dharmakīrti’s definition of the “purposeful action/causal efficacy” is well known, Prajñākaragupta several times refers to an idealistic view that acknowledges the dependence of the existence of all entities on the mind. Third, with regard to philosophy of religion, his arguments in defense of the Buddha’s omniscience are analyzed from two viewpoints, namely the process for its attainment and the proof of it. Finally, Prajñākaragupta’s philosophy is summarized as a path to the wisdom of non-duality from its opposite, dualistic view of reality.

Backward causation (bhāvikāran avāda) Before Prajñākaragupta, Dharmakīrti established a Buddhist theory of causation based on the idea that everything arises and perishes at each moment. He argues both that there is a continuous chain of causes and effects and that several causes co-operate simultaneously in producing a single effect. For instance, a seed, which, strictly speaking, continuously arises moment to moment, produces a sprout when sunshine, earth, water, and so on cooperate. Since one observes that a certain entity in appropriate conditions produces another entity, and the entity does not arise when the former entity is lacking, one infers the former from the latter, namely a cause from its effect, for example, fire from smoke. However, Dharmakīrti rejected the reverse inference of an effect from its cause, because there is always the possibility of impediments between a cause and its effect. Regarding the reverse inference, Dharmakīrti just argued that one can infer the fitness or possibility ( yogyatā) of an effect’s arising when one observes an entity being surrounded by conditions appropriate to its producing a possible effect. He states: When the fitness of an effect’s arising is inferred from a cause in appropriate complex of conditions, the fitness is called the “essential property” (svabhāva), because it does not depend on anything else. (PV I.7, Steinkellner 1991, 1999, 2013, Teil I, 17) In Dharmakīrti’s system of logic, there are two types of valid reason that can result in a proof. The first type is an essential property of an entity (svabhāvahetu, e.g., being produced for proving impermanence), and the other is an effect (kāryahetu, e.g., smoke for proving fire). Whereas the former is established in proving a property from another property, both of which refer to the same entity, the latter is used in the inference of an entity, which is a cause, from 529

Shinya Moriyama

another entity, which is its effect. Based on this distinction, Dharmakīrti maintains that the reverse inference of an effect from its cause can be described as a kind of inference based on an essential property as logical reason; that is, it is the appropriateness or completeness of a cause that proves another essential property, that is, the fitness for the production of its effect. The two properties (i.e., the cause’s completeness and the fitness for its effect’s arising) indicate one and the same state of a cause that necessarily produces its effect. However, he never argued the reverse inference by relying on an effect as logical reason, even though, as Steinkellner (1999) has pointed out, Dharmakīrti probably would have been interested in explaining the future attainment of Buddhahood or rebirth within his system of logic. It is Prajñākaragupta who goes one step further and claims that one can infer a future life from a present life by using the reverse inference. However, it should be noted that in this case, he does not accept any new type of valid reason such as a cause as logical reason (kāraṇahetu). Instead, by using the words “cause” and “effect” in different manners, he provides a peculiar explanation to interpret the present life as an “effect” resulting from its future “cause.” In doing so, the case is still categorized under the inference using effect as a reason (kāryahetu). The idea is based on the theory of future cause (bhāvikāraṇavāda), or so-called“backward causation,” in which a future entity plays the role of causing a present entity, against our commonsense view that cause precedes effect (Franco 2007, 2015). How, then, is the backward causation justified? On this point, he resorts to an argumentative device well known in Indian philosophy (Cardona 1967–68; Inami 1999) called “following copresence and co-absence” (anvayavyatirekānuvidhāyitva) as the grounds for defining “cause,” which can be paraphrased by a famous formulation of the Buddhist doctrine of dependent origination ( pratītyasamutpāda): “When A is present, B is also present,” and “When A is absent, B is also absent.” Thus, if a particular entity a is established as such a type of A, a is said to be the “cause” of b that belongs to type B. In our usual understanding of causality, one might assume a temporal relation between A and B; yet, the formulation does not imply any temporal order, at least not for Prajñākaragupta, who states: If causation is determined merely from the relation in which an entity’s occurrence is brought forth by another entity’s occurrence (tadbhāvabhāvitā), what does it contradict? What is the use of the temporal sequence ( pūrvaparabhāva) here? (PVA 69.11, v. 440) To be more precise, the formulation “If A, then B” is described in Sanskrit with the terms of a locative case ending (sati). Other commentators, such as Devendrabuddhi and Śākyabuddhi, interpret this locative as denoting a temporal order where the cause precedes its effect. Prajñākaragupta disagrees. Once the necessary connection of two types, A and B, is established, there is no longer any distinction between inferences of “cause to effect” and “effect to cause,” because in both cases, it is equally true that what is to be proved does not exist at the present, regardless of whether it is a past entity or a future one. That is to say, in as much as past and future entities are here considered as conceptual constructions, in the technical sense that they are not actual but fabricated by the mind, both inferences have the same value. If one acknowledges the validity of inference of a past entity that has already perished, similarly, one should also accept the validity of inference of a future entity that has not yet come into existence. Moreover, a folk belief largely shared at that time in India provides an important background to Prajñākaragupta’s idea. For instance, one’s present pleasant feeling is said to be influenced by one’s future happiness. Likewise, when one perceives an omen of death (ariṣṭa) in a person’s body, one believes that it is caused by the person’s future death. Those examples 530

Prajñākaragupta

run counter to our belief that the past determines the present, that there are cases in which we also believe that the future determines the present (PVA 67.30, v. 435, 68.29–31; Franco 2007). Although we know little about its exact impact on later Indian philosophy, there are at least two examples of the legacy of Prajñākaragupta’s argument. The first is Bhāsarvajña, who pointed out that Prajñākaragupta’s backward causation deviates from Dharmakīrti’s view: while the latter argues that the necessary relationship (avinābhāva) between the two is ascertained only after one’s empirical determination of causal relations,2 the former maintains that the causal relation is a priori determined as the necessary connection (Moriyama 1998). The second can be found in a short work by Jitāri (ca. 940–1000), Treatise on Future Cause (Bhāvikāraṇavāda), which has been very recently found and studied by Eli Franco (Franco 2015). In the treatise, Jitāri makes an argument based on backward causation while referring to Prajñākaragupta’s view.

Existence defined by perception Another innovative idea of Prajñākaragupta concerns the definition of existence (sattā/sattvam). While Dharmakīrti’s definition of existence as “purposeful action/causal efficacy” (arthakriyā) is well known (Katsura 1984; Dunne 2004, 272–78), Prajñākaragupta refers several times to another definition, “existence is perception” or “existence is what is perceptible” (upalambhaḥ sattā).” This resonates with George Berkeley’s famous phrase, Esse est percipi. With this famous phrase, the eighteenth-century Irish philosopher aimed at criticizing the commonsense view that presupposes the mind-independent existence of objects. According to Berkeley, the idea is ultimately grounded on an eternal spirit’s perception that acts even while others do not perceive an object, that eternal spirit being none other than God. Similarly, the eighth century Buddhist philosopher aimed to establish an idealistic view that acknowledges the mind-dependent existence of all entities. The interesting point of comparison between the two authors is that, although Prajñākaragupta refers to the theory of existence in several contexts, in one place (PVA 112.1–6), he argues for it by resorting to the Buddha’s omniscience regarding all entities in the past, present, and future. This showcases an interesting affinity between idealism and the argument for the religious authority in both traditions of Christianity and Buddhism (Moriyama 2015, Forthcoming). This idealistic view has its roots in Dharmakīrti’s discussion of the concept of “nonperception” (anupalabdhi) as a logical reason. According to his account, the absence of a perceptible object in a certain place is never directly perceived but inferred, and the inference is based on the “non-perception of an object that is perceptible under appropriate conditions (upalabdhilakṣaṇaprāpta)” (Kellner 2003). In this case, there must be a necessary connection between the logical reason and the target to be proved, that is, between the non-perception and the nonexistence of a perceptible object. Dharmakīrti explains the necessary connection by its contraposition: That is, existence is nothing but either perceptibility (upalabdhi) that is characterized by the real entity’s causal capability, or perception (upalabdhi), namely, a cognition’s occurrence based on the entity’s causal capability. (PVSV 4.9–11, Kellner 2003, 125–28; Steinkellner 2013, Teil I, 11) The Sanskrit term upalabdhi is grammatically analyzed as a derivative from the verb “to perceive” (upa-LABH), and the noun “perception” is analyzed from the perspective of both the object (karman) and the agent (kartṛ). Dharmakīrti plays on both possibilities, first on focusing on the perceptibility of the object, and second on focusing on the occurrence of the perceiving 531

Shinya Moriyama

cognition. In the first use of upalabdhi, which stands for a causal account of perception, a cognition occurs by the causal capability of its object, which has arisen in the moment preceding the perceptive cognition, which implies that there must be a temporal gap between an object’s existence and that which is cognized through its perception. However, the second use of upa­ labdhi suggests that an object’s existence is equated with the occurrence of its perception. In the two definitions of existence, some commentators like Śākyabuddhi takes the first alternative as standard and explains the second one as a metaphorical usage of the term upalabdhi (PVṬ, Vol. Je, 13b3; Steinkellner 2013, II, 50 n. 54). In contrast, Prajñākaragupta rejects the causal account of perception and claims that the object and its cognition are essentially not different.3 He explains as follows: If the term upalabdhi indicates the state of being perceived, the state is exactly the essence of an object. Alternatively, even though the term upalabdhi indicates the perception in the sense of agency, it means that the existence of a perceptible object is conceptually pervaded by its “perception.” And if x is not separable from y, one does not understand that x and y are different, because the difference is understood only in separable ones. Dharmakīrti said therefore that existence is perception because the latter does not deviate from the former. Moreover, the perception of objects does not occur due to perception that is different from the object and that occurs at the same time of the object, but rather due to “perception” characterized by “being perceived,” which is entailed in the object (tadanupraviṣṭopalabhyamānatā). Therefore, existence is nothing but perception. (PVA 633.4–8, Moriyama, forthcoming) Compared to Dharmakīrti’s original argument, Prajñākaragupta’s interpretation emphasizes the inseparability between an object’s existence and its perception by denying the idea that it is possible to separate the two as cause and effect. For him, there is no reason to suppose a gap between an object’s existence and its perception. This is because, he argues, the term “object” really refers to something that is already in the state of being perceived. That is, a mental image of the object necessarily appears in the perceiver’s cognition. In fact, his epistemological position is often called the position that accepts cognition endowed with mental images (sākārajñānavāda), which summarizes well in epistemological terms his claim that things are mind dependent.4 This Indian version of “esse est percipi” is further discussed in Prajñākaragupta’s commentary on verse III.53d (meyaṃ tv ekaṃ svalakṣaṇam) of Dharmakīrti’s Commentary on Epistemology, where Dharmakīrti argues that the two kinds of objects, that is, particulars (svalakṣaṇa) and universals (sāmānyalakṣaṇa), are ultimately reducible to the particular as the object that is capable of purposeful action (arthakriyā). For instance, an aggregate of atoms that forms the so-called “pot” is defined by the action of “holding water,” which is expected when one perceives or infers the pot.5 That is, the existence of a particular entity is determined through one’s act toward the entity after its perception or inference. Since the immediate object of perception and inference is causally connected to the particular that fulfills the purpose expected by the cognizer, Dharmakīrti claims that cognizable objects are reducible to the particular. To this, Prajñākaragupta proposes another, non-causal account that maintains the equality between an object’s existence and its perception at the ultimate level: The existence defined by perception (upalambha) is indeed the ultimate existence. The ultimate existence of entities is expressed in this way: “Existence is nothing but perception.” (PVA 213.22, v. 230) 532

Prajñākaragupta

Here we should pay attention to the term “ultimate,” which he uses in other contexts for qualifying the state of self-awareness or reflexive awareness (svasaṃvedana), in which a cognition not only cognizes an external object but is also reflexively aware of its cognitive state. To be more precise, every cognition arises holding two different aspects, namely the aspect of an object and that of its own grasping cognition. Thus, by focusing on the internal structure of cognition that is endowed with those two aspects, Prajñākaragupta even goes a step farther to argue that ultimately all cognitions are reducible to self-awareness. As he states elsewhere (PVA 31.22–23, Ono 1999, 83.2): “The means of valid cognition is only self-awareness as the single type of perception” (svasaṃvedanam evaikaṃ pratyakṣaṃ pramāṇam). For him, ultimately, there is only self-awareness, distinct from all other cognitions that are established in everyday activity (vyavahāra). According to Prajñākaragupta’s understanding, Dharmakīrti’s account works well in the analysis of those ordinary cognitions, in which an object’s existence should be defined by purposeful action, without introducing the analysis from the viewpoint of self-awareness. In contrast, at the ultimate level, an object’s existence is equated with the appearance of its mental image that is being perceived reflexively in the cognition. This viewpoint can be identified as Prajñākaragupta’s idealist approach to the definition of existence. In this connection, it is also noteworthy that a similar idea to distinguish the two levels of existence emerges again in the epistemological works of Jñānaśrīmitra, a late tenth- to early eleventh-century Buddhist philosopher. He distinguishes the existence defined by illumination ( prakāśa) from that defined by purposeful action or causal efficacy. While the latter presupposes the position that acknowledges external objects, the former is linked with Yogācāra idealism (SSŚ 398.24–399.2; cf. CAP 131.13–25). Of course, because the terminology of “illumination” differs from “perception,” it is still questionable whether Jñānaśrīmitra’s view is inspired by Prajñākaragupta’s. Nevertheless, we should not overlook the structural similarity between the two arguments.

The Buddha’s omniscience as the ultimate means of valid cognition The major contribution of Prajñākaragupta in the field of philosophy of religion is related to his arguments about the Buddha’s omniscience. To explain the point, we should briefly look at the religious philosophical background of Dharmakīrti’s epistemology, which is specifically discussed in the second chapter of the Commentary on Epistemology (Franco 1997). In the period from the late sixth to seventh century, a controversy arose regarding religious autho­ rity between Mīmāṃsā thinkers, who accept the eternal Vedic scripture as their authority; Nyāya-Vaiśeṣika thinkers, who accept a god (Īśvara) as their authority; and Buddhist thinkers, who consider the Buddha the ultimate authority. Kumārila (ca. late sixth century–early seventh century), a representative philosopher of the Mīmāṃsā school, argued against the Buddha’s omniscience. Dharmakīrti, avoiding a direct reply to the criticism, demonstrated the Buddha’s authority based on his practical knowledge on the primary purpose of human beings ( pradhānapuruṣārtha), namely the cessation of the cause of all sufferings and the means for its attainment, that is, the Four Noble Truths. Since the Buddha does not mislead about (avisaṃvādin) the path that he himself has already experienced (svadṛṣṭamārga), and because he presents the Four Noble Truths which are not yet known (ajñātārthaprakāśa) to other non-Buddhist teachers, he is called the “means of valid cognition” whereby people are led to the right cognition on the final purpose of life (Franco 1997, chapters 1 & 2; Krasser 2001; Steinkellner 2003; Moriyama 2014, chap. 1). However, later Buddhist philosophers, including Śāntarakṣita, Kamalaśīla, and Prajñā­ kar­agupta, selected a different way to defend the Buddha’s omniscience from the Mīmāṃsā 533

Shinya Moriyama

criticisms (McClintock 2010; Moriyama 2014). In the case of Prajñākaragupta, his commentary on the second chapter of the Commentary on Epistemology provides several interesting arguments for establishing the Buddha’s omniscience as a reliable source for radically inaccessible matters (atyantaparokṣa), including past and future lives ( paraloka) and the connection between karmic causes and their results, and so on, which remain hidden for us since our conventional means of valid cognition do not reach what lies beyond our senses. He remarks: Therefore our ignorance is removed only by a treatise that has been proclaimed by an omniscient being, but not by anything else. Thus, in this manner, only the words of an omniscient being are a means of valid cognition. Hence, ultimately ( paramārthatas), only the cognition of an omniscient being, but not another, is a means of valid cognition. This is the ultimate meaning of Dharmakīrti’s statement of PV II 5b, “a treatise is for the removal of ignorance.” (PVA 29.26–27, Ono 1999, 77.16–18; Moriyama 2014, 21) Identifying the Buddha’s omniscience and his teaching as the ultimate means of valid cognition, Prajñākaragupta distinguishes it from the conventional (sāṃvyavahārika) ones, namely our ordinary perception and inference. However, it should be noted that even such omniscience is classified as “conventional” in comparison to non-dual cognition (advayajñāna), because the omniscience presupposes entities that are established in our everyday activity based on the division of cause and effect (Iwata 2004; Moriyama 2014). The Mīmāṃsā opponents point out the impossibility for a human being, such as the Buddha, to attain omniscience because of the limitation of human ability. Dharmakīrti already argued that unlike physical ability like jumping, which has the limitation of its training, mental abilities can be developed infinitely through training. For instance, by repeatedly training oneself to arise compassion, one can develop an infinite compassion to the point of extending to all living beings. Following Dharmakīrti, Prajñākaragupta further argues that the same is possible for the Buddha’s meditation for the attainment of omniscience. Since he has compassion, namely the wish for the salvation of all living beings (PVA 53.15–17), he is motivated to undertake the practice for omniscience. Prajñākaragupta describes this kind of practice as the Buddha’s making various kinds of inferences for knowing past and future entities that relate to his own mental stream as well as others’ ones. Following the chain of cause and effect, he infers what has happened and what will happen and finally reaches to the direct intuition of every single phenomena. Therefore, Prajñākaragupta calls one side of the Buddha’s omniscience the “inference of all aspects of entities” (sarvākārānumāna) and another side the immediate cognition (sākṣātkaraṇa) of everything. This would imply that the notion of omniscience breaks the border between perception and inference as defined by Dignāga. However, since the Sanskrit term pratyakṣa (“perception”) etymologically (i.e., by vyutpattinimitta) indicates a “cognition that occurs in accordance with ( prati) each sense faculty (akṣa),” it should be distinguished from inference. To solve the issue, Prajñākaragupta replies that the term akṣa here indicates the mental faculty (manas) as the cause of the immediate cognition, and hence, the term pratyakṣa covers the “inference of all aspects of entities,” a cognition based on the mental faculty as well. Just as one is able to perceive fire from smoke spontaneously after the absolute repetition (atyantābhyāsa) of inference of fire from smoke, the Buddha is also able to perceive everything spontaneously after the absolute repetition of inferences of various kinds of cause and effect. In this manner, the process of how the Buddha has attained omniscience is explained as a combination of inference and perception (Moriyama 2014, Chap. 3).

534

Prajñākaragupta

The Mīmāṃsā opponents also point out the difficulty concerning the proof of an omniscient being. Kumārila already raised the objection that it is impossible in principle to prove that somebody else is an omniscient being unless the person who judges that person is, herself, an omniscient being. In this regard, Prajñākaragupta emphasizes the importance to verify the Buddha’s teaching even partially, to the extent that one is able to examine it (śakyavivecana). Commenting on the first half of verse II.280 of the Commentary on Epistemology, where the Buddha’s knowledge, which is considered true, lasting, and endowed with all qualities (tattvasthirāśeṣaviśeṣajñāna), is said to be justified by its capacity to “protect living beings,” namely the fact that the teaching of the Four Noble Truths can liberate them from suffering, Prajñākaragupta shows how it is possible to prove the Buddha’s omniscience from his teaching (PVA 164.28–12, Wakahara 1985; Moriyama 2014, Chap. 4). Suppose we listen to a teacher’s statement, “All things are perishing at every moment,” one aspect of the Truth of suffering. If we learn Dharmakīrti’s proof of momentariness based on the reason of “existence” as defined by causal efficacy (sattvānumāna), we are able to examine it and conclude that the proposition is true (Steinkellner 1968). This leads us to another assumption – that the teacher’s knowledge of the proposition is either transmitted from another teacher’s teaching or ascertained by himself. Even in the first case, another teacher’s knowledge is assumed to be either originated from a further different teacher’s teaching or ascertained by himself. In this manner, it is inevitable that one needs to finally assume an initial teacher who immediately perceived everything as momentarily perishing and ascertained the proposition established universally. Since the statement, “All things are perishing at every moment,” is a universal proposition, the teacher who initially held it is required to possess the knowledge that the proposition is established without any exception. If there were even one unseen entity for the teacher, the possibility would remain that the proposition is inconclusive. Therefore, in order to explain that the proposition is universally established, there is no other way than to assume that the teacher, the Buddha, is omniscient. In other words, the validity of inference of momentariness, which is verifiable even by our logical thinking, guarantees the high probability of the initial teacher’s omniscience, without which the universal validity of the proposition would not be grounded on all entities of this world. In addition, by applying the same way to other propositions the Buddha taught, it becomes more certain that he is omniscient, and as a consequence, we may have confidence in the Buddha’s words with respect to radically inaccessible matters.

Conclusion: the path toward the wisdom of non-duality Before concluding this chapter, let us turn to the topic of the wisdom of non-duality that is at the core of Prajñākaragupta’s Buddhist philosophy. He first mentions this special cognition in his commentary on Dharmakīrti’s second definition of the means of valid cognition, as that which “illuminates the object that is not yet known to us (ajñātārtha),” and states that this special cognition is the ultimate means of valid cognition. Commenting on the “object that is not yet known to us,” Prajñākaragupta interprets it as meaning the nature of non-duality and states that a means of valid cognition is nothing but that which illuminates non-duality. For him, Dharmakīrti’s system of epistemology is set as a program that leads people from their lower level of knowledge to the highest level of the wisdom. Starting from the position that presupposes the existence of external objects independent from the mind, one gradually attains higher understandings as one’s analysis becomes more and more subtle. For instance, in the view that cognitions are endowed with images that correspond to external objects, further investigation reveals a thorny issue, since there can be no numerical correspondence between

535

Shinya Moriyama

images and external objects. This is because the former are essentially non-divisible, while the latter are divisible into many atoms. Thus, for solving the problem, Dharmakīrti claims in the third chapter of his Commentary on Epistemology that cognition is empty owing to the neitherone-nor-many argument (209–219) and that cognition has the nature of variety-in-non-duality (citrādvaita) (220–222). As Inami (2011) has clarified in detail, Prajñākaragupta goes one step further and provides his own view of this section: “There can never exist any distinction in the ultimate sense, so that the distinctions between singular and multiple, and between variegated and unvariegated, are considered untenable” (Inami 2011, 178). Prajñākaragupta’s fundamental position on the non-duality of cognition can thus shed more light on how his own views on causality, ontology, and religious philosophy are articulated. For instance, his peculiar view of backward causation entails the view that the temporal sequence is fictional. This claim can now be understood as implying that time is seen by Prajñākaragupta from the perspective of non-duality, which abolishes the temporal differences of the past, present, and future. Likewise, the definition of existence as perception is also based on the doctrine of non-duality that denies the subject-object dichotomy. Finally, it is also remarkable that the Buddha’s role loses its authoritative position from the point of view of non-duality, since no distinction between self and other or between teacher and disciples remains. In other words, the Buddha’s role, in particular his omniscience, occupies its place only in a world of duality that is conceptually constructed by ordinary people who sees themselves as separated from and inferior to the Buddha. Overall, Prajñākaragupta’s philosophy seeks to lead from the conventional reality, based on duality, to the ultimate reality. As one’s attachments to the world of duality are gradually abandoned through investigation of the Buddha’s teaching, a new horizon of non-dual cognition opens little by little. The goal that Prajñākaragupta finally aimed to reach is the state of non-duality, which is called the “highest vehicle” (advayaṃ yānam uttamam, PVA 32.15, Ono 1999, 84.14f.). At the very end of The Ornament of the Commentary on Epistemology, Prajñākaragupta writes, “Village girls are longing for ornaments to make the body elegant due to their desire. I composed the Ornament (Alaṅkāra) of Dharmakīrti’s Commentary on Epistemology (Vārttika), but it is not out of my arrogance” (PVA 648.9–10, v. 622). His motivation as a commentator is thus based on the simple wish to clarify Dharmakīrti’s condensed arguments rather than to present his own view. However, his commentary includes in fact numerous original ideas that introduce novel perspectives and are different from Dharmakīrti’s account. Although Prajñākaragupta will say that they are the natural consequences that can be drawn from the depth of his predecessor’s ideas, in our eyes, they are good examples of the fascinating development of Buddhist epistemology through the “Commentary” tradition.

Abbreviations CAP Citrādvaitaprakāśavāda, ed. by A. Thakur, 1975 PV I Pramāṇavārttika, the first chapter (Svārthānumāna), ed. by R. Gnoli, 1960 PV II Pramāṇavārttika, the second chapter (Pramāṇasiddhi), ed. by Y. Miyasaka, 1971/72, and T. Vetter, 1990 PV III Pramāṇavārttika, the third chapter (Pratyakṣa), ed. by H. Tosaki, 1979/1985 PVA Pramāṇavārttikālaṅkāra, ed. by R. Sāṅkṛtyāyana, 1953 PVSV Pramāṇavārttikasvavṛtti, ed. by R. Gnoli, 1960 PVṬ Pramāṇavārttikaṭīkā, sDe dge Tibetan Tripiṭaka, vol. Je. SSŚ Sākārasiddhiśāstra, ed. by A. Thakur, 1987. 536

Prajñākaragupta

Notes 1 For bibliographical information on PVA, see Ono 1999, xi–xxv; Steinkellner and Much 1995, 74f. On Prajñākaragupta’s date, see Franco 2019. Although Sāṅkṛtyāyana’s edition of PVA is still useful, corrections are necessary for a better edition by relying on photocopies of manuscripts in Watanabe 1998 and Tibetan materials, including Ravigupta’s commentary on PV and sub-commentaries by Jayanta and Yamāri. Very recently, a Sanskrit manuscript of Yamāri’s subcommentary has been found and is now being edited by Eli Franco and his research group at the University of Leipzig. Apart from PVA, the Tibetan Tripiṭaka contains a treatise called the Sahopalambhaniyamasiddhi (sDe dge no. 4255/Peking no. 5753) attributed to Prajñākaragupta, though Iwata (1991, 255f.) questions whether it was actually composed by him. 2 Dharmakīrti argues in several different places that causal relation is determined by perception and nonperception. On the empirical process of determination, there was a controversy among his followers. See Kajiyama 1963. 3 In this connection, we know that there were two different opinions about the interpretation of the logical reason of non-perception (anupalabdhihetu): while a party represented by Dharmottara accepts the non-perception of the property that pervades another property (vyāpakānupalabdhi), which presupposes only a conceptual relation between the existence of perceptible objects and their perceptions, another party represented by Prajñākaragupta accepts the non-perception of intrinsic nature (svabhāvānupalabdhi), which presupposes the real identity between existence and perception. See Kellner 1997; Tani 2000, 186–95. 4 Concerning the later controversy around a Yogācāra position that cognition is endowed with images (sākāra) and another position that cognition is not endowed with images (nirākāra), Jñānaśrīmitra evaluates that Prajñākaragupta completed the former position, which was originally intended by Dharmakīrti. See SSŚ 367.18–20; Oki 1975, 94. 5 For the verse and its interpretation, see Franco and Notake 2014, 136–39. For Prajñākaragupta’s interpretation of svalakṣaṇa and its related concept pravartaka, see Kobayashi 2011; McCrea 2011; Miyo 2014. On Prajñākaragupta’s theory of sensory perception, see Yokoyama 2018.

Bibliography Cardona, George. 1967–1968. “Anvaya and Vyatireka in Indian Grammar.” Adyar Library Bulletin 31–32: 313–52. Dunne, John. 2004. Foundations of Dharmakīrti’s Philosophy. Boston: Wisdom Publications. Franco, Eli. 1997. Dharmakīrti on Compassion and Rebirth. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. ———. 2007. “Prajñākaragupta on pratītyasamutpāda and reverse causation.” In Pramāṇakīrtiḥ: Papers dedicated to Ernst Steinkellner on the occasion of his 70th birthday, Part 1, edited by Birgit Kellner, Helmut Krasser, Horst Lasic, Michael T. Much, and Helmut Tauscher, 163–85. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. ———. 2015. “Jitāri on Backward Causation (bhāvikāraṇavāda).” In Buddhist Meditative Praxis: Traditional Teachings & Modern Application, edited by K. L. Dhammajoti, 81–116. Hong Kong: Centre of Buddhist Studies University of Hong Kong. ———. 2019. “Prajñākaragupta.” In Brill’s Encyclopedia of Buddhism, Volume Two, edited by Jonathan A. Silk et al. Leiden: Brill. Franco, Eli, and Miyako Notake. 2014. Dharmakīrti on the Duality of the Object. Pramāṇavārttika III 1–63. Zürich and Berlin: LIT Verlag. Inami, Masahiro. 1999. “On the Determination of Causality.” In Dharmakīrti’s Thought and Its Impact on Indian and Tibetan Philosophy. Proceedings of the Third International Dharmakīrti Conference, Hiroshima, November 4–6, 1997, edited by Shoryu Katsura, 131–54. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. ———. 2011. “Nondual Cognition.” In Religion and Logic in Buddhist Philosophical Analysis. Proceedings of the Fourth International Dharmakīrti Conference, Vienna, August  23–27, 2005, edited by Helmut Krasser et al., 177–96. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Iwata, Takashi. 1991. Sahopalambhaniyama. Stuttgart: Franz Steiner Verlag. ———. 2004. “Prajñākaragupta’s Proof of the Buddha’s Authority.” In Three Mountains and Seven Rivers: Prof. Musashi Tachikawa’s Felicitation Volume, edited by Shoun Hino and Toshihiro Wada, 355–74. New Delhi: Motilal Banarsidass Publishers.

537

Shinya Moriyama Kajiyama, Yuichi. 1963. “Trikapañcakacintā: Development of the Buddhist Theory on the Determination of Causality.” Miscellanea Indologica Kiotiensia 4–5: 1–15. [Reprint in Y. Kajiyama, Studies in Buddhist Philosophy (Selected Papers), edited by Katsumi Mimaki et al., 475–89. Kyoto: Rinsen Book]. Katsura, Shoryu. 1984. “Dharmakīrti’s Theory of Truth.” Journal of Indian Philosophy 12: 215–35. Kellner, Birgit. 1997. “Non-Cognition (anupalabdhi) – Perception or Inference? The Views of Dharmottara and Jñānaśrīmitra.” Tetsugaku 49: 121–34. ———. 2003. “Integrating Negative Knowledge into Pramāṇa Theory: The Development of the Dṛśyānupalabdhi in Dharmakīrti’s Earlier Works.” Journal of Indian Philosophy 31: 121–59. Kobayashi, Hisayasu. 2011. “Prajñākaragupta’s Interpretation of Svalakṣaṇa.” Indogaku Bukkyōgaku Kenkyū 59 (3): 182–87. Krasser, Helmut. 2001. “On Dharmakīrti’s Understanding of Pramāṇabhūta and His Definition of Pramāṇa.” Wiener Zeitschrift für die Kunde Südasiens 45: 173–99. McClintock, Sara. 2010. Omniscience and the Rhetoric of Reason. Śāntarakṣita and Kamalaśīla on Rationality, Argumentation, and Religious Authority. Boston: Wisdom Publications. McCrea, Lawrence. 2011. “Prajñākaragupta on the pramāṇas and their objects.” In Religion and Logic in Buddhist Philosophical Analysis. Proceedings of the Fourth International Dharmakīrti Conference, Vienna, August 23–27, 2005, edited by Helmut Krasser et al., 319–28. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Miyo, Mai. 2014. “Controversy Between Dharmottara and Prajñākaragupta Regarding Pravartaka.” Indogaku Bukkyōgaku Kenkyū 62 (3): 223–28. Moriyama, Shinya. 1998. “Raise no ronshō ni miru Prajñākaragupta no mirai genin setsu.” [“Prajñāka­ ragupta’s Theory of Future Cause in the Proof of Future Lives.”] Indo Tetsugaku Bukkyōgaku Kenkyū 5: 44–57. ———. 2014. Omniscience and Religious Authority. A  Study on Prajñākaragupta’s Pramāṇa­ vārttikālaṅkārabhāṣya ad Pramāṇavārttika II 8–10 and 29–33. Zürich and Berlin: LIT Verlag. ———. 2015. “Prajñākaragupta no chikaku-sonzai setsu” [“Prajñākaragupta’s Theory of Existence Defined by Perception.”] Indo Tetsugaku Bukkyōgaku Kenkyū 22: 175–90. ———. Forthcoming. “Prajñākaragupta on sattvam upalabdhir eva or an Indian version of “esse est percipi.” In Gedenkschrift für Helmut Krasser, edited by Vincent Eltschinger, Chizuko Yoshimizu, et al. Oki, Kazufumi. 1975. “Citrādvaita riron no tenkai – Prajñākaragupta no ronjutsu” [“A Development of the Theory of Citrādvaita: An argument of Prajñākaragupta.”] Tōkai Bukkyō 20: 94–81. Ono, Motoi. 1994. “Prajñākaragupta ni yoru Dharmakīrti no pramāṇa no teigi no kaishaku  – Prajñākaragupta no shinriron” [“Prajñākaragupta’s Interpretation of Dharmakīrti’s Definition of Valid Cognition – Prajñākaragupta’s Theory of Truth.”] Indogaku Bukkyōgaku Kenkyū 42 (2): 885–75. ———. 1999. Prajñākaraguptas Erklärung der Definition gültiger Erkenntnis. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Patil, Parimal. 2009. Against a Hindu God. Buddhist Philosophy of Religion in India. New York: Columbia University Press. Schiefner, Anton. 1869. Tāranātha’s Geschichite des Buddhismus in Iniden. St. Petersburg: Commissionäre der Kaiserlichen Akademie der Wissenschaften. Stcherbatsky, Theodore. 1930/32. Buddhist Logic, 2 vols. Leningrad: The USSR Academy of Sciences. Steinkellner, Ernst. 1968. “Die Entwicklung des kṣaṇikatvānumānam bei Dharmakīrti.” Wiener Zeitschrift für die Kunde Süd- und Ostasiens 12–13: 361–77. ———. 1991. “Dharmakīrti on the Inference of Effect (kārya).” In Papers in Honour of Prof. Dr. Ji Xianlin on the Occasion of His 80th Birthday (II), edited by Li Zheng et al., 711–36. Jiangxi: Jiangxi Renmin Chubanshe. ———. 1999. “Yogic Cognition, Tantric Goal, and Other Methodological Applications of Dharmakīrti’s Kāryānumāna Theorem.” In Dharmakīrti’s Thought and Its Impact on Indian and Tibetan Philosophy. Proceedings of the Third International Dharmakīrti Conference, Hiroshima, November 4–6, 1997, edited by Shoryu Katsura, 349–62. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. ———. 2003. “Once More on Circles.” Journal of Indian Philosophy 31: 323–41. ———. 2013. Dharmakīrtis frühe Logik. Annotierte Übersetzung der logischen Teile von Pramāṇavārttika 1 mit der Vṛtti, 2 vols. Tokyo: The International Institute for Buddhist Studies. Steinkellner, Ernst, and Michael T. Much. 1995. Texte der erkenntnistheoretischen Schule des Buddhismus. Göttingen: Vandenoeck & Ruprecht. Tani, Tadashi. 2000. Setsunametsu no kenkyū. [A Study on Momentariness.] Tokyo: Shunjūsha.

538

Prajñākaragupta Tosaki, Hiromasa. 1979/1985. Bukkyō ninshikiron no kenkyū [A Study on Buddhist Epistemology], 2 vols. Tokyo: Daitō shuppansha. Vetter, Tilmann. 1990. Der Buddha und Seine Lehre in Dharmakīrtis Pramāṇavārttika. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. Wakahara, Yusho. 1985. “Āgama no kachi to zenchisha no sonzai shōmei.” [“Authority of Scripture and the Proof of the Omniscient Being.”] Bukkōgaku Kenkyū 41: 52–78. Watanabe, Shigeaki, ed. 1998. Sanskrit Manuscripts of Prajñākaragupta’s Pramāṇavārttikabhāṣyam, Facsimile Edition. Patna and Narita: Bihar Research Institute/Naritasan Institute for Buddhist Studies. Yokoyama, Akito. 2018. “Prajñākaragupta’s Theory of Sense-Perception (indriyapratyakṣa): With a Focus on the Criticism of the Nyāya-Vaiśeṣika School.” Journal of Indian and Buddhist Studies 66 (3): 148–52.

539

PART 7

Late-Period Commentators (Tenth–Twelfth Century)

LATE-PERIOD COMMENTATORS Introduction to Part 7

The tenth through twelfth centuries represent the start of the waning of the classical period of Buddhist philosophy in India. Yet this period is also marked by tremendous creativity, often in ways that modern scholars are only beginning to appreciate and understand, with Buddhist philosophers propounding new types of syntheses both in the debate halls of the great monastic universities of North India and in scholarly treatises. Tantric practice and thought are increasingly influential, while the epistemological theories of Dignāga and Dharmakīrti continue to hold considerable sway. Madhyamaka and Yogācāra continue to vie for dominance, even as thinkers like Ratnākaraśānti attempt to knit them together, sometimes with the aid of tantric or Buddha-nature theories. Scholars like Jñānaśrīmitra and Ratnakīrti continue to argue with non-Buddhist thinkers, extending conversations on topics including philosophy of language and the warrants for knowledge ( pramāṇas) such as perception and inference, even as practitioners like Saraha (see Part 2) offer an alternative path rejecting such scholastic approaches. Some philosophers travel outside of India, bringing new ideas to other regions and greatly influencing the course of Buddhist thought, as with Atiśa’s promotion of Candrakīrti in Central Tibet. In turn, Tibetans and others visit the monastic universities, seeking knowledge to bring back to their homelands, where Buddhist philosophy would continue to develop long past the time of its demise in India. According to Tibetan sources, Jitāri (tenth century) was associated with tantric traditions and clearly defended Madhyamaka in his doxographical writings. However, as Junjie Chu shows in his chapter, two recently discovered manuscripts of a treatise attributed to Jitāri, Topics of Debate, complicate our understanding of Jitāri’s own philosophical position. Topics of Debate makes possible a more comprehensive understanding of Jitāri’s philosophy than previously available texts because of the breadth of its scope: in seventeen chapters, he focuses on seventeen significant topics of debate among Buddhist, Brahmanical, and Jain philosophers. These chapters typically begin with an argument for Jitāri’s own position followed by refutations of objections and other positions. Chu gives an overview of each of these chapters and concludes that while Jitāri may have been influenced by Madhyamaka thought, he is better understood as a Yogācāra thinker in the tradition of Dharmakīrti. Jñānaśrīmitra (tenth century) belongs to a group of scholars from the great Buddhist university of Vikramaśīla. He also wrote from within the logical and epistemological tradition derived from Dignāga and Dharmakīrti. Jñānaśrīmitra analyzes cognitions into “appearance” DOI: 10.4324/9781351030908-47

543

Late-Period Commentators

( pratibhāsa) and “determination” (adhyavasāya). Any conceptual or nonconceptual awareness, according to Jñānaśrīmitra, involves the appearance of an image in consciousness. “Determination” is the process whereby a phenomenal image is given as presenting something other than itself; this is, Jñānaśrīmitra argues, conceptual awareness. But we are mistaken in taking a cognition that includes an image of something beyond itself as actually giving us an external object. Rather, as Lawrence McCrea makes clear in his chapter, for Jñānaśrīmitra, appearances in consciousness are just that  – appearances in consciousness  – for there is nothing that exists outside of consciousness. Jñānaśrīmitra defends an idealist position that there is only one undifferentiated mental image, what he calls “variegated nonduality” (citraadvaita). He deploys the distinction between appearance and determination to offer creative solutions to a variety of problems in ontology and metaphysics, philosophy of language, epistemology, logic, and philosophy of religion, even as he grounds himself as an exegete of the Dharmakīrtian tradition. Jñānaśrīmitra’s work was influential for much Buddhist philosophy in India and was also engaged by prominent non-Buddhist Indian philosophers for centuries. Ratnakīrti (eleventh century), his student from Vikramaśīla, concisely summarizes, defends, and elaborates on the views of his teacher. Like Jñānaśrīmitra, Ratnakīrti works within the broader Dharmakīrtian tradition of logic and epistemology and addresses a range of issues in Buddhist philosophy. In his chapter, Patrick McAllister explores Ratnakīrti’s ontology with a particular focus on two themes. The first is Ratnakīrti’s position that for anything to be real, it must be momentary, because if something were not momentary, it could not be a cause. The second is Ratnakīrti’s ontological idealism that nothing exists outside of the mind. As McAllister shows, Ratnakīrti devoted considerable attention to making sense of cognition and the Buddhist path from a perspective in which our everyday experience of enduring and external objects is thoroughly erroneous. The Mahāyāna view that ordinary persons perceive conventional reality and take it to be ultimately real, while a buddha does not perceive conventional reality, raises the problem of how a buddha can communicate with and help ordinary persons. According to Ratnākaraśānti (eleventh century), grasping subjects and the objects they grasp are both imagined. They are representational forms that arise in the consciousness of an ordinary person, but a buddha is precisely free from these erroneous representations and perceives reality as it ultimately is. As Gregory Max Seton shows, Ratnākaraśānti makes the controversial argument that, motivated by compassion, a buddha purposefully retains a degree of cognitive error, which enables perception of the external objects and the grasping subject even as a buddha simultaneously knows these representational forms are erroneous. Seton focuses on the account of consciousness and the theory of error Ratnākaraśānti employs as he attempts to develop an ontology and representationalist epistemology that would constitute the correct interpretation of both Nāgārjuna’s Madhyamaka and Asaṅga’s Yogācāra. We can see another kind of syncretic project in the work of Atiśa (982–1054), who integrates Candrakīrti’s and Bhāviveka’s methodological approaches to Nāgārjuna’s Middle Way philosophy, as well as teachings on the bodhisattva path and esoteric tantric practices. While Atiśa primarily follows Candrakīrti’s reductio ad absurdum method of demonstrating contradictions in the views of others, like Bhāviveka, he employs reasoning as an aid to cultivate insight, even as he eschews the approach to valid means of cognition ( pramāṇa) developed by Dignāga and Dharmakīrti. The result, as James B. Apple suggests, is a form of nominalism, according to which cognitions and objects are mere appearances and not ultimately real because they are mental imputations. Even the mind, itself dependently originated, is also a nominal designation, an appearance that cannot be established as inherently existing. Atiśa 544

Introduction to Part 7

calls his approach the “Great Middle Way,” integrating a multiplicity of teachings and practices to support progress on the path to a nondual awakening, ultimately letting go of even nonconceptual wisdom. Abhayākaragupta (twelfth century) was active at both Nālandā and Vikramaśīla, the two most important centers of Buddhist learning in North India. His writings typically consist of passages from earlier Buddhist texts, without naming the author or text, compiled into a coherent whole. Despite what might be regarded as a lack of originality, his ordering of ideas proved influential in both India and Tibet. In his chapter, Kazuo Kanō explores five themes in Abhayākaragupta’s texts: the scriptural authority of Mahāyāna texts, the two truths, the two truths and an evaluation of the mind-only doctrine, the illusory but beneficial mind, and Buddha-nature. According to Abhayākaragupta, the doctrine of the Two Truths is the foundational Buddhist view: all other Buddhist doctrines need to be understood in the framework of the Two Truths. Abhayākaragupta applies this framework to the mind and the achievements of a bodhisattva, which are both beneficial even if they are illusory because they can result in ordinary beings being freed from suffering. Thus, he concludes, even if something exists only conventionally, if it is beneficial, it should be accepted in some form as true.

545

33 JITĀRI A Later Buddhist Master of Debate1 Junjie Chu

Jitāri2 is a renowned Buddhist philosopher who exercised a strong influence on the later period of Indian Buddhist philosophy. He is regarded by Durvekamiśra (c. 970–1030), an important commentator in his period, as his teacher or guru (Dharmottarapradīpa 149.24, Hetubinduṭīkāloka 411.2); the initial formal reasoning of his “proof of the omniscience of the Buddha,” as well as the conclusion of his “proof of cognitive invalidity of the Veda,” are quoted by Ratnakīrti (Ratnakīrti-Nibandhāvaliḥ 31,13–17; 101,16–17), another influential philosopher in the eleventh century. Further, in the Tibetan tradition, Jitāri was said to be the teacher of Atiśa (cf. Roerich 1949, 243), a key person in the second transmission of Buddhism to Tibet in the eleventh century. Unfortunately, we know very little about his life due to the lack of reliable and precise chronological materials. His date can be fixed only conjecturally on the basis of Tibetan historiographical sources. According to studies of these materials by modern scholars (Vidyabhusana 1921, 337f.; Iyengar 1952, viii; Tucci 1956, 249–54, 1971, 249ff.; Shirasaki 1981b, 345–42), the information concerning Jitāri’s date can be summarized as follows: Jitāri’s father lived at the court of King Sanātana of Varendra, a vassal of the Pāla kings. It is said that Jitāri received the title of Paṇḍịta from King Mahāpāla who, according to some sources, reigned until 940 CE. As mentioned, it is also said in the Tibetan sources that Jitāri was the teacher of Atiśa when the latter was a youth. Atiśa was born in 980 and arrived in Tibet in 1042. Comparing these dates, Jitāri is thought to have flourished in the second half of the tenth century (c. 940–1000). The early seventeenth-century Tibetan scholar Tāranātha’s History of Indian Buddhism (rGya gar chos ’byung, GGCB) includes a passage describing Jitāri’s life and conveying the following two significant points: (1) Jitāri is closely connected with the tantric practice of Mañjughoṣa, a tantric form of Mañjuśrī. It is said that Jitāri’s father sNying po’i zhabs (Garbhapāda) performed the tantric initiation of Guhyasamāja for the king Sanātana. For this reason, when Jitāri was sent to a school at the age seven, he was persecuted by other Brahman boys, and was told: “Being a Buddhist tantric practitioner your father gave the low caste (śūdra) queen a higher status and, while worshipping, he mixes the low and high castes without discrimination.” To subdue them, Jitāri’s father decided to bestow the tantric empowerment of Mañjughoṣa on him and let him DOI: 10.4324/9781351030908-48

547

Junjie Chu

practice. Tāranātha sums up Jitāri’s achievement of tantric practice: “He learned whatever his father had mastered – the Guhyasamāja, Cakrasaṃvara, Hevajra, etc. He also attended many other teachers. Uniquely, he was able to listen to all doctrines from Śrī Mañjughoṣa himself ” (cf. GGCB 233,6‒234,17, 234,9–11; Chimpa and Chattopadhyaya 1970, 290–92). (2) With regard to Jitāri’s Mahāyāna scholarship, he is affiliated with Śāntideva’s Madhyamaka teaching tradition. Tāranātha concludes: “He taught a lot on various doctrines and his fame spread widely. He composed short commentaries on the Compendium of Training (Śikṣāsamuccaya), Engaging in Bodhisattva Conduct (Bodhicaryāvatāra), the Essence of Space Sūtra (Ākāśagarbhasūtra), etc. He composed about one hundred treatises on many sūtras and tantras” (GGCB 235,2–235,4; cf. Chimpa and Chattopadhyaya 1970, 292). Sadly, we have no other historical materials to prove the reliability of this description. However, the first point – that is, Jitāri’s association with Mañjughoṣa – is reflected in several introductory verses in his own works.3 For example, the verse of praise (maṅgalācaraṇa) at the beginning of his Topics of Debate (Vādasthānāni), which is also included in the eleventh-century anthology of court poetry known as the Jewel Treasury of Elegant Sayings (Subhāṣitaratnakośa), reads: May he who consecrates his worshiper into the kingdom of his law with anointing liquid, golden red as saffron, poured from the golden amphora, his foot, wherein his lovely toes are ceremonial buds: may Mañjuśrī watch over you for your happiness and good. (Translation in Ingalls 1965, 67; Sanskrit in Kosambi and Gokhale 1957, 6, 9–12.) Concerning the second point  – that is, his relation to the Madhyamaka tradition  – to my knowledge, clear support can be found only in his Verses on the Classification of Buddhist Systems and its commentary (Sugatamatavibhaṅgakārikā and Sugatamatavibhaṅgabhāṣya; the Madhyamaka section is translated in Shirasaki 1986a, 26ff.). This is a doxographic work which follows the pattern of Āryadeva’s Compendium on the Essence of Knowledge (Jñānasārasamuccaya, JSS), explaining the four Buddhist philosophical systems, namely the Sarvāstivāda, Sautrāntika, Yogācāra, and Madhyamaka. As seen previously, Tāranātha mentions that Jitāri wrote commentaries on Śāntideva’s Compendium of Training and his Engaging in Bodhisattva Conduct. According to Ruegg, both Āryadeva and Śāntideva are related to the religio-philosophical movement of the synthesis of Madhyamaka and Vajravāya (Ruegg 1981, 106); that is to say, they are the Madhyamaka philosophers who are influenced by tantric or Vajrayāna theory and practice. In Verses on the Classification of Buddhist Systems verses 7–8 (D 8a2–3), which are almost identical to JSS verses 27–28,4 it is said clearly that the awareness (vijñāna) postulated in the Yogācāra system is not accepted by the Mādhyamika as the ultimate reality because it is devoid of either a single or a multiple essence (ekānekasvabhāva), just like a sky-lotus, and the latter accepts only the reality that is free from the four extremes of existence, nonexistence, both, and neither (Verses on the Classification of Buddhist Systems verses 8a3–4; JSS 26b7–27b1; cf. Mimaki 1976, 189). An extended discussion on this topic can be found in Commentary on the Verses on the Classification of Buddhist Systems (46b5ff ). Based on Verses on the Classification of Buddhist Systems and its commentary, there seems to be no problem with regarding Jitāri as a Mādhyamika. However, disagreements arise with regard to Jitāri’s opinion about the cognitive image of objects (ākāra). In his commentary on Verses on the Classification of Buddhist Systems verse 7, Jitāri criticizes Dharmottara’s position. The controversy is based on their different interpretations of the famous Dharmakīrtian 548

Jitāri

argument that a cognition and its object are not different because they are necessarily perceived together (sahopalambhaniyama). Matsumoto concludes that “Jitāri should be regarded as a Mādhyamika who advocates Satyākāravāda,” that is, the theory that cognition bears the true image of an object (Matsumoto 1981, 966). Shirasaki has a different opinion: he considers Jitāri a Mādhyamika who advocates Nirākāravāda, the theory that cognition does not bear the image of its object (cf. Shirasaki 1986a, 14ff.).5 It is remarkable that the previously mentioned studies are based mainly on Jitāri’s doxographic works and in some cases are influenced by later Tibetan doxographic works as well. However, more convincing conclusions could be made by comparing these doxographic treatises to Jitāri’s own philosophical works. Of Jitāri’s philosophical works, although a considerable number were brought to Tibet, only a few were translated into Tibetan. The following works are included in the Tibetan canon: 1. Ascertainment of Property and the Property-Bearer (Dharmadharmiviniścaya = Chos dang chos can gtan la dbab pa, D4262; cf. the synopsis in Kyuma 2003). In the initial verse, the topics of the treatise are summarized. The distinction between property and the propertybearer is viewed from five vantage points: (1) the natures, (2) the different aspects, (3) the categorical divisions, (4) the particular determinations, and (5) the etymological interpretations. 2. Logic Introduced to Unlearned Persons (Bālāvatāratarka = Byis pa ’jug pa’i rtog ge zhes bya ba, D4263; critically edited in Shirasaki 1981a, 32–52). This is a treatise on epistemology for beginners which follows Dharmakīrti’s Ascertainment of the Means of Valid Cognition (Pramāṇaviniścaya) and Drop of Reason (Nyāyabindu). It is divided into three chapters: perception ( pratyakṣa), inference for oneself (svārthānumāna), and inference for others ( parārthānumāna). 3. Instruction on the True Nature of a Reason (Hetutattvopadeśa = gTan tshigs kyi de kho na nyid bstan pa, D4261; the Sanskrit edition  =  Tucci 1956, 261–74; Sanskrit-Tibetan and Japanese indices are available in Miyasaka 1964). As Tucci points out in his introduction, in this treatise, Jitāri follows closely the Entry to Reason (Nyāyapraveśa), a work of Śaṅkarasvāmin, the disciple of Dignāga. It starts with the introductory verse of the Entry to Reason and it follows the content and structure of the latter very closely, even repeating verbatim many sentences from the latter. This shows that Dignāga’s logic had not been completely suppressed by that of Dharmakīrti even as late as the tenth or early eleventh century. It is worth mentioning that, concerning the content, text structure, and argumentative style, these works are very different from his Topics of Debate, to which we now turn. As has been reported elsewhere,6 the present writer now has access to two newfound manuscripts (hereafter named “A” and “B”) of a philosophical work attributed to Jitāri that had been brought to Tibet. The work contains twenty-one chapters, of which six have previously been edited and published by other scholars; the remaining fifteen are completely new.7 At the beginning of the work, after the salutation to the Buddha and the verse of praise already cited above, Jitāri prefaces his work in the initial verse as follows: In compliance with the wish of friends, putting my shyness aside, some topics of debate (vādasthānāni) [between Buddhists, Brahmans, and Jainas] are written here according to my ability, according to my recollection. It would thus seem that Topics of Debate was the title of the collection as a whole. However, titles of philosophical works in Sanskrit do not usually appear in plural form, and the term 549

Junjie Chu

might also have been used merely as a description of the content of the work.8 Since there is no general colophon at the end of the work, no certainty on this matter can be arrived at. Thus, for lack of anything better, Vādasthānāni or Topics of Debate will be used as the title of the work.9 Although no general colophon is present, the end of each chapter does include a colophon containing the name of the author and the title of the treatise. Needless to say, colophons cannot always be relied on to convey the original title of a work. In the case under discussion, it is clear that they were not written by Jitāri himself, whose name appears in honorific forms such as jitāripāda and mahāpaṇḍitaśrījitāripāda (always in the plural), which he was highly unlikely to have used to refer to himself. Furthermore, if we take titles appearing in the colophons as the titles of each chapter, there would be two cases where a title was doubly used for two different texts; this suggests also that the colophons were added by different scribes later on. In addition to the titles appearing in the colophons, however, a list of the chapters is written on the cover page of manuscript A. Unfortunately, because of poor legibility, only seventeen titles can be clearly identified. In comparison with the titles that appear in the colophons, those contained in the list have one obvious advantage: they describe the contents or the focus of each chapter more precisely, and there is no repetition of titles. For these reasons, in what follows, I use the list to identify each chapter, supplementing the illegible part of the list with the titles found in the colophons. Topics of Debate is now considered a single work containing multiple chapters, since Jitāri gives them a common title by composing the initial verse for all of them. However, the chapter divisions were not made beforehand – as a matter of fact, these chapters are independent in content from one another, exhibiting a variety of topics or perspectives. It is clear that they also circulated singly and in various combinations. Although Jitāri gives regular indications of the changes of subject matter with an introductory sentence usually placed at the beginning each chapter – something like “Here, at the beginning, the theory of the generic property is first refuted” (A1b2) and “Now, the nonexistence of the Self is proved” (A11b3) – he does not seem to make any reference to any formal division of his work. All this strengthens the hypothesis that we are indeed dealing here with a set of single compositions which have been collected together later for convenience. In Topics of Debate, Jitāri addresses a wide range of topics, most of which have been subjects of long-lasting debates among Buddhist, Brahmanical, and Jaina philosophers, with occasional discussions originating from inside the Buddhist system. He generally starts his discussions with a statement of formal reasoning ( prayoga), which forms the main argument, or the main topic, of the chapter. This is usually followed by the refutation of various objections that claim that the reason Jitāri uses in his reasoning is invalid in one of three usually recognized ways: by being “unestablished” (asiddha), “inconclusive” (anaikāntika), or “contradictory” (viruddha). In what follows, I list twenty of the twenty-one chapters10 according to the order they appear in manuscript A. Each title is accompanied by a short description of the main topic of the work, based chiefly on the available formal reasoning at the beginning of the chapter. Because the critical edition of the most of these chapters is not yet complete, and due also to space limitations, a more precise or comprehensive description is not possible. Here is a list of the contents of Jitāri’s Topics of Debate: 1. Negation of the Existence of Universals (Jātyādiniṣedha; A1b1–5b6, B1b1–6b3 = Jātinirākṛti in Tucci 1971, 251–54; Iyengar 1952, 72–80; Bühnemann 1985, 30–38; cf. Shirasaki 1983). The generic property is unreal because it cannot be said to be different from or identical to the individual. All things that cannot be said to be different from or identical to the individual are not real, just like a sky-lotus. 550

Jitāri

2. Refutation of the Existence of Universals (Sāmānyanirākṛti; A6a1–8a5, B6b3–10a1). The universal cannot be treated as existent in its related particulars, because it is not thus perceived. When one thing is not perceived to be existent in another thing, even when the conditions for its perception are fulfilled, this thing cannot be treated as existent in that thing, just like a horn on the head of a horse. 3. Refutation of the Existence of God (Īśvaranirākaraṇa; A8a5–11b3 = Bühnemann 1985, 39–43;11 cf. Shirasaki 1995). The refutation is centered around the formal reasoning used by the opponent to prove the existence of God: All products are perceived to be inseparable from their producer, just like a pot. And products in the world such as a body, and so on, have also an creator, namely God. The refutation comes to the conclusion that the world’s diversity is caused by karma and does not have a single universal producer. 4. Proof of the Nonexistence of the Self (Nairātmyasiddhi; A11b3–12a3, B10a1–10b3 = Bühnemann 1985, 29). A  living body cannot possess a Self, because it does not have any connection with a Self. Anything that is not connected with another thing cannot possess that thing through the connection, just like the Himalaya mountains with the Malaya mountains. 5. On the Cognitive Invalidity of the Veda (Vedāprāmāṇya; A12a3–14b4 = Vedāprāmāṇyasiddhi in Bühnemann 1985, 23–26; quoted in its entirety in Ratnakīrti-Nibandhāvaliḥ 99,16– 101,17, as identified by Bühnemann 1985, 13; cf. Eltschinger 2003). The Vedic verbal testimonies are not valid cognitions concerning external objects, because they have no connection with them. Anything that has no connection with another thing cannot be a valid cognition for that thing, just like an ass cannot be the valid cognition for fire. 6. Proof of the Existence of Mere Cognitive Representation (Vijñaptimātratāsasiddhi; A14b4–20a6, B49a1–55b2, cf. Chu 2020). In contrast to his Verses on the Classification of Buddhist Systems and its Commentary, in this chapter, Jitāri aims to prove the Yogācāra theory that all things that become manifest or appear in cognition are only cognition’s self-manifestation. He does so by refuting the theory of existence of external object (bahirarthavāda), which is divided into two branches: one accepts that cognition possesses an image of the external object, and the other does not accept that cognition possesses an image of the external object. The main target of his refutation is the second one, that is, the theory that cognition does not possess an image (nirākārajñānavāda). His formal reasoning runs: Anything that becomes manifest in cognition is cognition itself, just as the conceptual construction of a blue thing. And a visible thing, and so on, becomes manifest in cognition; therefore, this is an instance of a cognition with the image of objects. In the context of countering the attack that the reason is “inconclusive,” Jitāri refutes the opponent’s thesis that the manifestation of the external object in cognition is caused by cognition. He does so by using an argument in the form of “four alternative propositions,” that is, negating all four logically possible propositions with regard to the relationship between the manifestation and the object – difference, non-difference, both, and neither – with the conclusion that the manifestation of the object is merely the cognition’s self-manifestation. And thus, the reason, that is, “becoming manifest,” cannot be inconclusive, since the manifestation never occurs in any other way, and the thesis that is ascertained through the reason free from all fallacies, each of which is proved not to be existent, should be accepted. 7. Negation of the Naiyāyika’s Concept of the Whole Consisting in Parts of a Gross Object (Avayaviniṣedha; A20a6–24b6, B41a1–46b1). All seen things like a pot cannot be regarded as being single, because they are involved with contradictory properties; whatever is involved with contradictory properties cannot be a single thing, just like a variegated cloth, and so on. 551

Junjie Chu

  8. Proof of the Theory of Exclusion (Apohasiddhi; A24b6–32b1, B32a1–40b3). Every expression (vācaka) in verbal form such as “pot” refers to a superimposed single excluded thing, because every expression takes a superimposed single excluded thing as its sphere of reference, just like a nonexistent verbal form.   9. On the Non-Perception of the Pervader (Vyāpakānupalambha; A32b1–46a1). At the beginning, Jitāri says that he composes the treatise for the purpose of examining the proper nature of the means of cognition that invalidates the occurrence of the reason “being existent” in the opposite of what is to be proved (viparyayabādhakapramāṇa), which is applied in the proof of the Buddhist doctrine of the momentariness.12 According to Jitāri, this should not refer to the perception of the contrary of the pervader (vyāpakaviruddhopalambha); rather, it should be understood as the non-perception of the pervader (vyāpakānupalambha), which is in line with Dharmakīrti’s intention. 10. Negating the Superiority of Brahmanhood (Brāhmaṇyaniṣedha; A46a1–57b4). Jitāri questions the opponent who assumes to be a twice-born (i.e., a Brahman or other member of the first three classes) about his high self-esteem, which is claimed to be connected to the property not in common with the Śūdra, and so on. Jitāri negates the existence of such a distinctive property; he lists all possible interpretations of this property (i.e., the birth, the activity, the family, or the special capacity), then he negates each of them. For example, he negates the special birth with the formal reasoning: That which is not known through a means of valid cognition cannot be treated as existent, just like a sky-lotus. The birth uncommon to Śūdra, and so on, is not known through the means of valid cognition in those who assume to be twice-born. He draws the conclusion that the designations of the four traditional classes (varṇa) of Brāhmaṇa, Kṣatriya, Vaiśya, or Śūdra are only different in names, which are devoid of connection to any real substance, and one should not adhere to them merely due to the merely nominal difference. 11. Examination of the Theory of Non-Momentariness (Akṣaṇikavādavicāra; A57b4–62b2, B18a – 22b). The proponents of non-momentariness are divided into two groups according to whether they regard things as permanent or not. The first group holds that the permanent things exist without a cause, like ether; they can never and in no way come to a cessation of their proper nature. The second group maintains that things exist with a cause and have an impermanent nature, so they come necessarily to destruction at any time in any manner, but they are not momentary, because when they would cease to attain uninterruptedly their proper nature, everything would thus become momentary. These two are refuted in detail. 12. Proof of the Omniscience of the Buddha (Sarvajñasiddhi; A62b2–64a4, B11b  – 117b3 = Bühnemann 1985, 27–28). The Buddha’s speech, like “all produced things are momentary,” is a valid cognition that is reliable and has an ascertained objects because every speech that is non-belying and has an ascertained object presupposes directly or indirectly the knowledge that apprehends intuitively its object, like the expression “fire burns.” Note that here, “omniscient” refers to “knowing all that is useful” but not “knowing everything whatever” (cf. McClintock 2010, 133 and n. 338; Moriyama 2011, 2014, 84–86). In this chapter, Jitāri proves that the Buddha’s teachings which are useful for liberation are reliable. 13. On the Theory of Backward Causation (Bhāvikāraṇavāda; A64a4–69b3, Franco 2015, 90–99). Whatever has a positive and negative concomitance followed by another thing is the cause of that thing, just like fire is the cause of smoke. And every future effect has necessarily positive and negative concomitances followed by its cause, so this future effect is the cause of that cause. Here, Jitāri emphasizes that the determining factor of a causal 552

Jitāri

relationship consists in the positive and negative concomitance, not other factors, such as the temporal sequence of their manifestation. 14. On the Theory of the Generic Property or Universal (Jātivāda; A69b3–70b3). The existence of the generic property cannot be proved, because it is not capable of producing an effect. Everything that is not capable of producing an effect cannot be proved to be existent, even by taking existence as its basis (sattāśrayatvena), just like a sky-lotus. 15. Proof of the Human Authorship of the Vedic Injunction (Śrutikartṛsiddhi; A70b3–77b4, B26b – 131b3). Vedic statements, such as “a person who wants to go to heaven should scarify the fire-offer,” are human products. Every verbal statement is a human product, just like a verbal statement of a passerby. 16. On the Invalidity of Vedic Verbal Testimony (Śabdāprāmāṇya; A77b5–85b4). Verbal testimonies can be a valid cognition only because they lead to attain the object of the statement, not due to being related to the mere intention of the speakers . . . (the remaining part of the sentence is illegible). 17. Destruction of the Causal Complex (Sāmagrībhaṅga; A85b5–87b1). The opponent holds that object and its cognition have a subject-object relationship on account of their being commonly dependent on the same causal complex (ekasāmagrī), that is, the sense faculties. This is refuted on the basis of the nonexistence of such a complex. According to Jitāri, a thing that is whole is called “complex,” but the sense faculty cannot be the whole cause, because this can neither be proved by perception, since it is supersensory, nor by inference, since inference is preconditioned by perception. 18. Introduction to Momentariness (Kṣaṇabaṅgopanyāsa; A87b1–90a5). All things, such as visible things, visual senses, and so on, are momentary because they are existent; when a thing is not momentary, it is not existent, just like in the case of a sky-lotus. 19. Proof of Momentariness (Kṣaṇabhaṅgasiddhi; A90b1–93b4). The opinions of Arcaṭa and Dharmottara are reported, and they are labeled advocates of the “internal pervasion” (antarvyāptivādin) and the “external pervasion” (bahirvyāptivādin), respectively. These are the two theories about the nature of the pervasion of the reason “existence” by the property to be proved “momentariness.” In the case of the “internal pervasion,” the pervasion is apprehended in the subject of the inference itself without reference to an external case, so no example is necessary, whereas in the case of the “external pervasion,” the pervasion is apprehended outside of the subject of inference with reference to an external case, that is, something other than the subject that is usually indicated with an example.13 Arcaṭa’s reasoning runs: Whatever is existent is momentary, and a visible thing, visual sense, and its cognition, and so on, are existent; therefore they are momentary. In this reasoning, the pervasion is apprehended within its subject “every existent thing,” so no other example is given. Furthermore, according to this theory, the invalidity of the reason is excluded by the means of valid cognition, which invalidates the occurrence of the reason in the opposite to the thesis to be proved, as it is formulated as follows: A thing that does not have the capacity to produce an effect in a successive or non-successive way is not capable of producing an effect at all, just like a sky-lotus. And a non-momentary thing is not capable of producing an effect in a successive or non-successive way, so it is not capable of producing any effect and thus is not existent, too. This is because the nature of being existent is equal to the nature of being capable of producing an effect, and this capacity is pervaded by successiveness or simultaneousness in producing an effect. On the other hand, Dharmottara’s reasoning is as follows: Whatever is existent is momentary, just like a pot. And these pots, and so on, are existent, so they are momentary. Here, in 553

Junjie Chu

contrast to Arcaṭa’s reasoning, an example is given. But the necessity of the example in this reasoning is questioned; a detailed discussion is provided. 20. Refutation of Non-Absolutism (Digambaramataparīkṣā; A93b4–97b5, B23a  – 125b3  = Anekāntavādanirāsa in Iyengar 1952, 80–85; cf. Shirasaki 1974; Toshiaki 1978). In the Jaina system, it is maintained that the substance (dravya) keeps its intrinsic nature, while the modes ( paryāya) of that substance can be different, just like gold keeps its essential nature, while its modes can be various gold ornaments. Jitāri refutes this by pointing out the incompatibility between a single substance and multiple modes. In all these chapters, Jitāri defends the most important Buddhist doctrines and refutes various non-Buddhist theories by following Dharmakīrtian epistemologico-logical theoretical patterns and argumentative methods. Although Jitāri does not seem to offer many innovative ideas or clearly independent theories, the Topics of Debate should still occupy an important place in the history of post-Dharmakīrtian Yogācāra philosophical literature, as it summarizes or discusses from a new perspective the most important topics that later Indian Buddhists were debating with various non-Buddhist systems. The main target of his criticisms should be the Mīmāṃsaka, and he mentions Kumārila by name many times. There are also other cases where he has as his opponents different Buddhist interpreters of certain theories. This enriches significantly our knowledge about later development of the Yogācāra system and its interaction with various non-Buddhist or Buddhist opponents. With regard to Jitāri’s philosophical position in the Topics of Debate, in my view, he is a Yogācāra who follows Dharmakīrti’s system, although he may be influenced by the Madhyamaka system to some extent. In many chapters, his Yogācāra background is quite clear. For instance, in chapter 6, Proof of the Existence of Mere Cognitive Representation, he proves that all manifestations of external objects are merely the self-manifestation of cognition. So, in this chapter, it is more natural to say that he establishes the fundamental theory of the Vijñānavāda or the Yogācāra system that the mere cognitive representation (vijñapti) or awareness (vijñāna) exists rather than that he negates it from the Madhyamaka point of view. That task is reserved for his Verses on the Classification of Buddhist Systems and its commentary, where he says unmistakably that the “awareness” regarded in the Yogācāra system as the ultimate reality is not accepted from the Madhyamaka point of view. As for his opinion about the cognitive image of an object, it is also clear that he maintains the theory that awareness has the image of the object (sākāravāda) on the basis of the Vijñānavāda or the Yogācāra system rather than supporting the theory that cognition does not have the image of object (nirākāra) from the Madhyamaka point of view, as he is interpreted to hold in his Verses on the Classification of Buddhist Systems and its Commentary. In any case, his position in the Topics of Debate is different from that in the Verses on the Classification of Buddhist Systems and its Commentary. According to my present knowledge, in the former, no statement can be found in agreement with those in the latter. Indeed, the argument “being devoid of the single or multiple essence” used by Śāntarakṣita in his Madhyamakālaṃkāra 1, which is often regarded as the indication of author’s Madhyamaka position, is not found to be used in the Topics of Debate, and “four alternatives” is often used to negate various non-Buddhist theories, but it is not found to be used against the Yogācāra. If we do not assume that there are two philosophers who bear the same name, we have to consider the possibility that Jitāri articulates his different philosophical positions in his two works. In his philosophical work Topics of Debate, he refutes various mainly non-Buddhist theories from the Yogācāra point of view, and in his doxographical work Verses on the Classification of Buddhist Systems and its Commentary, he refutes Yogācāra theory from the Madhyamaka position. In this reading, we can say that he speaks differently to different audiences. Another 554

Jitāri

possibility, however, is that these two works were composed in different periods and that in between these two periods, his thought underwent a radical change from the one philosophical perspective to the other. Since the edition and the philosophico-historical studies on the Topics of Debate are still in the early stage, no decisive conclusion can be offered.

Notes 1 I am indebted to Dr. Sara L. McClintock, whose valuable comments and suggestions improved my English and made the presentation more clear and precise. 2 Regarding the Sanskrit form of his name, although Tibetan sources suggest something like Jetāri, on the basis of colophons in the published texts as well as in the newfound manuscripts, the correct form must be Jitāri, with dGra las rnam rgyal or dGra las rgyal ba as the Tibetan translation. 3 Additionally, a large number of tantric works are attributed to him (cf. Shirasaki 1981b, 336–28). However, these are beyond the scope of the present study. 4 The Sanskrit verses of JSS 27 and 28 are quoted in Mokṣākaragupta’s Language of Logic (Tarkabhāṣā; 70,4–5) and in Prajñākaramati’s Commentary on Engaging in Bodhisattva Conduct (Bodhicaryāvatārapañjikā, 174,11–12) respectively. See Mimaki 1976, 188. 5 Another related topic discussed is how Jitāri regards Dharmakīrti’s thought. According to Ruegg 1981, 100; Shirasaki 1986b, Jitāri regards Dharmakīrti as a Mādhyamika master. This conclusion gives rise to an interesting discussion in Steinkellner 1990, where the author points out that there is not sufficient evidence to prove Dharmakīrti a Mādhyamika, and the source for Jitāri’s view is Śāntarakṣita’s Commentary on the Ornament of the Middle (Madhyamakālaṃkāravṛtti; Steinkellner 1990, 82f.). 6 See Chu and Franco 2016. In the following, some discussions related to the manuscripts are based on that article, with some information updated. 7 According to Kellner et  al., there is a third manuscript of Jitāri’s Proof of the Existence of Mere Cognitive Representation (Vijñaptimātratāsiddhi; see no. 6), and there is also a copy of the Deliberation on the Rule of the Necessary Co-Perception of a Cognition and Its Object (Sahopalambhaniyamasamarthana) which is also ascribed to Jitāri. See “Preface” in Kellner et al. 2020, xiv. However, these are not available to the present author. 8 See Dignāga’s initial verse in Pramāṇasamuccaya 1.1ab: “for the purpose of establishing the valid means of cognition, from my earlier ideas scattered in various treatises, here a single collection (samuccaya) is composed.” 9 Note that in Iyengar 1952, “Vādasthānāni” is also used as the title of Jitāri’s work which contains only two sections: Jātinirākṛti and Anekāntavādanirāsa. 10 The last chapter, Negation of Ungrammatical Language (Apaśabdanirākṛti; A112b2–7  =  B47a1– 48b5), is not included because the legibility in both manuscripts is very poor, and for the time being, I am unable to offer an understandable transliteration. 11 In the manuscript Bühnemann uses, the title is missing; a tentative title “*Īśvaravādimataparīkṣā” is suggested, see Bühnemann 1985, 19. 12 For the technical term and the related topics, see Steinkellner 1991, 314ff.; Shiga 2011; Sakai 2019, 2020. 13 For more detailed explanations of these two terms, “internal pervasion” and the “external pervasion,” see Mookerjee 1935, 398–400; Kajiyama 1958, 362.

Bibliography Āryadeva. Tibetan translation of Jñānasārasamuccaya (Āryadeva), see D 3851=P 5251. Bühnemann, Gudrun, ed. 1985. Jitāri: Kleine Texte beschrieben und ediert. 2nd expanded ed. Wien: Arbeitskreis für tibetische und buddhistische Studien, Universität Wien. Chimpa, Lama, and Chattopadhyaya, Alaka, trans. 1970. Taranatha’s History of Buddhism in India. New Delhi: Motilal Banarsidass. Chu, Junjie. 2020. “Jitāri’s Vijñaptimātratāsiddhi.” In Reverberations of Dharmakīrti’s Philosophy, Proceedings of the Fifth International Dharmakīrti Conference Heidelberg, August 26 to 30, 2014, edited by Birgit Kellner, Patrick McAllister, Horst Lasic, and Sara McClintock, 1–19. Vienna: Austrian Academy of Sciences Press.

555

Junjie Chu Chu, Junjie, and Eli Franco. 2016. “Rare Manuscripts of Works by Jitāri.” In Sanskrit Manuscripts in China II, edited by Xuezhu Li, Horst Lasic, Xuezhu Li, and Horst Lasic, 15–48. Beijing: China Tibetology Publishing House. Daisetz, T. Suzuki, ed. 1957. The Tibetan Tripitaka, Peking Edition – Kept in the Library of the Otani University, Kyoto. – Reprinted under the Supervision of the Otani University. Tokyo-Kyoto: Tibetan Tripitaka Research Institute. Eltschinger, Vincent. 2003. “La Vedāprāmāṇyasiddhi de Jitāri. Introduction et traduction.” Journal Asiatique 291 (1–2): 137–72. Franco, Eli. 2015. “Jitāri on Backward Causation (bhāvikāraṇavāda).” In Buddhist Meditative Praxis: Traditional Teachings and Modern Application, edited by K. L. Dhammajoti, 81–116. Hong Kong: Centre of Buddhist Studies. Ingalls, Daniel H. H., ed. 1965. An Anthology of Sanskrit Court Poetry, Vidyākara’s “Subhāṣitaratnakośa.” Cambridge, MA: Harvard University Press. Iyengar, H. R., ed. 1952. Tarkabhāṣa and Vādasthāna of Mokṣākaragupta and Jitāripāda. Mysore: Hindusthan Press. Jitāri. Tibetan translation of Sugatamatavibhaṅga (Jitāri), D3990)=P 5461. ———. Tibetan translation of Sugatamatavibhaṅgakārikā (Jitāri), see D3899=P 5296. Kajiyama, Yuichi. 1958. “On the Theory of the Intrinsic Determination of Universal Concomitance in Buddhist Logic.” Journal of Indian and Buddhist Studies 7 (2): 360–64. Kellner, Birgit, Patrick McAllister, Horst Lasic, and Sara McClintock, eds. 2020. Reverberations of Dharmakīrti’s Philosophy, Proceedings of the Fifth International Dharmakīrti Conference Heidelberg, August 26 to 30, 2014. Vienna: Austrian Academy of Sciences Press. Kosambi, D. D., and V. V. Gokhale, eds. 1957. The Subhāṣitaratnakośa, Compiled by Vidyākara. Cambridge, MA: Harvard University Press. Kyuma, Taiken, ed. 2003. “Jitari ni kiserareru Dharmadharmiviniscaya ni tsuite [On Dharmadharmiviniscaya ascribed to Jitāri].” Indogaku Bukkyōgaku Kenkyū 51 (2): 900–4. Malvania, Dalsukhbhai, ed. 1955. Paṇḍita Durveka Miśra’s Dharmottarapradīpa, Being a Subcommentary on Dharmottara’s Nyāyabinduṭīkā, a Commentary on Dharmakīrti’s Nyāyabindu. Patna: Kashi Prasad Jayaswal Research Institute. Matsumoto, Shiro. 1981. “On the Philosophical Positions of Dharmottara and Jitāri.” Journal of Indian and Buddhist Studies/Indogaku Bukyōgaku Kenkyū 29 (2): 966–69. McClintock, Sara L. 2010. Omniscience and the Rhetoric of Reason: Śāntarakṣita and Kamalaśīla on Rationality, Argumentation, and Religious Authority. Somerville, MA: Wisdom Publications. Mimaki, Katsumi. 1976. La Réfutation bouddhique de la permanence des choses (Sthirasiddhidūṣaṇa) et la preuve de la momentanéité des choses (Kṣaṇabhaṅgasiddhi), 2 vols. Paris: Institut de civilisation indienne. Miyasaka, Yūsho. 1964. “A Sanskrit-Tibetan and Japanese Index to the Hetutattvopadeśa of Jitāri: Comparing with the Chinese of Buddhist Logical Terms.” Mikkyō Bunka 68: 31–57. Mookerjee, Satkari. 1935. The Buddhist Philosophy of Universal Flux: An Exposition of the Philosophy of Critical Realism as Expounded by the School of Dignaga. Calcutta: University of Calcutta. Moriyama, Shinya. 2011.“Pramāṇapariśuddhasakalatattvajña, sarvajña and sarvasarvajña.” In Religion and Logic in Buddhist Philosophical Analysis, Proceedings of the Fourth International Dharmakirti Conference Vienna, August 23–27, 2005, edited by Helmut Krasser, Horst Lasic, Eli Franco, and Birgit Kellner, 329–40. Wien: Verlag der Österreichischen Akademie der Wissenschaften. ———. 2014. Omniscience and Religious Authority. A  Study on Prajñākaragupta’s Pramāṇavārttikālaṅkārabhāṣya ad Pramāṇavārttika II 8–10 and 29–33. Zürich and Berlin: LIT Verlag. Roerich, George N., trans. 1949. The Blue Annals (Parts 1–2). Calcutta: Royal Asiatic Society of Bengal. Ruegg, David Seyford. 1981. The Literature of the Madhyamaka School of Philosophy in India. Wiesbaden: Harrassowitz. Sakai, Masamichi. 2019. “Antarvyāpti and Antarvyāptivāditva: Arcaṭa on the Buddhist Inference of Momentariness and Durvekamiśra’s Comparison of Him with Dharmottara.” Journal of Indological Studies 30/31: 99–128. ———. 2020. “Dharmottara on the viparyaye bādhakapramāṇa and trairūpya in Dharmakīrti’s sattvānumāna.” In Reverberations of Dharmakīrti’s Philosophy, Proceedings of the Fifth International Dharmakīrti Conference Heidelberg, August 26 to 30, 2014, edited by Birgit Kellner, Patrick McAllister, Horst Lasic, and Sara McClintock, 375–89. Vienna: Austrian Academy of Sciences Press.

556

Jitāri Sanghavi, Sukhlalji, and Muni Shri Jinavijayaji, eds. 1949. Hetubinduṭīkā of Bhaṭṭa Arcaṭa with the SubCommentary Entitled Āloka of Durveka Miśra. Baroda: Oriental Institute. Shiga, Kiyokuni. 2011. “antarvyāpti and bahirvyāpti re-examined.” In Religion and Logic in Buddhist Philosophical Analysis, Proceedings of the Fourth International Dharmakirti Conference Vienna, August 23–7, 2005, edited by Helmut Krasser, Horst Lasic, Eli Franco, and Birgit Kellner, 423–36. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Shirasaki, Kenjō. 1974. “Jitāri no Anekāntavāda hinhan [The Anekāntavāda of Jitāri].” Indogaku Bukkyōgaku Kenkyū 22 (2): 999–1003. ———. 1981a. “The Bālāvatāratarka.” Indo Gakuhō 3: 23–52. ———. 1981b. “Jitāri, hito to shisō [Jitāri: Life and Thought].” In Sōden no kenkyū: Kimura Takeo kyōjū koki kinen, edited by Takeo Kimura, 320–48. Kyoto: Nagata Bunshōdō. ———. 1983. “Jitāri no ‘Fuhen-jitsuzai-ron-hihan [Japanese translation of Jitāri’s Jātinirākr. ti]’.” Bukkyōshi Kenkyū 26: 1–19. ———. 1986a. “Sugatamatavibhaṅgabhāṣya dai yon shō chūganha no kyō gi wayaku [A Japanese Translation of the Sugatamatavibhaṅgabhāṣya of Jitāri: Chap. IV on Madhyamaka Doctrine].” Nanto Bukkyō 55: 1–104. ———. 1986b. “Dharmakīrti wa chūganronsha de aru [Dharmakīrti Is a Mādhyamika].” Bukkyō Ronsō 30: 110–14. ———. 1995. “Jitāri no Īśvaravādimataparīkṣā [The Īśvaravādimataparīkṣā of Jitāri].” Kōbe Joshi Daigaku Bungakabu Kiyō 28 (2): 105–24. Steinkellner, Ernst. 1990. “Is Dharmakīrti a Madhyamika?” In Earliest Buddhism and Madhyamaka, edited by David Seyfort Ruegg and Lambert Schmithausen, 72–90. Leiden: E. J. Brill. ———. 1991. “On the Logic of Svabhāvahetu In Dharmakīrti’s Vādanyāya.” In Studies in the Buddhist Epistemological Tradition: Proceedings of the Second International Dharmakirti Conference, June, 11–16, 1989, 311–24. Wien: Verlag Der Österreichischen Akademie Der Wissenschaften. ———. 2005. “Dignāga’s Pramāṇasamuccaya, Chapter 1, a Hypothetical Reconstruction of the Sanskrit Text with the Help of the Two Tibetan Translations on the Basis of the Hitherto Unknown Sanskrit Fragments and the Linguistic Materials Gained from Jinendrabuddhi’s Ṭīkā.” www.oeaw.ac.at/ias/ Mat/dignaga\_PS\_1.pdf. Takasaki, Jikidō, Yamaguchi Zuihō, and Hakamaya Noriaki, eds. 1981–1984. Sde dge Tibetan Tripiṭaka Bstan ’gyur, preserved at the Faculty of Letters, University of Tokyo. Vol. 1–20, Inmyōbu (Tshad – ma). Tokyo: Sekai Seiten Kankō Kyōkai. Tāranātha. rGya gar chos ’byung dgos ’dod kun ’byung, spyi’i ang 175 deb ang TBE167001. ’Dzam thang par ma la gzhi bcol ba yin. Ser byes rig mdzod kyis zhu sgrig byas. www.serajeyrigzodchenmo.org/ Thakur, Anantalal, ed. 1987. Ratnakīrtinibandhāvaliḥ (Buddhist Nyāya Works of Ratnakīrit). Patna: K. P. Jayaswal Research Institute. Toshiaki, Tamaru. 1978. “Jitāri no Anekātnavādanirāsa [On Jitāri’s Anekātnavādanirāsa].” Bukkhōgaku Kenyū 34: 22–40. Tucci, Giuseppe, ed. 1956. “Hetutattvopadeśa of Jitāri and Tarkasopnāna of Vidyākaranśānti.” In Minor Buddhist Texts, 247–309. Rome: Istituto italiano per il Medio ed Estremo Oriente. ———. 1971. “The Jātinirākṛti of Jitāri.” In Opera Minora, 249–54. Part I. Roma: G. Bardi. Reprint of “The Jātinirākṛti of Jitāri.” Annals of the Bhandharkar Oriental Research Institute (Poona) (11) (1930): 54–58. Vaidya, Paraśurāma Lakṣmaṇa, ed. 1960. Bodhicaryāvatāra of Śāntideva with the Commentary Pañjikā of Prajñākaramati. Darbhanga: Mithila Institute. Vidyabhusana, Satish Chandra. 1921. A History of Indian Logic: Ancient, Mediaeval and Modern Schools. Calcutta: Calcutta University.

557

34 JÑĀNAŚRĪMITRA Variegated Non-Duality Lawrence McCrea

Jñānaśrīmitra (c. 1025 CE) was among the most accomplished and celebrated philosophers active during the last great flowering of Buddhist intellectual culture in India. He wrote extensively on a variety of philosophical topics in support of positions he took to be grounded in the logical and epistemological tradition descending from Dignāga and Dharmakīrti. While his philosophical writings do not appear to have survived in India, he and his works were still known and quoted by non-Buddhist Indian philosophers centuries later, even after Buddhism had largely disappeared as a serious intellectual force in India; his views are quoted and discussed, for example, by Udayana (eleventh century), Mādhava (fourteenth century), and Śaṅkaramiśra (fifteenth century). He is listed as one of the six famous Dvārapaṇḍitas (“Gateway Scholars”) of the great Buddhist university of Vikramaśīla, and it is clear that he was a major figure in the educational and public disputational culture of that institution. Many of the more important Buddhist intellectuals of the succeeding generation declare themselves, or are said by others, to have been pupils of Jñānaśrīmitra, including Atiśa, Advayavajra, Ratnakīrti, and Yamāri (on Atiśa, see Tārānātha 1990, 302; on Advayavajra, see Tatz 1987, 698; on Ratnakīrti, see Thakur 1957, 17–18). Jñānaśrīmitra’s twelve extant Sanskrit philosophical works (all preserved in a single manuscript and none existing in Tibetan translation) are all topical monographs, or prakaraṇas, rather than full commentaries on the works of Dharmakīrti or other predecessors in the Buddhist epistemological tradition. Many of these philosophical essays cover topics that had long become standard among the Buddhist epistemologists: the momentariness of all objects (kṣaṇabhaṅga),1 the exclusionary or negative nature of linguistic and conceptual content (apoha),2 the nature of and mode of determining the pervasion relationship that undergirds inference (vyāpti),3 the inferential establishment of absences through non-apprehension (anupalabdhi),4 the refutation of the existence of God (īśvara-dūṣaṇa), and so on. Though the range of topics he covers is largely standard, there are some distinctive issues that preoccupy him and some features of his methods and his conceptual inventory set him apart. Building on the insights of the earlier Dharmakīrtian commentator Prajñākaragupta, Jñānaśrīmitra articulates a consistently idealist reading of Buddhist epistemology, deriving from this a remarkably uniform methodology in which nearly all philosophical problems are resolved through an analysis of cognitions into two contrasting aspects: “appearance” ( pratibhāsa) and “determination” (adhyavasāya). All awareness, both conceptual and nonconceptual, consists in the 558

DOI: 10.4324/9781351030908-49

Jñānaśrīmitra

appearance of a certain image (ākāra) in our consciousness. Conceptual awareness is that which, through the process of determination, takes one or more of the cognitions containing these phenomenal images to be intentional – that is, to contain or present images of something other than themselves. All such intentional determinations are erroneous – nothing exists apart from the mental images themselves  – but all worldly activity and belief, and all language, relies inescapably upon these ultimately false determinative projections. Considered independently of these false projections, what appears in our consciousness (and therefore what really exists) is nothing but a single, undifferentiated mental image, which Jñānaśrīmitra, following Prajñākaragupta, terms “variegated non-duality” (citra-advaita).

“Appearance” and “Determination” Earlier philosophy in the Dignāgan/Dharmakīrtian tradition had largely conducted its philosophical analysis, and its critique of the rival realist analyses of the Brāhmaṇical Nyāya and Mīmāṃsā traditions, in terms of two sets of binary oppositions: between the two authoritative means of awareness (  pramāṇa), namely perception (  pratyakṣa) and inference (anumāna), and between two types of knowable objects, namely self-characterized particulars (sva-lakṣaṇas) and commonly characterized universals (sāmānya-lakṣaṇas). The former are understood as unique individuals, absolutely dissimilar and sharing no common features whatsoever. The latter comprise all purported common properties (including natural kinds such as “cow” as well as property-universals such as “blue”) which are thought to unite these individuals into groups or classes, but, at least for Dharmakīrti and his followers, these common properties are purely mental constructions with no ultimate objective reality. A key tenet of the Buddhist epistemologists is that there is a one-to-one correspondence between the two pramāṇas and the objects they cognize, perception apprehending only particulars and inference only conceptually constructed universals. What is distinctive about Jñānaśrīmitra’s reworking of Dharmakīrtian epistemology and ontology is that these two linked binaries are overshadowed by, and in fact radically reconceived in the light of, the appearance/determination binary. Jñānaśrīmitra’s basic model of the role of appearance and determination within the operation of the pramāṇas is derived not from Prajñākaragupta, whom he most typically looks to as his model and inspiration, but from Dharmottara, with whom he has a more complex and often contestatory relationship. It was Dharmottara who first clearly asserted that each of the two pramāṇas has not just one, but two objects – one “grasped,” that is, directly apprehended, and the other “determined.” Jñānaśrīmitra adopts this dual object model and makes it basic to his entire epistemology.5 One of the clearest statements of his basic position on this issue can be found in his Investigation of Pervasion (Vyāpticarcā), his analysis of the pervasion relationship that underlies inference and of the manner in which knowledge of this pervasion relationship is acquired. A large portion of the work is devoted to addressing specific criticisms of Dharmakīrtian inference theory offered by rival philosophical traditions. One of the more important interlocutors here is the Nyāya philosopher Trilocana, author of a (lost) work entitled The Flower Cluster of Reasoning (Nyāyamañjarī), with whom Jñānaśrīmitra engages frequently in his works. Taking the basic Dignāgan position to be that perception has only particulars for its objects, Trilocana argues that this will make inference altogether impossible. Inference depends on prior awareness of an invariant relationship between the inferential reason and the thing to be proven by it (e.g., in the standard example, between the smoke that one sees rising from a mountain and the fire whose presence there one infers from it). Dharmakīrti and his followers hold that this relationship is learned by a combination of perception (of, e.g., fire occurring together with smoke) and non-apprehension (e.g., not observing smoke occurring where fire is absent).6 But, if one can perceive only particulars, and all inference depends 559

Lawrence McCrea

on prior awareness of a link between two universals, there can be no pramāṇa, no valid source of knowledge, that will allow one to know such a relationship. Jñānaśrīmitra summarizes Trilocana’s argument as follows: But Trilocana raises the following objection: Since both perception and nonapprehension have only particulars as their objects, how can one come to know the connection between two universals by means of them? If you say that what we come to know is the connection of “what is excluded from non-smoke” with “what is excluded from non-fire,” we may still ask by what means can this connection be known? It cannot be known by perception, since this has only particulars for its objects. Nor can it be known by inference, since inference itself depends upon first knowing this connection.7 Jñānaśrīmitra’s response to this objection proceeds by challenging Trilocana’s account of the Buddhist theory of perception. It is not in fact the Buddhist position that perception has only particulars as its objects; universals too are cognized via perception, but not in the same way particulars are. It is here that Dharmottara’s account comes into play. Dharmottara argued that all episodes of valid awareness, both perceptual and inferential, have two objects, one “grasped” and one “determined.” When Dignāga and Dharmakīrti insist that perception has only particulars as its object, what they really mean is that perception has only a particular as its grasped object.8 Adopting this model derived from Dharmottara, Jñānaśrīmitra argues that perception is able to apprehend universals and pervasion relationships between them as the second, determined type of object. He seeks to undercut Trilocana’s argument by examining the Buddhist epistemologists’ claim that “Perception has the specific individual as its object.” There are two ways to understand the predicate of this claim: that perception has only the specific individual as its object (such that connection with any other kind of object is excluded) or that perception does indeed have the specific individual as its object (such that non-connection with the specific individual is excluded, without ruling out connection with other objects as well). Jñānaśrīmitra argues that Trilocana has misconstrued the Buddhist’s claim by taking it to entail the first sort of exclusion rather than the second: To the objection of Trilocana, we reply as follows: Does the claim that “Perception has the specific individual as its object” mean that it also has the specific individual as its object, or that it only has the specific individual as its object? On the first view, that there is exclusion of non-connection [of perception with a specific individual] there is nothing to prevent perception having a universal as its object. How then is there no opportunity for perception and non-apprehension of the two universals [linked in the inference warranting relationship]? When you say that there is “exclusion of connection with anything other than a universal,” do you mean that this is so according to our own established conclusion, or that it is really so? Not the first, since, according to us, both pramāṇas have both particulars and universals as their objects, since these objects are divided into those that are “grasped” and those that are “determined.” Whatever appears in any particular awareness is what is “grasped.” But that thing toward which one acts is “determined.” This being the case, in perception it is a particular that is grasped, and a universal that is determined. For inference, on the other hand, the case is just the reverse. Now, when one perceives something that serves as a means of action, and desires the pragmatic effect to be achieved by that means, even though that perception apprehends only a single moment, it does have a universal as its object, 560

Jñānaśrīmitra

insofar as it bears on the object-continuum. But, in a situation in which one learns the pervasion relationship, one perceives only a single individual, but this perception too has a universal for its object, since it takes as its object all particulars of the same class. When one has the thought “I ascertain by perception that smoke is born from fire,” one is aware of a determination arising from this perception that incorporates all individual instances of smoke, excluding whatever is different from smoke. This is just like cases where one sees merely visible form, but, because one determines that one has seen, for example, a pot, which is a specific collection of visible, gustatory, olfactory, and tactile properties, the perception has that pot as its object.9 Here Jñānaśrīmitra explicitly adopts Dharmottara’s two-object model of both perception and inference as his own. Jñānaśrīmitra applies this theory specifically to the problem raised by Trilocana’s critique: How can one apprehend, through perception alone, the pervasionrelationship between, for example, smoke and fire? Because all perception has a second, determined and universal, object in addition to the momentary particular that it directly grasps, perception is able to, and indeed necessarily does, make us aware of universals through determination. That any given perceptual awareness grasps only a momentary particular poses no problem, as determination creates a second, universal object toward which one directs one’s subsequent activity. This is as much true when we perceive a purportedly individual object such as a single pot (which is not a true individual, according to the Buddhist epistemologists, but already a universal constructed by grouping individual sensibilia into an imagined whole), as when we determine class-universals based on a finite set of observations – such as that “smoke (in general) is born from fire (in general).” Trilocana’s misreading of the Buddhist position, as Jñānaśrīmitra understands it, hinges on a misunderstanding of what exactly a “universal” is on the Buddhist account. As Jñānaśrīmitra makes clear, a universal is not a distinct kind of entity – inaccessible to perception but required for inference. Rather, it is merely one possible way of conceptualizing the particulars that perception grasps: There is no other thing called a “universal.” Rather, it is the particulars themselves that are called a “universal” when their differences from one another are not separately discerned. When those differences are separately discerned, each can be designated by its own term, “particular.” So, even if all the particulars distinct from what is other do not appear in a perceptual cognition, and hence cannot be its “object” in the sense of being grasped by it, nevertheless it cannot be denied that they are its object in the sense of being determined by it.10 The Buddhist view of universals, as typically presented by both themselves and their opponents, is that they are mentally constructed and hence wholly fictitious objects, “non-things” (a-vastu). But Jñānaśrīmitra insists that this is not correct. A “universal” is simply a set of particulars conceptualized as a class rather than as individually distinct entities. Hence perception can, through the process of determination, allow us to be aware of universals and the universal pervasion relations between them.

Exclusion (apoha) Some of Jñānaśrīmitra’s most original and influential writing revolves around his analysis of the Buddhist theory of “exclusion” (apoha), the view first advanced by Dignāga that the content of all linguistic and conceptual awareness takes the form of a negation, excluding the 561

Lawrence McCrea

conceptualized item from some domain without ascribing to it any objectively real feature or property. Jñānaśrīmitra devoted a full monograph to the topic,11 in which he not only surveys and refutes the most important criticisms of the theory offered by anti-Buddhist philosophers but analyzes and critiques a variety of internal Buddhist positions on the question. This long essay proved quite influential, and, along with the briefer summary of the same arguments by Jñānaśrīmitra’s pupil Ratnakīrti in his own Proof of Exclusion (Apohasiddhi), redefined the Buddhist position on this central issue. Jñānaśrīmitra’s discussion in particular drew a detailed critique from Udayana in his Determination of the Essence of the Self (Ātmatattvaviveka) and thereby established itself as a touchstone for later critiques from Nyāya and other Brāhmaṇical systems. Here, as in so many other areas, Jñānaśrīmitra’s distinctive and innovative treatment of this standard Buddhist topic turns on a careful discrimination of the roles of appearance and determination in linguistic and cognitive processes. By his time, Buddhists had been arguing among themselves for centuries about the proper understanding of the nature of the exclusion said to define the content of linguistic and conceptual awareness. They asked: Should we take the content of a conceptual awareness to be (1) a pure negation (which exists only internally, as a mental image), (2) an external object qualified by such a negation, or (3) some hybrid of the two? (All three of these positions – sometimes dubbed “negativist,” “positivist,” and “synthesist” – have been advanced by authors within the Buddhist epistemological tradition at one time or another; see Mookherjee 1993, 131–33; Katsura 1986, 171.) Jñānaśrīmitra was the pioneering and most prominent advocate of the “synthesist” position – that the “object” of conceptual awareness necessarily involves both a positive and a negative component. And, significantly and typically, he links these two aspects to the two basic cognitive modes of appearance and determination. He summarizes his own view of exclusion in a single verse, on which his entire Monograph on Exclusion (Apohaprakaraṇa) is constructed as a commentary: First of all, it is the external object that is primarily expressed by words. This being the case, exclusion is understood as a qualifier of that. One of these – the external object – is conditionally adopted as an object due to determination; the other – the exclusion – is conditionally adopted as an object due to appearance. But really, nothing at all is expressed. While the traditional position of the Buddhist epistemologists is that linguistic awareness has only a negation or “exclusion” as its referent (so that the term “cow” signifies “what is not a non-cow”), Jñānaśrīmitra stresses, as many of his predecessors did not, that conceptual content must have a positive component and that this component is, at least phenomenologically, the predominant element in such awareness, exclusion serving only as a necessary subordinate element. It is this that accounts for our common intuitions about linguistic reference. Even the exclusion theorist acknowledges that when we use the expression “There is a cow,” or recognize an object before us as “a cow,” we typically feel ourselves to be speaking of and recognizing a physical object external to our own language usage and thought. This is the “external object” that is “primarily expressed by words.” But this purported external object is not what appears in a conceptual awareness episode; rather, it is what is determined by it. Through analyzing the components of inferential/verbal awareness in terms of the appearance/determination binary, Jñānaśrīmitra is able to explain how such awareness functions on a conventional level without admitting the ultimate reality of any verbal referent or conceptual object at all. Given the need to reconcile two seemingly contradictory positions earlier taken by the Buddhist epistemologists – that words have exclusion as their referent and that words have no referents at all – Jñānaśrīmitra ultimately concludes that the question is perspectival: one 562

Jñānaśrīmitra

may legitimately speak of different things being the referents of words from the perspective of appearance, of determination, or of ultimate truth. He closes out his discussion as follows: So, when one is asked, “How is exclusion expressed by a word?” we answer, “As an element of that [external object that is primarily expressed],” the meaning of which is as described. If the question is, “Why are the mental image, or the particular, or the contingent features not expressed?” these questions are dispensed with in order by saying, “This is because of the absence of determination, the absence of appearance, and the absence of both.” But if the question is “What is it that is expressed by words?” then, having set out these options (1) on the basis of appearance, (2) on the basis of determination, or (3) really, the answers are, in order, (1) “the image that is excluded from what is other, that resides in conceptual awareness”; or (2) “the particular that is excluded from what is other”; or (3) “nothing.” This has already been said. Therefore, establishing the position that words and inferential reasons have exclusions as their objects is for the sake of making it known that all properties are inexpressible. For Jñānaśrīmitra, all conceptual awareness presupposes the existence of objects which are both apparent to us and available to us as objects of action (even if that action is only verbal or mental). But ultimately there can be no object that meets both of these criteria. Cognitive images appear in our awareness, but these, being momentary, are not amenable to deliberate action; determination constructs the imagined external particulars toward which all our activity (even verbal and mental activity) is directed, even though these determined particulars do not actually appear in our awareness. But there exists no one thing that can both appear and be determined. “Exclusion” must be accepted as an element that qualifies the (fictitious) external particular determined by a conceptual awareness but cannot be a self-sufficient object of such awareness. Hence he defends the Dignāgan theory of exclusion but only as a partial account of awareness on the conventional level. Ultimately, language has no object at all, neither a positive entity nor an exclusion.

Conditionally Adopted Positions A key feature of Jñānaśrīmitra’s treatment of exclusion, as can be seen from the previous summary, is his relegation of many of even the most characteristic and prestigious positions of the Buddhist epistemological tradition – such as the claim that the objects of linguistic or conceptual awareness are exclusions (apoha) – to the realm of conventional, rather than ultimate, truth. For him, conceptual awareness really has no object at all, universal or particular, internal or external. It is only conventionally that may speak of anything as the “object” of such an awareness. Why then did Buddhist thinkers beginning with Dignāga insist that exclusion is the object of inferential/verbal awareness if this is not really the case? To explain this, Jñānaśrīmitra develops a sophisticated theory of “conditionally adopted positions” (vyavasthā), that is, positions that one may legitimately adopt in a philosophic argument even while knowing them to be untrue: By relying on a little bit of the truth, a certain conditionally adopted position is, for a specific purpose, constructed by us, in one way, even though the actual state of affairs is different, just as in examples such as the “self ” or the “arising of a thing.” For “arising” can be a property only of an existing object qualified by a prior absence. By relying on a little bit of the truth, namely, the prior absence, we conditionally adopt the position that “there is arising of a nonexistent thing” in order to foreclose 563

Lawrence McCrea

any worries about the doctrine that effects preexist in their causes. Or, by relying on the conceptual construction of a single continuum, we conventionally say: “Who else will experience the result of an action done by this very person?” in order to frustrate the deceptive view that there is the passing away of what has been done and the onset of what has not been done. He then applies this theory specifically to his interpretation of the theory of exclusion: Here too, the idea that linguistic expression takes a positive entity as its object is just the same [in that it, too, is a conditionally adopted position]. Here we conditionally adopt the position that exclusion, even though it is really just a necessarily attendant awareness, is the object of conceptual awareness, in order to set aside any suspicion that we accept the position pushed by our opponents that it is only the positive entity that is really expressed. And therefore, we don’t talk in terms of just the positive entity [when describing the semantic value of a word]. But when someone pushes the position that “exclusion alone is the primary meaning of a word,” then we put forth the positive entity as well. As stated, “First of all, it is the external object that is primarily expressed by words.” But in perception, because there is no disagreement of this sort, it is proper that one should not conditionally adopt this position.12 It turns out that for Jñānaśrīmitra, both perceptual and inferential/verbal awareness necessarily have both a positive and a negative component. What differentiates perceptual from inferential/verbal awareness is not that one has a positive and the other a negative “object” but that the rhetorical context in which the Buddhist theory is advanced justifies the claim of a negative object in the second case but not the first. In the case of conceptual and verbal awareness, the advancing of the exclusion-based account, although untrue, is justified by the need to counter the realist contention that the positive entity alone is expressed. And it is the theory of conditionally adopted positions that provides this justification. Through this explanation of the “conditionally adopted position,” Jñānaśrīmitra clearly theorizes, perhaps for the first time and certainly the most perspicaciously, a process that had been underway for some time. Almost since the first layer of commentary on Dharmakīrti’s works, philosophers in the Buddhist epistemological tradition had been engaging in a mostly tacit program of rational reconstruction, adjusting, expanding, or reformulating Dharmakīrti’s key concepts and arguments to confront new objections and to maximize the coherence and explanatory power of the system as a whole and often adopting stances that would likely have been anathema to Dharmakīrti himself. (Dharmottara’s dual-object model of the pramāṇas is a good case in point.) What is most striking about Jñānaśrīmitra’s treatment is the openness with which he acknowledges and elucidates this process, openly dismissing the overt positions of Dignāga and Dharmakīrti while simultaneously justifying them as “white lies,” ostensibly deliberate misrepresentations rendered virtuous through the pedagogical or rhetorical purposes they serve.13

Lower- and Higher-Order Conventional Reality By thus relegating key elements of the established doctrine of the Dignāgan and Dharmakīrtian tradition to the realm of conventional rather than ultimate truth, Jñānaśrīmitra is in part doing what commentators often do, updating the stances of their forbears to meet new philosophical and dialectical needs. But, as we have seen already in his development of the idea of 564

Jñānaśrīmitra

conditionally adopted positions, he is also unusually open and self-conscious in making this move, and, when he questions the ultimate truth of the overtly stated views of the philosophers whose tradition he ostensibly upholds, he seems to call into question the ultimate value of their having advanced these views in the first place. By linking the legitimacy of conventional assertion to the ability to articulate a legitimate purpose for the assertion of things not strictly and ultimately true, Jñānaśrīmitra raises overtly the question of how we can meaningfully assess the value or usefulness of philosophical positions as distinct from their accurate representation of what is really the case. If all the claims of philosophers, being verbal and thus necessarily conceptual, are thereby in the final analysis misrepresentative of ultimate reality, how then are they to be weighed against one another? If the “truths” presented by the Buddhist epistemologists are merely conventionally rather than ultimately true, why should they be at all preferred to the views of rival Brāhmaṇical philosophers or, for that matter, to the philosophically untutored beliefs of ordinary people? In response to such objections, Jñānaśrīmitra finds it necessary to develop a more sophisticated account of conventional truth. As far back as the time of Nāgārjuna, Buddhist philosophers had routinely spoken of “the two truths,” conventional (saṃvṛti) and ultimate ( paramārtha); many things that may legitimately taken as “true” for everyday practical purposes – such as the existence of perduring extra-mental objects – must be rejected as ultimately false when subjected to philosophical analysis. But Jñānaśrīmitra goes beyond this binary division, arguing that different conventional truths must be seen as ranked in a continuum ranging from lower (adhara) to higher (uttara), with the latter being systematically preferred to the former, even though both necessarily fall short of ultimate truth. This issue is discussed most fully in the opening section of Jñānaśrīmitra’s Treatise on Moment by Moment Destruction (Kṣaṇabhaṅgādhyāya). As a committed idealist, Jñānaśrīmitra is confronted with a basic problem at the outset of his extensive proof of momentariness. What is the point of proving, and what does it even mean to prove, that mind-external objects must exist only as series of causally related moments if one will only go on to argue that no such objects exist at all? Jñānaśrīmitra takes the position that the long-held Buddhist tenet of momentariness must be defended as a higher-order convention, in preference to the lowerorder convention upholding the existence of temporally persistent objects, and it is in connection with this argument that he offers his most fully developed theorization of higher- and lower-order conventional truths: Even though the causal capacity of objects is only conventionally real, it must necessarily be accepted if this investigation is to have any validity. And, because of this, postulation of purportedly stable entities such as “universals” and “inherence”14 should not be accepted, in so far as they are falsified by the proof of moment-by-moment destruction adduced from this causal capacity of objects, provided one accepts that objects have pragmatic efficacy. (If, on the other hand one does not accept that objects have pragmatic efficacy, then there would have to be conventional acceptance of reality for something incapable even of producing an awareness of itself; and this should be considered to be an even lower-order convention.) Therefore, given that progressively higher-order (uttara) conventions are conditionally established (vyavasthāpyamāna) as “ultimately real” in comparison with progressively lower-order conventions, the conventional belief in pragmatic effectiveness, having no convention higher than it, must be accepted as “ultimately true,” and whatever exists with this capacity of pragmatic effectiveness is called “ultimately real.” This is just like what happens when a child thinks there is a face in the surface of a mirror, and, in contrast to this belief, one says to the child as an ultimate truth that “What is seen there is not a real face,” even though, in contrast with 565

Lawrence McCrea

this belief, there is a further “ultimate truth” that the rays of light from one’s eyes, turned back upon oneself by striking the mirror, actually apprehend one’s own face. Thus whenever successively later conceptual constructions prove more powerful than those that precede them, we say that each earlier one is “conventional,” and each one that follows after is “ultimately true,” up until the point when pragmatic effectiveness is achieved. The belief that external objects exist only momentarily is not in the final analysis any more true than the view that they exist as perduring over time. Both views are ultimately false, as a philosophically educated view will hold that no such objects can possibly exist in the first place. Nevertheless, the view that such objects exist as series of causally related moments is to be preferred as a higher-order convention, being “stronger” than the lower-order convention it displaces. We are not told what exactly the criteria of “strength” in this context are, except that they are to be explained in terms of “pragmatic effectiveness” (arthakriyā). One conventional view is ranked as stronger and therefore “higher” than another based on its usefulness for some purpose, not because it more accurately approaches or approximates ultimate truth. In light of Jñānaśrīmitra’s theory of conditionally adopted positions discussed previously, presumably such purposes include rhetorical effectiveness in philosophically refuting the views of others as well as soteriologically assisting the holders of such beliefs in the quest for enlightenment.

Jñānaśrīmitra’s “With-Image” Idealism and the Struggle with Ratnākaraśānti One of the central preoccupations of Jñānaśrīmitra’s philosophy concerns an entirely intraBuddhist debate regarding the ultimate ontological status of the mental contents or images (ākāra) that appear in our awarenesses. Two of his surviving works, together comprising more than a third of his extant philosophical oeuvre, are devoted to this topic: he wrote a A Treatise Proving That Cognitions Contain Real Images (Sākārasiddhiśāstra), an extended prose piece devoted programmatically to this topic, as well as a shorter (but still lengthy) verse summary of the same argument, his Brief Summary of the Argument That Cognitions Contain Real Images (Sākārasaṃgrahasūtra). Jñānaśrīmitra begins his Treatise Proving That Cognitions Contain Real Images with a general statement of his idealist ontology, which holds that all that really exists is a single multiform (citra) cognitive image: It has been established that this entire triple world is nothing but consciousness. Here in this work we explain the greater or lesser strength of internal divisions within this consciousness. To explain: By the steps of refuting mind-external objects and the like, This world is, first of all, not external, since in itself it has a form apparent to consciousness, and due to the necessary co-apprehension [of supposed objects with the cognitions that apprehend them]. No other thing can possibly be established through reasoning, and, even if it were possible, it would not have the characteristic of being apprehensible. Therefore, all this world is mere consciousness. The world is one variegated sensory image, comprising colors such as gold, sounds such as the third note of a musical scale, tastes such as sweet, odors such as the pleasantly fragrant, tactile sensations such as soft, and mental sensations such as pleasure and displeasure. Thus there is the following scriptural passage, “Indeed, O Monks, these three worlds are mind only.”15 566

Jñānaśrīmitra

Jñānaśrīmitra, following Prajñākaragupta and the idealist Yogācāra tradition more broadly, takes the basic position that all that exists is mere consciousness, with no corresponding extramental objects. The very fact that we are aware of objects at all is sufficient to establish that these “objects” are mental. To be apparent to awareness is precisely the mark of the mental. This is often summed up by the principle of “co-apprehension” (sahopalambha) first articulated by Dharmakīrti that “there is no distinction between ‘blue’ and the awareness of blue, since they are necessarily apprehended together” (Pramāṇaviniścaya 1.54ab); one can never be aware of any purported “object” such as blue as distinct from one’s awareness of it, and hence there can be no evidence for the existence of such an object. But this basic position, by Jñānaśrīmitra’s time widely (if not universally) shared by Buddhist intellectuals, gives rise to a further intra-Buddhist debate about the nature and epistemic status of the purported internal contents of our awareness. Are the images that appear in our awareness themselves ultimately real, or are they themselves fictions or illusions ultimately to be displaced? Jñānaśrīmitra argues forcefully that these images are not, and in fact cannot, be displaced or overturned in any way, and are ultimately real (are, in fact, all that is ultimately real). In making this case, he summarizes what he takes to be the arguments of the two principal Buddhist opponents of this view, the Mādhyamikas and the “No Image” (nirākāra) Yogācāras. He begins by briefly presenting the Mādhyamika position: On this point, some people [i.e., the Mādhyamikas] say: Let us agree that sensory pro­ perties such as white, and internally sensed properties such as pleasure have the nature of pure consciousness. Even so, they cannot escape the emptiness of nature characteristic of all things. If you seek to establish that external objects cannot exist by introducing an argument based on the impossibility of their being either unitary or multiple, there is no reason why this argument would recoil from properties such as brightness even if they bear a purely mental nature [since they would still be inexplicable as either unitary or multiple]. Therefore, it is better to say that all properties [whether mental or extramental] are just empty, as is set forth in the Buddhist scriptures in many places, and also in Dharmakīrti’s Critique of the Sources of Valid Awareness (Pramāṇavārttika), where it is said that, “In whatever way one considers the objects of cognition, they collapse in just that way” (Pramāṇavārttika, chapter on perception, 209). The view that all is consciousness is merely a conventional and not an ultimate truth.16 Having thus briefly surveyed the Mādhyamikas’ argument against the ultimate reality of awareness’s phenomenal content, Jñānaśrīmitra then turns to introduce and confront his principle adversary, the “Without Image” Idealist (Nirākāra-yogācāra) school as represented by his contemporary and rival Ratnākaraśānti. He begins by summarizing this view, which maintains that, while all the purported objects of our ordinary awareness are ultimately “blocked” (bādhita), that is, overturned and displaced by other cognitions and hence to be rejected as false (alīka), “illumination itself ” ( prakāśamātra) remains as the constant and ineradicable nature of all awareness, which can never be falsified or overturned and hence must be accepted as ultimately real: Others [i.e., Ratnākaraśānti] resolve this in the following manner: Let us accept that the forms of awareness are false, in that they are blocked. But the illumination itself is true, and accepted by our tradition, because an awareness consisting of this illumination is perceptual, since it cannot possibly be blocked. “Illumination” itself is the innate form of an illuminating awareness; it cannot be introduced into it through some confusion, such that the awareness of it could be an error. Specific phenomenal content such as “blue,” because it is a form different from the awareness itself, could 567

Lawrence McCrea

be introduced through a confusion, and hence the awareness of it could be an error. So, there is an opportunity for the blocking of phenomenal content such as blue, but not of illumination. Negation should always be understood to be of that which is false or conceptually constructed.17 Jñānaśrīmitra’s refutation begins by attacking the notion of the “illumination” which the Nirākāra theorist imagines to remain unfalsified when all specific awareness-contents have been falsified. If in all cognitions what appears to us are just specific cognitive images (ākāra), how can we ever come to be aware of this supposed “illumination” as distinct from the images that invariably appear along with it? Regarding this view the following must be considered: What exactly is this “illumination” as distinct from phenomenal content such as blue? If there is no distinction, then how can you avert the fear of illumination being blocked? In any case, there is no valid authority to believe that it exists as something distinct. Perception cannot be the authority, since it is impossible in the awareness of blue to apprehend anything other than the perceived blue. Nor can inference be the authority since one can never become aware of any potential inferential sign’s having an inference-warranting connection with something that has never previously been perceived. Furthermore, what could be more of an embarrassment than to claim that the “illumination” existing in one’s own stream of cognitions is known through inference, which only apprehends things beyond the range of one’s perception?18 If the existence of “illumination” as distinct from the specific phenomenal content of any awareness cannot be established by either perception or inference (the only two valid means of knowledge accepted by the Buddhist epistemologists), then there is no way one could ever be aware of this “illumination” as distinct from the image presented by the awareness. And, in the absence of any evidence for illumination as a distinct element, there is no way one can argue that the cognitive image is blocked or overturned while the illumination remains somehow unfalsified. An awareness and its image are ultimately inseparable, and there is no way the one can be falsified without the other. This line of argument, if successful, effectively undermines Ratnākaraśānti’s position but does not by itself establish Jñānaśrīmitra’s countervailing view, that all the images that appear in our awareness remain unfalsified and are therefore to be accepted as ultimately real. As a central pillar of his case for the ultimate reality of mental content, Jñānaśrīmitra seeks to undermine the entire notion that the presentational content of any cognition can ever be “blocked” or falsified in any way – an argument principally directed against Ratnākaraśānti but which effectively undermines the Mādhyamika opponent as well, as it makes it ultimately impossible to falsify the presentational content of any awareness. Jñānaśrīmitra’s argument turns, once again, on analyzing the process of falsification through the familiar rubric of appearance and determination. The falsification or “blocking” of one cognition by another necessarily requires the determination of the second cognition as being a cognition which apprehends the first cognition as “false,” which, Jñānaśrīmitra argues, can occur only at the level of conventional reality. This argument is presented most forcefully in the following passage from the fourth chapter of the Treatise Proving That Cognitions Contain Real Images: But how can there be any falsity (alīkatā) of the nature of an experiential awareness? You may say, “Because another awareness blocks it.” But what is it that is “blocked,” and by what? You may argue that, “There is blocking of the whole collection of 568

Jñānaśrīmitra

variegated mental images, by the non-apprehension of a factor that would necessarily pervade them.” But this is protecting the cow after you’ve already sold it: A relation of blocker and blocked has been defended in Prajñākaragupta’s Commentary on Dharmakīrti’s Critique of the Sources of Valid Awareness, but only on the conventional level, since a contradiction has been established on the theory of “appearance of the non-existent,”19 which would pervade the relation of blocker and blocked.20 The idea that one cognition could block another depends on the belief that a second cognition could somehow reveal that the seeming content of the first was neither a real external object or the cognition’s own real form but something totally non-existent. But this can be accepted only on the level of conventional truth, because it is ultimately not possible for a cognition to reveal anything other than itself and therefore for one cognition to reveal the falsity of another. Jñānaśrīmitra then proceeds to flesh out this claim of impossibility by asking precisely how one cognition could block the content of another: Furthermore, an awareness cannot block itself, since its consciousness of its own form would overpower any possible blocking. But, if one awareness were to be blocked by another, would that other awareness be one with the same object, or a different object? And, would it be simultaneous or non-simultaneous with the awareness that it blocks? Supposing it has the same object, it would then be incoherent to say that this awareness has the object of the first cognition as its object, but nevertheless blocks that very object. But how could the object of the first cognition be blocked by a cognition with a different object, since all the second cognition does is to present its own object? And why shouldn’t the second cognition be blocked by the first? Furthermore, one cognition could not be blocked by another occurring simultaneously: If multiple objects, whether similar or dissimilar, were to appear at the same time, how could one know which is blocked by which? But, if they occur at different times, this is all the more reason why there is no occasion for one to block the other. This would lead to impossibilities such as the ancient king Mahāsaṃmata being in conflict with the much later king Śrīharṣa.21 You may argue that what is meant by “blocking” is that one cognition makes known that another has no object, and that there is no contradiction in this occurring between two cognitions occurring at different times. But if the image to be blocked did in fact appear in the prior cognition, then it surely did exist at that time; how then could it be “blocked”? If, on the other hand, it did not appear in the initial cognition, then it simply does not exist on its own accord – what need is there of any further negation by the later blocking cognition? And, in any case, the so-called “blocking” cognition could not be said to have the same object as the earlier cognition.22 No cognition can coherently be said to block itself, but no other cognition, whether simultaneous or successive, could block it either. For this to occur, it would be necessary for the two cognitions to be “of ” the same thing but to present it differently, and this Jñānaśrīmitra shows to be impossible. It is only conventionally that one could imagine any cognition to represent or misrepresent anything other than itself: You might argue that there is prohibition of what is superimposed.23 But we have already explained that superimposition can be established only on the level of conventional 569

Lawrence McCrea

reality, and hence the blocking of it also cannot go beyond the conventional. You may argue as follows: Since “error” is simply the arising of a successive cognition in reliance on a specific kind of dependent co-origination, in just the same way “blocking” is just the arising of an awareness of the form “This is like this, and not otherwise.” If so, then it is established that there is really no blocking at all. A given awareness arises through taking on a certain image, but nothing is either established or blocked by it, just as an awareness of sweetness arising immediately after an awareness of blue neither establishes nor blocks that prior awareness. If you say that there is a difference, since in this case there is blocking of the other through determination, we respond as follows: To say that this blocking is “established only through determination” is precisely to say that it is conventional. Therefore the meaning of “blocking” cannot be that it makes known that another awareness lacks an objective foundation.24 Seen simply in terms of their phenomenal appearance as cognitions, one following the other in sequence, there is no way a second cognition could make known the unreality or falsity of the image appearing in the first. Instead, this can happen, in Jñānaśrīmitra’s account, only on the level of conventional reality. “Determination” consists of taking the phenomenal image that appears in one’s awareness to be an image “of ” something else and is necessarily a form of conceptual awareness and, as such, ultimately erroneous. It is only by falsely projecting the image that constitutes an awareness as being the image of something outside that awareness that one can take one awareness to falsify another. At the ultimate level, the level of appearance at which no awareness reveals anything beyond itself, blocking is impossible and unimaginable. Throughout this entire argument, we see again how the appearance/determination binary grounds Jñānaśrīmitra’s approach to the most important philosophical problems he deals with. He shows at great length that any theory of the “blocking” of one cognition by another entails taking the second, blocking cognition to be a cognition of something other than itself, which necessarily depends upon the process of determination. What appears in any cognition is nothing more than the image (ākāra) that appears in it; taking it to cognize anything other than this image is precisely what Jñānaśrīmitra understands determination to consist of. But all determination is conceptual and therefore a form of error. Hence the appearance-content of all awareness remains unfalsified and must be accepted as ultimately real.

Conclusion Through his thoroughgoing rereading of the Dignāgan/Dharmakīrtian tradition through the lens of the binary appearance/determination rubric and his theoretically self-conscious deve­ lopment and deployment of the “conditionally adopted position” as his key hermeneutic tool, Jñānaśrīmitra is able to effectively position himself both as the culmination of the half millennium of Buddhist epistemological thought that precedes him and as a creative and often provocatively innovative thinker in his own right. His work offers an excellent example of the ways in which Indian philosophers operating within broadly scholastic intellectual traditions could satisfy the sometimes conflicting imperatives of exegetic faithfulness and original and creative philosophizing. His rich and still very much understudied oeuvre stands as one of the monuments of late first-millennium Indian philosophy and one of the most sophisticated attempts to synthesize the diverse mass of Buddhist epistemological theory into a coherent system.

570

Jñānaśrīmitra

Notes 1 See his Treatise on Momentary Destruction (Kṣaṇabhaṅgādhyāya), partly edited and translated in Kyuma 2005. 2 See his Monograph on Exclusion (Apohasiddhi), edited and translated in McCrea and Patil 2010. 3 See his Investigation of Pervasion (Vyāpticarcā), edited and translated in Lasic 2000. 4 See his Secret of Non-Apprehension (Anupalabdhirahasya), edited and translated in Kellner 2007. 5 On the history of the theory of determination within the Dharmakīrtian tradition and Dharmottara’s transformative role in it, see McCrea and Patil 2010, 16–20; McCrea and Patil 2006, 318–31; Dreyfus 1997, 354–64. 6 The precise nature and number of such prior observations and non-observations required to acquire knowledge of a pervasion relationship was itself a matter of controversy, and Jñānaśrīmitra devoted another short work to this problem, his Proof of the Relation Between Cause and Effect (Kāryakāraṇabhāvasiddhi, in Thakur 1959, 317–22). 7 Thakur 1959, 161. 8 See Dharmottara, Nyāyabinduṭīkā 1.12; for further discussion of Dharmottara’s account and Jñānaśrīmitra’s deployment of it, see also McCrea and Patil 2006. 9 Thakur 1959, 166. 10 Thakur 1959, 166. 11 McCrea and Patil 2010. 12 Thakur 1959, 204–5. 13 On the conditionally adopted position as “white lie,” see Patil 2007. 14 “Universals” (sāmānya) are held by the Brāhmaṇical realists to be eternal unchanging class-properties or essences manifested in individual members of a class such as “cow”; “inherence” (samavāya) is the relation by which these universals inhere in individual class-members. 15 Thakur 1959, 367. The original scriptural source of the quotation is unknown, but was earlier quoted by Vasubandhu (Introduction to Viṃśikā verse 1) and several other authors following him. 16 Thakur 1959, 367–68. 17 Ibid., 368. As Kajiyama has shown, this is a summary of Ratnākaraśānti’s presentation of his own position in his Instruction in the Perfection of Wisdom (Prajñāpāramitopadeśa). See Kajiyama 1965, 36–37. 18 Thakur 1959, 368. 19 One of the classical theories of error first laid out by Maṇḍanamiśra in his Vibhramaviveka, the doctrine holds that what appears in an erroneous awareness is something wholly unreal (asat). The Buddhist epistemologists reject this doctrine, and Prajñakāragupta argues against it in his commentary on Pramāṇavārttika, Pratyakṣa 331, as part of his overall rejection of the idea that any cognition can ultimately make known anything other than itself (Prajñākaragupta 1943, 356). 20 Thakur 1959, 437. 21 Mahāsaṃmata is a legendary king, said to have lived long before the time of the historical Buddha. “Śrīharṣa” refers to Harṣavardhana, who ruled a large North Indian kingdom centered in Kanauj from 607–647 CE. 22 Thakur 1959, 437. If, as supposed here, the content to be blocked did not appear in the initial cognition in the first place, there is no way the later cognition that supposedly overturns it can be seen as bearing on the same object – hence the content to be overturned was never affirmed in the first place, and the notion of “blocking” is rendered incoherent. 23 That is to say, the image the first cognition was thought to “grasp” was never really part of that cognition but was erroneously superimposed on it. The second “blocking” cognition therefore does not overturn the real nature of the prior one but merely a misidentification wrongly imposed on it. 24 Thakur 1959, 437.

Works Cited Dreyfus, Georges B. J. 1997. Recognizing Reality: Dharmakīrti’s Philosophy and Its Tibetan Interpretations. Albany: State University of New York Press. Kajiyama, Yuichi. 1965. “Controversy Between the Sākāra- and Nirākāra-vādins of the Yogācāra School – Some Materials.” Indogaku Bukkyōgaku Kenkyu 14 (1): 26–37.

571

Lawrence McCrea Katsura, Shoryu. 1986. “Jnānaśrīmitra on Apoha.” In Buddhist Logic and Epistemology: Studies in the Buddhist Analysis of Inference and Language, edited by Bimal K. Matilal and Robert D. Evans, 171–84. Dordrecht: D. Reidel. Kellner, Birgit. 2007. Jñānaśrimitra’s Anupalabdhirahasya and Sarvaśabdābhāvacarcā: A Critical Edition with a Survey of his Anupalabdhi Theory. Wien: Arbeitskreis für tibetische und buddhistische Studien, Universität Wien. Kyuma, Taiken. 2005. Sein und Wirklichkeit in der Augerblicklichkeitslehre Jñānaśrīmitra’s Ksanabhangadhyaya I. Paksadharmatādhikāra: Sanskrittext und Ubersetzung. Wien: Arbeitskreis für tibetische und buddhistische Studien, Universität Wien. Lasic, Horst. 2000. Jñānaśrīmitras Vyāpticarcā: Sanskrittext, Übersetzung, Analyse. Wien: Arbeitskreis für tibetische und buddhistische Studien, Universität Wien. McCrea, Lawrence J., and Parimal G. Patil. 2006. “Traditionalism and Innovation: Philosophy, Exegesis, and Intellectual History in Jnānaśrīmitra’s Apohaprakaraṇa.” Journal of Indian Philosophy 34: 303–66. ———. 2010. Buddhist Philosophy of Language in India: Jnānaśrīmitra on Exclusion. New York: Columbia University Press. Mookherjee, Satkari. 1993. The Buddhist Philosophy of Universal Flux: An Exposition of the Philosophy of Critical Realism as Expounded by the School of Dignāga. New Delhi: Motilal Banarsidass. Patil, Parimal G. 2007. “Dharmakīrti’s White Lie: Philosophy, Pedagogy, and Truth in Late Indian Buddhism.” In Pramānakīrtih: Papers Dedicated to Ernst Steinkellner on the Occasion of His Seventieth Birthday, Part 2, edited by Birgit Kellner, Helmut Krasser, Michael T. Much, and Helmut Tauscher, 597–619. Wien: Wiener Studien zur Tibetologie und Buddhismuskunde. Prajñākaragupta. 1943. Pramāṇavārttikabhāṣyam. Edited by Rahul Sankrityayana. Patna: Kashi Prasad Jayaswal Research Institute. Tārānātha. 1990. History of Buddhism in India. Translated by Lama Chimpa and Alaka Chattopadhyaya. New Delhi: Motilal Banarsidass. Tatz, Marm. 1987. “Life of the Siddha Philosopher Maitrīgupta.” Journal of the American Oriental Society 107 (4): 695–711. Thakur, Anant Lal, ed. 1957. Ratnakīrtinibandhāvali. Patna: Kashi Prasad Jayaswal Research Institute. ———. 1959. Jñānaśrīmitranibandhāvali. Patna: Kashi Prasad Jayaswal Research Institute.

572

35 RATNAKĪRTI Aligning Everyday Experience with Momentariness and Idealism Patrick McAllister

The Scholar and His Philosophical Works Ratnakīrti was a Buddhist scholar (mahāpaṇḍita) active in the monastery of Vikramaśīla in today’s Bihar during the first half of the eleventh century CE. Today, Ratnakīrti is best known for his philosophical works. All of these works were composed in Sanskrit, and most of them can best be characterized as concise and analytic summaries of the much longer and more complex works composed by his teacher, Jñānaśrīmitra. In their texts, Jñānaśrīmitra and Ratnakīrti present a highly developed formulation of the main tenets of Dharmakīrti’s system. Their works provide a coherent and detailed analysis of our conventional understanding of everyday reality and argue that this conventional understanding is an erroneous fiction based on ignorance. Given that our conventional understanding is based on ignorance, Jñānaśrīmitra and Ratnakīrti then theorize the nature of liberation and how it can be attained under such conditions. There are several recurring topics in Ratnakīrti’s philosophical oeuvre. In A Proof of an Omniscient Being,1 he discusses the possibility of a buddha’s omniscience. His arguments about omniscience are also applied to the refutation of god-like beings and their characteristics in A Refutation of the Proof of God.2 His considerations of universal momentariness and of the way in which everyday activity seemingly engages with permanent objects were developed into a theory of existence that serves as the ontological basis for all his works. These considerations are found in A Positive Proof of Momentary Cessation, A Negative Proof of Momentary Cessation, A Refutation of the Proof of Enduring Things, A Proof of Exclusion, and A Theory of Multifaceted yet Non-Dual Appearance.3 The main subject of this last work and A Refutation of Other Continua4 is the philosophy of mind. In these two works, Ratnakīrti develops a typical Buddhist form of idealism, which results from his analysis of the mind as always having what he calls a multifaceted, yet non-dual appearance and his proof that we cannot know other minds. Writing in the tradition of the logico-epistemological school of Indian Buddhist philosophy, all his treatises showcase his mastery of logic. He explicitly discusses technical problems of epistemology and logic in A Treatise on the Existence of Other Means of Valid Cognition5 and in An Assessment of Pervasion.6 The following sketches will focus on his most prominent ontological theories. Both of these, momentariness and the non-existence of anything but the mind, fly in the face of everyday experience. Ratnakīrti, like his predecessors, was pushed to account for how the engagement with enduring, extra-mental objects could possibly be an illusion. DOI: 10.4324/9781351030908-50

573

Patrick McAllister

Momentariness and the Definition of Reality Ratnakīrti’s two Proofs of Momentariness seek to establish the truth of the following inference:7 To exist is to be momentary; for example, a pot exists and is therefore momentary. And these things under discussion do exist.8 To be valid, this kind of inference (a parārthānumāna; see Kajiyama 1998, 86) requires certain elements: First, a relation between a reason – here, “to exist” – and the probandum – here “to be momentary” – is stated to hold. This relation, called “pervasion” (vyāpti), implies that the reason cannot be true of some thing without the probandum being true of the same thing. Second, an example of such a relation is provided, which should be acceptable to both the proponent and the opponent. In the last step, the reason is stated to apply to the case under consideration. Ratnakīrti specifies that the reason used in this inference is one that consists in the proper nature of a thing (svabhāvahetu).9 We should therefore understand that, just like an oak is a tree by its very nature of being an oak, things that exist are momentary by their very nature; in other words, to exist is to cease or perish momentarily. Both of Ratnakīrti’s proofs of momentariness take up the validity of this inference. A Positive Proof of Momentary Cessation first discusses how to define “existence” and to what extent it qualifies the objects that the inference is about and then asks whether there are – and whether we can know about – cases in which things are momentary by their very nature of being existent.10 Ratnakīrti begins this discussion as follows: One cannot declare that a reason shows an unperceived object without first eliminating the doubts about whether the reason might be a pseudo-reason. And pseudoreasons are of three kinds: unestablished, contradictory, and ambiguous reasons.11 The core of the argument of A Positive Proof of Momentary Cessation consists in showing that the reason for inferring momentariness, “existence,” is not one of these three so-called pseudo-reasons (unestablished, contradictory, and ambiguous, a differentiation that long preceded Ratnakīrti). Broadly speaking, an unestablished reason is one that is not actually present in the object that the inference is about (it may be true that an oak is a tree, for instance, but if the object of the inference is not actually an oak, the inference does not apply). A contradictory reason is one that can never have joint presence with the property that should be established (so there must be at least one instance in which existence and momentariness, or “being an oak” and “being a tree,” occur together). An ambiguous reason is one that occurs both together with and apart from the inferred property and as such is not a sure indicator of the latter property (some existent things would be momentary, others not, like some oaks would be trees, others not). In other words, we can think of Ratnakīrti’s analysis of the three potential faults of the reason “existence” as investigating (1) the relevant definition of the property of exis­ tence and whether the things under discussion have this property, (2) the possibility that these existent things are momentary, and (3) the necessity that these existent things be momentary. Ratnakīrti’s investigation into whether “existence” might be an unestablished reason focuses on the analysis of existence. He first states that there is a generally acceptable definition of existence: the state of producing an effect (arthakriyākāritva). This definition of exis­ tence had been accepted within Buddhism at least since Dharmakīrti. According to Ratnakīrti, this definition can serve as common ground for himself and his opponents: on the one hand, it is the notion of existence implicitly presupposed by anyone who engages in everyday activities, and on the other, it is only this kind of existence that can be generally established by perception and inference, the two means of knowledge ( pramāṇa) acceptable to both Buddhists and virtually all of their opponents (see Kajiyama 1998, 30–38). Other definitions had been 574

Ratnakīrti

proposed by competing philosophical systems, yet, Ratnakīrti argues, since such definitions meet these two criteria only by resort to causality, it would be useless to discuss them in detail. Existence is thus analyzed in causal terms: to exist is to produce an effect. The next question to examine is how this characteristic is linked to momentariness. That is, can things that produce an effect be momentary? If this cannot be affirmed, then the purported reason “exis­ tence” would be a pseudo-reason of the second type, a contradictory one. If it were impossible for momentary things to produce an effect, then to say that something exists would preclude that thing’s momentariness. At this point, Ratnakīrti has to consider another logical peculiarity of the proof of momentariness. In the usual inference of its type, “This is a tree because it is an oak,” there are three groups of things that have well-defined relations to each other: the thing currently under consi­ deration, all or at least some other things for which it has been established that they are oaks and trees, and everything that is not an oak (though perhaps a tree). In the tradition of logical analysis that Ratnakīrti subscribes to, an inference generates knowledge only when the property used as a reason (“being an oak,” “existing”) is “pervaded by” the property to be established (“being a tree,” “being momentary”). This relation of pervasion can usefully be compared to sets: the instances of oaks are a proper subset of the instances of trees, and the set of trees contains other types of trees apart from oaks. If this is true, then the reason “being an oak” is proof of “being a tree”; in other words, “being an oak” is pervaded by (or subsumed under) “being a tree.” This approach generates two problems for the inference of momentariness. First, the set of existing things and the set of momentary things are, according to the claim of the inference, the same: there are no momentary things that do not exist, and no things exist that are not momentary. Second, there are no (existing) things that do not belong to this set. In other words, there is nothing that could be adduced as an example for being both existent and momentary. Apart from being a formal requirement of inferences in Ratnakīrti’s tradition, the question is larger: if there were no other corroborating instances, how could we be sure the inference is true? Since the inference is about all existing things, every “similar instance” – everything for which it is established that it is momentary, independently of this inference – would be included in the group of things that the inference is about and would, for that reason, not yet have been proven to be momentary (or the present inference would be unnecessary). Conversely, it is impossible to present anything that does not belong to this group and yet illustrate the inference through another case, since by definition such a thing would not exist. Ratnakīrti’s formulation of the inference of momentariness simply states “as in the case of a pot” as an example of an instance of momentariness in which the reason “existence” is present. His opponent asks how the pot could possibly be a similar instance: since the subject of the inference comprehends all things, the pot included, and the inference aims at proving its momentariness, the pot thus cannot be used as the example case for which momentariness is supposed to be already established. Hence the opponent claims that there is no way to establish the pot’s momentariness: perception does not establish it (we experience temporally extended things); nor does the inference from existence, as that inference is currently under scrutiny, and its employment for its own proof would lead to an infinite regress, and there is no other inference that can establish momentariness (and if there were, the present inference from existence would be useless). Ratnakīrti’s answer focuses first on the notion of the capability to produce an effect.12 For example, a pot which contains water is the cause of the water’s remaining in a certain place; consequently, it is evident that the pot is presently capable of carrying water. But what about past or future times in which the pot was or will be? If capability is determined through observed effect, then Ratnakīrti claims there are only three options: either the pot produces 575

Patrick McAllister

the very same effect of carrying water at those other times, it produces another effect, or it produces no effect at all. The first option can be dismissed without a long discussion – if true, it would then follow that the pot is now producing an effect that has already been produced, a contradiction in terms. The pot’s effect on the water yesterday, its spatially and temporally determined containment, is not being produced now. Even if the pot broke now, it would not spill the water yesterday. Thus the pot today or tomorrow cannot be considered capable of holding the water yesterday. The second option deserves closer consideration. Both Ratnakīrti and the opponent agree that one can coherently maintain that the pot is capable of producing the effect it is currently producing. But should one say that it is capable or incapable of the future or past effects? Evidently, for Ratnakīrti, if the pot produces the present effect because it is capable of it, and if it were currently capable also of producing the future effects, there would be no reason it should produce only the present effect but not the future ones in the present moment. Therefore, the pot must be considered currently incapable of producing the effects that one thinks it produced before or will produce later. However, to think that the same pot is both capable and incapable of the same thing (i.e., the future effect) would require assigning it contradictory properties. It would be more plausible to assume that the thing changes from something that is not capable to something that is. The third option, that the pot stops producing effects altogether, is incoherent. As the present pot producing a present effect shows, capability is temporally directly adjacent to the actual production of that effect. But if a thing we call “capable” were at some moment not causing anything at all, then this criterion would be contradicted and we should not call it “capable.” The only possibility that Ratnakīrti considers coherent is that the pot changes its state from not being capable of a certain effect to being capable of that effect. Implied in this notion of capability is that when something is capable of an effect, it must produce it – if it did not, there would be no reason to call it capable of it. The very idea of potential causation is thus rejected. Ratnakīrti formalizes and refines this argument with a logical tool that allows for inferences that have no examples. This tool consists of two parts: the formulation of an “unwanted consequence” ( prasaṅga) and the formulation of a “contraposition of the unwanted consequence” ( prasaṅgaviparyaya). The “unwanted consequence” is an inference showing that a certain assumption – usually one held by the opponent – implies an untenable commitment. The “contraposition” is an inference that, based on the negation of that untenable commitment, proves the opposite of the previous assumption. Ratnakīrti’s formulation of the unwanted consequence is as follows: That which can be treated coherently as producing something at some time must produce that thing at that time. Like the last group of causes in a causal line must produce its specific effect. And, according to you, this pot is something that is cohe­ rently treated as producing past or future effects, both at the time of producing the present effects as well as at the time after all effects have been produced. So there is an unwanted consequence based on a reason that is a thing’s nature.13 This reasoning is based on the opponent’s assumption that the pot stays the same throughout its usage, be it yesterday, today, or tomorrow. It can thus correctly be thought of, talked about, and used as the same single thing that produces effects in the past and in the future. But according to Ratnakīrti, if it really were always capable of those effects, it would always have to produce them in every moment, even after it has already produced them. There would be continuous causation of all effects. This is, obviously, not something that the opponent could accept. 576

Ratnakīrti

The argument seems to be liable to an obvious objection: ordinarily, we agree that the capability for an effect does not mean that this capability has to be exercised. A grain is, after all, commonly considered to be capable of producing a shoot, but it does not do so unless it is in the soil in the right conditions. Likewise, a present pot is commonly held to be capable of carrying tomorrow’s water. Ratnakīrti’s answer to this objection relies on the notions of causality and capability that were described previously: a thing cannot truly be called capable of producing a certain effect until the moment when it is actually producing that effect. In the moments leading up to that moment of production, the thing should be characterized only as “the cause of the cause” (kāraṇakāraṇatva) of the intended effect but not yet the cause of that effect. In the case of a grain, the property of “generating a shoot” is applicable to a grain only in that state where it actually is producing a shoot. Before that, it is gradually attaining – in combination with other factors like humidity, temperature, soil, and so on – a state in which it can properly be described as capable of producing a shoot. For Ratnakīrti, then, capability is synonymous with “being a cause.” Capability must always be exercised immediately, since there is no coherent explanation of why it should be exercised at one time rather than another. This line of argumentation is complemented with the following “contraposition of the unwanted consequence”: That which at some time does not produce something cannot at that time coherently be treated as being capable of producing something. Like millet, insofar as it does not produce a rice sprout, cannot be treated coherently as being capable of produ­cing a rice sprout. And this pot, at the time of producing the present effect and at the time after all its effects have been produced, does not produce a past or future effect. So the logical reason of not apprehending a pervading element separates the capable moment from the incapable moment.14 This reasoning aims to prove that when something does not produce a certain effect, it cannot be treated as capable of producing that effect. It is the contraposition of the previous reasoning in that it takes as its premise the negation of what was there the consequence and deduces the negation of this premise in the previous “unwanted consequence.” The absurd consequence would be that if it were correct to say that a pot were capable of producing past and future effects, the pot would have to produce all of these effects now. The contraposition concludes that since the pot does not produce these effects now, it is not correct to say that it is capable of producing these effects. This reasoning is intended to establish the corroborating example for the general inference of momentary cessation. These two supporting inferences jointly establish that the example in the inference of momentariness – the pot – is not always in the same state with regard to the production of an effect. And since the states of not being capable and being capable constitute contradictory states, the thing to which these states belong cannot be said to remain the same throughout such changes in states. A Negative Proof of Momentary Cessation complements A Positive Proof of Momentary Cessation by investigating whether there are dissimilar instances, cases where momentariness and existence are both absent. Since it is difficult, using the definition of existence as causal efficiency, to prove that some things lack the property of existence, the actual purpose of this second work is to show that the opposite is impossible. That is, non-momentariness must be shown to be incompatible with existence defined as cause and effect both within the model where a cause precedes its effect and within the model wherein cause and effect 577

Patrick McAllister

coexist. If that can be shown to be the case, then the only conclusion is to accept that exis­ tence implies momentariness.15

Concepts and Activity Without a World Ratnakīrti develops his epistemological framework in A Proof of Exclusion and A Theory of Multi­ faceted yet Non-Dual Appearance, as well as in additional discussions contained in A Refutation of Other Continua. This framework implies the ontological background established and explained previously in his texts on momentariness. It is useful to treat these texts in accordance with the degree to which the existence of external objects is tolerated in them: in A Proof of Exclusion, Ratnakīrti examines how everyday engagement with extra-mental, middle-sized objects is possible. In A Theory of Multifaceted yet Non-Dual Appearance and A Refutation of Other Continua, he proposes an analysis of the mind that results in a strong form of idealism and solipsism. The professed aim of A Proof of Exclusion is to establish the object or referent (artha) of words.16 By “referent of words,” as the discussion soon shows, Ratnakīrti actually means universals – in the broad sense of any commonness that the discontinuous and unrepeatable particulars proved in A Positive Proof of Momentary Cessation and A Negative Proof of Momentary Cessation are thought to share. These “universals” are the objects of any concept-driven activity, be it linguistic, mental, or physical. Any everyday activity that engages with temporally extended or similar objects has this “referent of words” as its object: to say “An apple is on the table,” to think a thought that can be expressed by those sounds, or to purposefully act in a way that is coherent with an apple being on the table are all activities that are engaged with “universals” in this sense. The apple that one thinks of as the same from the time it is first seen on the table to the time it is picked up to be eaten consists, according to the analysis of momentariness, of a causal continuum of apple-moments; “this apple” is already an abstraction of the individual moments, a result of what Ratnakīrti calls determination (adhyavasāya), the judgment that something (a sequence of apple-moments) is something else (a continuous apple). A continuous apple is closer in its nature to a universal (like cow-hood) than to a particular: both the continuous apple and cow-hood must be temporally extended entities. We have seen that, for Ratnakīrti, to be a real thing is necessarily to be momentary, discontinuous, and unrepeatable. Any notion that posits an object as either temporally extended, continuous, or repeatable does not accord with reality and is therefore erroneous. Nevertheless, such a notion can be practically true – reliable in that it enables activity that meets an agent’s expectations by directing her towards obtaining or avoiding desired or undesired effects of particulars. Concepts are “true enough” for someone to engage with particulars in everyday activity, but they do not let a person know those particulars in a stricter sense. What, then, is the nature of universals or temporally extended objects that are “practically true”? The answer is given in the theory of exclusion.17 This theory was first formulated by Dignāga around the fifth century and remained a point of much contention between Buddhists and rival groups for at least seven centuries. Its central postulation is that things are deemed similar to each other – be it in terms of outward appearance, causal effects, genus, or otherwise – because, in reality, they share the same difference from, or “exclusion from,” other things. Though this exclusion is often introduced as the referent of words, it is in fact the object of any conceptual cognition. An enduring middle-sized object like the apple on the table would thus be analyzed not as a substance in which certain slowly changing properties inhere but rather as something that is defined only through a set of distinctions from other things: momentary phases that are causally linked in a certain way so as to be judged a continuum. The phases constituting such a sequence can be judged to be the same in some way 578

Ratnakīrti

(say, as distinguished from what are not apples or from what is in a different place); any phase in such a continuum can cause, in someone already used to the idea of dealing with temporally extended objects, the notion of a slowly changing object in the same place. Ratnakīrti, like many authors in the tradition of Buddhist epistemology who have broached this subject, admits frankly that this analysis is very different from what everyday experience looks like to most people: moving about in the world, one does have a strong feeling that there are middle-sized, continuous objects. His justification of this gap relies on two arguments: (1) an ontological one that shows that, apart from exclusion, no other object of words and concepts is actually possible and (2) an epistemological one that explains how these appearances can cohere with the world as it really is. Ratnakīrti characterizes his opponents as proposing one of three kinds of objects as possible referents of words: either a particular, a true universal (like cow-hood) that may or may not be connected to a particular, or the form of awareness. This “form of awareness” (buddhyākāra) is the technical term for the object that any given awareness is directly aware of and with which it is identical. For the present discussion, it is best to think of it as the internal object of cognition. For Ratnakīrti, this form of awareness is a real thing like a particular, but not a mind-external, one. A pure particular, unqualified by a genus, and the form of awareness are discarded as viable candidates for the referent of a word without much discussion: speakers do not use the word “cow” for only one particular cow, be it a particular in the strict sense of the theory of momentariness or even in the loose sense of a single cow, and the form of awareness that a person has when hearing or saying a word is not what a speaker usually refers to. Both of these objects are useless as the referent of words: though they can be referred to, neither the one particular cow nor the imagined cow are things that can satisfy a central and observable function of the word “cow” – that it refers to all cows. Hence, A Proof of Exclusion consists mostly of refutations that a universal or genus is the referent of a word, either directly or as connected to – but still separable from – particulars. Ratnakīrti’s own definition of the word referent is this: “The word referent is a positive element qualified by other-exclusion.”18 With this formulation, A Proof of Exclusion effectively characterizes the referent of a word as something that has the structure of a property and its substrate (dharma and dharmin); the property in this case is other-exclusion (apoha), and that which is qualified by it is a “positive element.” Ratnakīrti tells us that this term, “positive element,” refers to two things: an external particular19 and the form that a particular awareness has. This internal form is also a particular, but not an extra-mental, one. While the external particular does not appear in conceptual cognition, whose objects are exclusions, both items are factors in any conceptual cognition and, as such, constitute two aspects of the object of conceptual cognition. Yet it is only the external particular that is considered the referent of a word (though, as we shall see, not directly but rather through its quality of other-exclusion). This follows from the fact that any predication that is performed in a conceptual cognition must refer to an external object. One of Ratnakīrti’s examples concerns the property “existence”: to attribute existence to the form of a particular awareness is redundant, since a present form must exist. To negate the present form as existent is incoherent because, again, predications that are performed in a conceptual cognition must refer to an external object. Predicating existence or its absence about external things is, however, useful and not redundant. The underlying reason for this argument is that the internal form is the object of a special kind of perception, that of awareness by itself (Ratnakīrti and his tradition call this selfawareness, svasaṃvedana). Insofar as a thing is known through perception, it is fully known: 579

Patrick McAllister

at least since Dharmakīrti, Buddhist epistemologists agreed that the distinction of properties and substances was not real and that perception shows its object as it really is; this means that any perception must show its object fully, with all the parts that might later be analyzed as properties or substrates. Ratnakīrti’s point is that this form appearing in awareness must, by definition, be known in all its aspects. Predications about it would thus be redundant: anything true that could be said about it would already be known, and anything false about it would be contradictory. A conceptual cognition does not apprehend an external thing through such a direct perception, by definition. Rather, the external thing is only indicated, and not shown, through determination. Determination is a judgment which, caused by the specific nature of the mindinternal particular form, itself causes activity directed at an object that is construed as external and enduring. Its truth consists in enabling everyday activity according to an agent’s expectations, even though the temporally extended object that the agent takes to be acting towards it does not exist. The two “positive elements” are keyed to two modes of awareness: the form of awareness is known through self-awareness, in a perceptual mode, and is thus known in all its particular characteristics; the external particular is known indirectly through a determination that, based on the particular form of awareness, instigates activity, the agent of which believes to be acting towards a temporally stable, unitary object. This fictional object might, if the cognition is “true enough,” align with an actual, external particular. Ratnakīrti’s definition of a word’s object attributes the property of other-exclusion to these two positive elements. This property is likewise analyzed as internal and external since it qualifies both positive elements. In the case of the external element, the exclusion from others is said to result from the specific set of causes that generate the particular: through these, the particular is “restricted” to a particular causal capacity that makes it useful (or not) for a particular aim. For Ratnakīrti, it is clearly coherent to say that two particulars can share an “exclusion from others” and that they have nothing real in common.20 This allows him to maintain, on the one hand, that two particulars are really completely different from each other and, on the other, that their treatment as having something in common is successful. For Ratnakīrti, this position is tenable because an “exclusion from others” is a form of absence or negation but not a substantially real universal (as it is for all his opponents). The property of “being a horse” can be absent from two particulars without those two particulars therefore sharing a substantially real property. They can thus be held to be excluded from horses without commitment to a real, shared property. What his opponents generally take to be the effect of universals – that certain lumps of matter share a common property like “being a cow” – is analyzed in terms of sets of differences. To be a cow is to be qualified by a certain set of differences from other things that are not cows. Insofar as a cow is an animal, it will share these differences with those other animals (usually, they are distinguished from horses). But to all these differences common to cows and other animals, there must be differences specific to cows (notably, a dewlap, horn, tail, and so on). Each cow, of course, has its further individual differences from all other cows. And, if analyzed carefully, each moment of an individual cow’s existence has its own set of differences from all other moments in that causal chain. This theory of universals is thus fully compatible with the theory of momentariness. Ratnakīrti does not even pretend that this is “how things appear” in conceptual cognition. All these sets of differences, which should be infinitely many, do not appear in cognition at all. Ratnakīrti’s solution to this is a rather surprising one: the form of awareness, the mind-internal positive element described previously, is what appears, yet there is always a simultaneous grasping of the differences from other things. This apprehension of the differences is equated by Ratnakīrti to a capacity (śakti), modeled on the stock example of the grasping of absence: 580

Ratnakīrti

as the perception of an empty stretch of floor is capable of generating infinite ascertainments (all of the form “X is not here”), so the internal positive element, shown by self-awareness, contains the ability to form such ascertainments. This means that to be a competent user of the word “cow” is to be able to, upon hearing the word “cow,” act (in speech or otherwise) according to that set of differences that the word “cow” is accepted to refer to. Such a person could say both “This is not a horse” and “This is an animal,” or, to use Ratnakīrti’s example of hearing the command “Tie up the cow!”, tie up a cow but not a horse. The emphasis here is on the possibility because not every difference has to come to awareness every time that the word is heard. In other words, the form that an awareness takes at any particular time is the basis for a certain set of possible predications of differences. This also suggests that the attribution of “other-exclusion” to the “positive element” must be understood ontologically not only in the case of the external particular but also in the case of the form of awareness, where it could also be understood as concerning the content of the cognition: the external particular and the form of awareness are both particulars in that they are caused by a specific set of causes and for that reason also have particular causal capacities. What appears to cognition whenever the word “cow” is heard (or used) need not be the same every time;21 the forms of these awareness-events just need to be similar in the sense that they cause a person to act in accordance with the same set of differences. If there is such an act, it will have been caused by an awareness whose causal capacity is defined by the causal capacity of the awareness that arose on hearing (or using) the word “cow.” Hearing a statement such as “Tie up the cow!” twice will, in a listener who understands the statement, twice create an awareness whose form is such that it delimits the activities compatible with the command to tying up the cow; the listener would also, in all likelihood, say that she has received the same command, understood it the same way, and did the same thing in accordance with it. But none of these instances of sameness are real. Instead, two spoken sentences caused two awareness-events that, due to the particular characteristics of the awareness at that time (i.e., previous experiences that influence the person’s understanding of that command, along with a deep habituation to erroneous beliefs that there are enduring, extra-mental objects), resulted in two actions that one might judge to be the same (i.e., whose observation would lead to other awareness-events that share similar sets of differences). The last important feature of Ratnakīrti’s theory of exclusion is that he lays great stress on the simultaneity of the appearance of the positive element and its quality, the exclusion from others. It is not possible to say that one of them is the primary and the other the secondary element in a conceptual cognition, in terms of the sequence in which one becomes aware of them. Since Ratnakīrti does not accept a real difference between properties and the things they qualify, they are also not known apart from each other: to perceive something is to perceive both what can be analyzed as the substrate and what can be analyzed as its properties in the exact same moment. This holds true for the substrate that is the form of cognition and its ephemeral quality, other-exclusion, since the substrate is known through a perception of the type self-awareness. For the external thing with its quality, there is no discussion of the simultaneous perception of the property and its bearer because it is not directly perceived through any conceptual cognition. Ratnakīrti illustrates this with the word indīvara, which is a name for blue water lilies that cannot be analyzed as a compound of two parts, one for “blue” and one for “water lily”: someone who knows what the word means does not become aware of two different things in sequence, the color blue and a water lily (as could perhaps be the case if one heard the phrase “blue water lily”); rather, there is a simultaneous awareness of both the color blue and a water lily. It is the same in the case of the word “cow,” which refers to what is 581

Patrick McAllister

different from non-cows: hearing it, a competent user of that word will have a form of cognition that is qualified by the according exclusion of non-cows. As stated before, this quality is a capacity of the cognition, not an actual presentation of everything that is not a cow. The theory of other-exclusion is thus a device to bridge the apparent gap between the realm of particulars and that of conceptual cognitions, which appear to deal with universals, unreal insofar as they are devoid of causal ability. This theory reinterprets these universals as the differences or exclusion from others. They are the only general things acceptable to Ratnakīrti because they are not real things in the causal sense. In the final analysis of our interaction with everyday objects, Ratnakīrti would conclude that there is no gap because the conceptual can be fully reduced to particulars as well: the infinitely many sequences of causes and effects that constitute the extra-mental world and the infinite causal sequence that constitutes awareness. Any lasting or repeatable phenomenon is unreal and ephemeral, figuring in any cognition only as a horizon at which activity may be directed.

The Transition to Idealism This justification of the difference between how everyday objects purportedly appear and how they are reasonably analyzed allows Ratnakīrti to make two important further points: first, he can both explain and deconstruct everyday experiences of extra-mental objects in terms of awareness-events that are erroneously directed “outward,” facilitating a move towards idealism; second, he can tie the Buddhist theory of nescience and the possibility of liberation to the very core of everyday experience. Further details of this strategy can be culled from A Theory of Multifaceted yet Non-Dual Appearance. This text is a highly condensed and focused treatise based on Jñānaśrīmitra’s works that discuss the teaching that awareness has a real form (sākāravāda) and whose central concerns – to show that awareness must always have a form; what that form is; and what this means for the analysis of cognition, reality, and liberation  – are not yet well understood.22 There is no doubt, however, that, at least for A Theory of Multifaceted yet Non-Dual Appea­ rance, there is a clear link between the following tenets: cognition always has a form of a well-defined kind; determination (the primary result and expression of the nescience of unenlightened beings) works without dependence on any extra-mental, real entities; and liberation and its opposite, saṃsāra, can be understood in terms of determination. A Theory of Multifaceted yet Non-Dual Appearance discusses these matters in a tightly defined context: Ratnakīrti assumes that any and all extra-mental objects have been proven to not exist; all theories that awareness does not have or show a form at all (or that its form is somehow ephemeral or unreal) have been refuted; and the whole world has been established as only the appearance of a form to awareness, like what appears in a dream.23 A rather specific form of idealism is thereby presupposed, in which external objects are like objects in a dream, devoid of extra-mental existence but real – causally efficacious – in the context of the dream or insofar as they are cognitions. The main thesis that Ratnakīrti seeks to establish in A Theory of Multifaceted yet Non-Dual Appearance is that awareness has a single yet variegated form.24 Typically, this point is presented as an inference: What appears, that is one. Like the form of blue occurring amidst other forms. And this collection of various forms, such as white, the sound “ga,” sweet, fragrant, soft, pleasure and its opposite, etc., appears. This is a proof using an essential nature as a reason.25 582

Ratnakīrti

This inference is formally valid: to appear (the reason) is to be one (the probandum). The example supplied is an apprehension of the color blue, as blue never appears by itself but instead appears always together with other “forms” (shapes, sounds, etc.). The reason is then said to apply to a variegated form of awareness, all aspects of which appear together at the same time and therefore must be a unit. The discussions focus largely on the unity the form of cognition possesses: according to Ratnakīrti, everyone agrees that something appears and that it is something manifold that appears – after all, everyone experiences a world of variegated objects. But Ratnakīrti claims that what appears to be manifold is actually a single, unitary thing. In being aware of something blue, one is not aware only of blue: blue appears amidst a group of other appearances (both external ones like sounds, smells, shapes, as well as pleasure and other factors attributable to the subject of the experience). Yet in the end, the whole world, including its temporal and spatial distinctions, is nothing but a single, variegated form that is of the nature of awareness (see Kajiyama 1965, 12). The opponent objects to Ratnakīrti’s inference that the example (the form of blue) is actually not one with the variegated yet unitary form of awareness that is under discussion in the inference since the two forms, the example and the subject of the inference, appear at different times in different places and look different. But if they were identical, the example would not be admissible since the example in an inference needs to be distinct from the case that the inference is about.26 Ratnakīrti’s answer is that the counter instance – a form that would not be unitary but less or more than one – is known to not appear. How is this known, the opponent immediately asks, since in order to know this counter instance, it would certainly have to appear to cognition? The trap is obvious: if it appears and is not one, the main inference is shown to be based on a wrong relation; if it does not appear, then how can we ascertain that it is not one? Ratnakīrti responds that the opponent cannot possibly mean that it should appear directly: insisting on that requirement would make most inferences impossible, since certainty that the reason is never present where the property to be proven is absent is a precondition for inference (e.g., the certainty of the absence of oak where there is no tree), but how could this co-absence ever be certain if it were required that all cases where the property to be proven is absent must appear to awareness? The counter instances are presented, rather, through determination. In this case, the actual counter instances need not appear but are cognized indirectly, as a group of cases for all of which this co-absence is true; in other words, all counter instances are known by a unifying general feature, the absence of the reason and the probandum. The inference that lets one know this group in the present case is of the same type as the one used to establish the momentariness of the example in the inference to general momentariness.27 Ratnakīrti’s discussion of determination presents itself as an explanation of how a counter instance can be known. The larger issue, however, is how determination makes it possible to apprehend something other than cognition without this other object having to be really different from cognition. Awareness is always one, and therefore anything that appears in it would be both real and identical with it, insofar as it appears in it. Yet insofar as what appears in awareness is determined – that is to say, produces activity that is directed at something that is not awareness – one may speak of external objects. This is how one can meaningfully act upon things with an ontologically problematic status, that is, past and future things, impossible things, and things whose existence is doubtful. They are real as appearances to awareness, but they are not real as they are acted towards. The main inference of A Theory of Multifaceted yet Non-Dual Appearance allows Ratnakīrti to give a clever example of this. The inference holds only if it can be shown that what is not of a single form – that is, any counter instance – cannot appear. To establish this, however, the counter instance must be known in some way. This is the puzzle that must be solved by 583

Patrick McAllister

determination: it should give cognitive access to something that is not directly present and that does not have to exist in the way that it is accessed, even though it must exist in some way so that it can be accessed. The basis of Ratnakīrti’s solution was already described in the treatment of other-exclusion previously: the form of awareness directly appears and has the capacity to cause activity that conforms to its properties (the other-exclusions that qualify it). Speaking from the position of radical idealism in A Theory of Multifaceted yet Non-Dual Appearance, Ratnakīrti adds to this analysis that the form of awareness is not really caused by external things at all but only by awareness itself, given that external things do not exist and hence cannot cause anything. Likewise, determination may direct activity outwards, and one of those activities can be acting towards an object in conformance with the form of awareness. But determination neither is caused in any way by an external thing, nor does it show an external thing, nor does it let anyone really obtain an external thing. All it truly does is give rise to the next momentary phase of awareness, continuing a process that only the achievement of liberation can end. Liberation is defined accordingly: determination is saṃsāra; its destruction liberation (mokṣa).28 Determination is thus not only a central function within Ratnakīrti’s epistemology but also a function of the nescience that makes living beings stay within the cycle of existence. Determination’s main operations, externalizing and generalizing objects, are what makes everyday activity possible but also what makes everyday activity an expression of the suffering of unenlightened beings. To end determination is thus, for Ratnakīrti, to attain liberation.

Notes 1 “A Proof of an Omniscient Being (Sarvajñasiddhi)” (Thakur 1975, 1–31). 2 “A Refutation of the Proof of God (Īśvarasādhanadūṣaṇa)” (Thakur 1975, 32–57; see Patil 2009). 3 “A  Positive Proof of Momentary Cessation (Kṣaṇabhaṅgasiddhiḥ – anvayātmikā)” (Thakur 1975, 67–82); “A  Negative Proof of Momentary Cessation (Kṣaṇabhaṅgasiddhiḥ – vyatirekātmikā)” (Thakur 1975, 83–95); “A Refutation of the Proof of Enduring Things (Sthirasiddhidūṣaṇa)” (Thakur 1975, 112–28); “A Proof of Exclusion (Apohasiddhi)” (McAllister 2020, 47–82); “A Theory of Multifaceted yet Non-Dual Appearance (Citrādvaitaprakāśavāda)” (Thakur 1975, 129–44). 4 “A Refutation of Other Continua (Santānāntaradūṣaṇa)” (Thakur 1975, 145–49). 5 “A Treatise on the Existence of Other Means of Valid Cognition (Pramāṇāntarbhāvaprakaraṇa)” (Thakur 1975, 96–105). 6 “An Assessment of Pervasion (Vyāptinirṇaya)” (Lasic 2000). 7 Mimaki (1976, 9–45) has shown how Ratnakīrti recasts the historical discussions of the Buddhist theory of momentariness as a logically ordered whole. He is shown to have separated the discussions in the Examination of Enduring Entities (Sthirabhāvaparīkṣā) and the Examination of the Relation of Action and Result (Karmaphalasambandhaparīkṣā) of Śāntarakṣita’s “A Summary of What Is the Case (Tattvasaṅgraha)” (Krishnamacharya 1926), an encyclopedic work of Sanskrit classical philosophy, into three works: A Positive Proof of Momentary Cessation, A Negative Proof of Momentary Cessation, and A Refutation of the Proof of Enduring Things. It is reasonable to assume that the characteristics described by Mimaki (1976, 9–45) would apply also to Jñānaśrīmitra’s “Section on Momentary Cessation (Kṣaṇabhaṅgādhyāya)” (Thakur 1987, 1–159) so that Ratnakīrti’s creativity would lie, as usual, in the reorganisation and compression of his teacher’s material rather than in the masterful analysis of the history of the arguments for momentariness. For a very useful tabular overview of the correspondences, see Mimaki 1976, 45. 8 A Positive Proof of Momentary Cessation 67. 9 See A Negative Proof of Momentary Cessation 83. 10 This question marks the main difference to the second text, A Negative Proof of Momentary Cessation, which discusses whether there are (and whether we can know) instances in which existence is present but not momentariness. 11 See A Positive Proof of Momentary Cessation 67. 12 This paragraph and the next are essentially a paraphrase of A Positive Proof of Momentary Cessation 68, though the illustration of the pot containing water has been added. 13 See A Positive Proof of Momentary Cessation 68. See Mimaki 1976, 56 for an analysis of this inference.

584

Ratnakīrti 1 4 See A Positive Proof of Momentary Cessation 69. 15 This is a paraphrase of Ratnakīrti’s closing statement in A Negative Proof of Momentary Cessation 95. 16 A Proof of Exclusion, 47: “Exclusion is declared as the referent of words.” 17 See Patil 2003; Patil 2009; McAllister 2020 for the following paragraphs. 18 A Proof of Exclusion, 49. 19 That the “external positive element” is a particular is the present author’s interpretation (see McAllister 2020, 235–58). Patil (2009, 281–83) understands Ratnakīrti to be talking about an external, but general, object, a continuum as a whole in the case of everyday objects. 20 Compare also Dharmakīrti’s famous example of the medical herbs in “Inference for Oneself (The First Chapter in the Additional Explanations to Dignāga’s Compendium of Means of Valid Cognition)” in Gnoli (1960, 74), translated by Eltschinger et al. (2018, 85) as: “Or else, to give another example ( yathā), one observes that in spite of their diversity, certain plants and not others are capable, whether individually or collectively, of alleviating fever, etc.” 21 But Ratnakīrti does clearly assume that a sort of generalized image-and-speech-sound appears in hearing words (A Proof of Exclusion, 60). But insofar as this is always a particular form of awareness, it cannot ever be truly the same as another. 22 See Moriyama 2011, 2012 for a translation into Japanese (which the present author does not read) and the introduction to Yūichi Kajiyama 1965 for a general characterization of the form and history of the idealism defended by Ratnakīrti. For Jñānaśrīmitra, see “A Systematic Exposition of the Proof of Cognition Having a Form (Sākārasiddhiśāstra)” (Thakur 1987, 367–513) and “Collected Statements on Cognition Having a Form (Sākārasaṃgrahasūtra)” (Thakur 1987, 515–78). Ratnakīrti’s A Theory of Multifaceted yet Non-Dual Appearance also includes material from Jñānaśrīmitra’s “Treatise on Exclusion (Apohaprakaraṇa)” (Thakur 1987, 201–32; trans. in McCrea and Patil 2010), attesting to the close connection of the topics. 23 See Kellner and Taber 2014 for a general characterization of this form of idealism. Ratnakīrti bases himself on Dharmakīrti’s sahopalambhaniyama argument, which, on the principle that two things that always occur together are identical, aims to establish that cognition and its object are not different from each other. Ratnakīrti quotes Dharmakīrti’s “First Chapter in the Ascertainment of the Means of Valid Cognition (Pramāṇaviniścayaḥ)” verse 54ab (Vetter 1966) at A Theory of Multifaceted yet Non-Dual Appearance 129. 24 To say that awareness “has a form” is to say that awareness shows that form or that one is aware of the form. The sākāra accepts that awareness is self-illuminating and holds that awareness really assumes the form of its object. 25 A Theory of Multifaceted yet Non-Dual Appearance 129. See Kataoka 2017; McAllister 2017a for investigations of the sections in which determination is discussed. “The sound ‘ga’ ” is the third note of a common solmization in Indian classical music. 26 For the background of this position (technically called bahirvyāptivāda), see Bhattacharya 1986. 27 See Kajiyama 1998, 116 ff., n. 310 for a succinct description of such an inference. 28 See A Theory of Multifaceted yet Non-Dual Appearance 137, 9–16.

Bibliography Bhattacharya, Kamaleswar. 1986. “Some Thoughts on Antarvyāpti, Bahirvyāpti, and Trairūpya.” In Buddhist Logic and Epistemology. Studies in the Buddhist Analysis of Inference and Language, edited by Bimal Krisha Matilal and Robert D. Evans, 89–105. Dordrecht: Reidel. Eltschinger, Vincent, John Taber, Michael Torsten Much, et al. 2018. Dharmakīrti’s Theory of Exclusion (apoha). Part 1, On Concealing: An Annotated Translation of Pramāṇavārttikasvavṛtti 24, 16–45, 20 (Pramāṇavārttika 1.40–91). Tokyo: The International Institute for Buddhist Studies. Gnoli, Raniero, ed. 1960. The Pramāṇavārttikam of Dharmakīrti: The First Chapter with the Autocommentary; Text and Critical Notes. Rome: Istituto Italiano per il Medio ed Estremo Oriente. Kajiyama, Yūichi. 1965. “Buddhist Solipsism – A Free Translation of Ratnakīrti’s Saṃtānāntaradūṣaṇa.” Journal of Indian and Buddhist Studies (Indo-Gaku Bukkyō-Gaku Kenkyū) 13 (1): 420–35 (9–24). ———. 1998. An Introduction to Buddhist Philosophy: An Annotated Translation of the Tarkabhāṣā of Mokṣākaragupta. Reprint with corrections in the author’s hand. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien.

585

Patrick McAllister Kataoka, Kei. 2017. “Dharmottara’s Notion of āropita: Superimposed or Fabricated?” In Reading Bhaṭṭa Jayanta on Buddhist Nominalism, edited by Patrick McAllister, 217–50. Vienna: Austrian Academy of Sciences. Kellner, Birgit, and John Taber. 2014. “Studies in Yogācāra-Vijñānavāda Idealism I: The Interpretation of Vasubandhu’s Viṃśikā.” Asiatische Studien/Études Asiatiques 68 (3): 709–56. Krishnamacharya, Embar, ed. 1926. Tattvasaṅgraha of Śāntarakṣita With the Commentary of Kamalaśīla, 2 vols. Baroda: Central Library. Lasic, Horst. 2000. Ratnakīrtis Vyāptinirṇaya: Sanskrittext, Übersetzung, Analyse. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. Matilal, Bimal Krishna, and Robert D. Evans, eds. 1986. Buddhist Logic and Epistemology. Studies in the Buddhist Analysis of Inference and Language. Dordrecht: Reidel. McAllister, Patrick. 2017a. “Competing Theories of Conceptual Cognition. Dharmottara and Trilocana vs. Dharmakīrti?” In Reading Bhaṭṭa Jayanta on Buddhist Nominalism, edited by Patrick McAllister, 291–321. Vienna: Austrian Academy of Sciences. ———, ed. 2017b. Reading Bhaṭṭa Jayanta on Buddhist Nominalism. Vienna: Austrian Academy of Sciences. ———. 2020. Ratnakīrti’s Proof of Exclusion. Vienna: Austrian Academy of Sciences. McCrea, Lawrence J., and Parimal G. Patil. 2010. Buddhist Philosophy of Language in India: Jñānaśrīmitra on Exclusion. New York: Columbia University Press. Mimaki, Katsumi. 1976. La réfutation bouddhique de la permanence des choses (sthirasiddhidūṣaṇa) et la preuve de la momentanéité des choses (kṣaṇabhaṅgasiddhi). Paris: Institut de Civilisation Indienne. Moriyama, Shinya. 2011. “An Annotated Japanese Translation of Ratnakīrti’s Citrādvaitaprakāśavāda (1).” South Asian Classical Studies 6: 51–92. ———. 2012. “An Annotated Japanese Translation of Ratnakīrti’s Citrādvaitaprakāśavāda (2).” South Asian Classical Studies 7: 135–67. Patil, Parimal G. 2003. “On What It Is That Buddhists Think About – apoha in the RatnakīrtiNibandhāvali.” Journal of Indian Philosophy 31 (1–3): 229–56. ———. 2009. Against a Hindu God: Buddhist Philosophy of Religion in India. New York: Columbia University Press. Thakur, Anantalal, ed. 1975. Ratnakīrti-nibandhāvaliḥ: Buddhist Nyāya Works of Ratnakīrti. 2nd  ed. Patna: Kashi Prasad Jayaswal Research Institute. ———. 1987. Jñānaśrīmitranibandhāvali: Buddhist Philosophical Works of Jñānaśrīmitra. 2nd  ed. Patna: Kashi Prasad Jayaswal Research Institute. Vetter, Tilmann. 1966. Dharmakīrti’s Pramāṇaviniścayaḥ 1. Kapitel: Pratyakṣam. Einleitung, Text der tibetischen Übersetzung, Sanskritfragmente, deutsche Übersetzung. Vienna: Austrian Academy of Sciences.

586

36 RATNĀKARAŚĀNTI The Illumination of False Forms Gregory Max Seton

Introduction Renowned as the “Omniscient One of the Degenerate Age” (Sankrtyayana 1935, 35 n. 4), Ratnākaraśānti (c. 970–1045 CE) was considered one of the foremost Indian Buddhist scholars and tantric adepts (siddha) of his time (Guide, 231a1; Chimpa and Chattopadhyaya 1990, 295, 313; Roerich 1988, 206).1 But his signature contribution to philosophy was his unique ontoepistemological answer to the Mahāyāna buddhological problem of how a buddha, who perceives the ultimate reality, can communicate with and bring benefit to ordinary persons, who perceive merely conventional reality. Since Ratnākaraśānti’s answer to this question relies on his buddhological assertion that a buddha retains cognitive error (bhrānti), it impli­citly contradicts the foundational Buddhist premise of a buddha’s infallibility and cognitive perfection and, not surprisingly, was considered controversial and criticized by most subsequent scholars. But his tantric commentaries, into which he often seamlessly wove this philosophical viewpoint, appear to have been widely appreciated for their ability to address the many pressing contemporary soteriological questions, such as why and how an obviously fallible human being might attain buddhahood within a single lifetime through antinomian tantric practices. This chapter will focus on explaining Ratnākaraśānti’s philosophical position, which claims to be the correct interpretation of both the Madhyamaka philosophy founded by Nāgārjuna and the Yogācāra philosophy explained by Asaṅga and Maitreya (Guide, 231a3). By claiming this, Ratnākaraśānti distinguishes his own position from the positions of prior Buddhist scholars. Among these, he explicitly calls Dharmottara and Śubhagupta materialists (bahirārthavādin) and refutes them (Proving Mere Representation, 309a2–3). He also refutes the Yogācāra viewpoint of Prajñākaragupta as nonsensical (Proving Mere Representation, 309a5–7; Instructions, 143a6 – b2, 148b1–49a5; Nondual Variegated, 137) and the “pseudo-Mādhyamika” (dbu ma ltar snang ba) Candrakīrtian viewpoint as nihilistic materialism (Guide, 226b4; Instructions, 143b4). Despite his finding more common ground with the epistemological position of Kamalaśīla, Ratnākaraśānti rebuts its refutation of the ultimate reality of mind (Instructions, 151b1–6).2 Before explaining Ratnākaraśānti’s philosophical intervention, we will examine the central Mahāyāna problem of how an awakened buddha’s immanent awareness can know what an ordinary confused person perceives.

DOI: 10.4324/9781351030908-51

587

Gregory Max Seton

The Immanent Awareness Problem and Ratnākaraśānti’s Intervention According to the Mahāyāna, ordinary persons needlessly suffer in cyclic existence, because, just as in a dream, their sleeplike ignorance causes them to cling to a dreamlike personal identity that experiences a constant state of existential dis-ease (saṃskāraduḥkhatā). However, a buddha (lit. Awakened One) overcomes this ignorance and suffering by directly perceiving the ultimate reality of all dreamlike phenomena perceived by ordinary persons. After awakening, a buddha comes to have three so-called bodies, which are ultimately not distinct but are merely the transcendent and immanent aspects of buddhahood perceived as different bodies through the subjectivity of unawakened ordinary persons, partially awakened beings (bodhisattva), and fully awakened ones (buddha) (Thurman 2004, 93–97; Mahāyāna Scriptures, ix.56–66). First, a buddha’s consciousness is transformed into a nonphysical “natural body” (svābhāvikakāya) – also called the true body (dharmakāya) – which embodies the true nature of a buddha that she only knows. Due to a buddha’s prior aspirations to help others awaken from their dreamlike states, the transcendent natural body manifests in the perception of partially awakened bodhisattvas as an immanent body of enjoyment (sāṃbhogikakāya) and in the perception of unawakened ordinary persons as an immanent body of emanation (nairmāṇikakāya). Despite agreeing that buddhas manifest three bodies to help awaken the two other types of beings, the two main Mahāyāna traditions, Madhyamaka and Yogācāra, debated what it is like for a buddha to fully overcome ignorance and still know the reality of deluded persons in order to communicate with them and lead them out of their own dreamlike cyclic existence. On the one hand, the Mādhyamika skeptics, who claimed to refute all ontological reals, essentially sidestepped the epistemological question by suggesting that what a buddha is truly like is beyond all conceptual proliferations (niṣprapañca) and that his so-called bodies only appear from others’ perspectives to be communicating with them. On the other hand, the monistic Yogācāra schools – who held consciousness to be the only ontologically real thing – offered nuanced epistemological explanations of how the immanent awareness of a buddha could perceive simultaneously the ontological reality accessible only to buddhas and the conventional reality perceived by ordinary deluded persons. According to Yogācāra ontology, while ordinary persons may feel as though they experience things external to consciousness, those things cannot be established as anything other than mind. That is to say, ordinary persons may think that they are seeing various colored things that exist somehow in the world. But their consciousness (vijñāna) – that is, the intentional “being conscious of ” something – is merely illuminating and knowing perceptual content (ālambana) in “representational forms” (ākāra) that are made of a mental substance rather than material substance external to mind. Although dualists might argue that external material things can be inferentially established as producing those representational forms, the Yogācāra scholars respond, in effect, that the putative correspondence between external phenomena and their mental representations could never be established, since consciousness, which has no way to access anything other than its own representational forms, could never know anything outside itself or verify the correspondence between its own mental representations and the supposedly external referents of those representations (e.g., Twenty Verses 3, 30). Furthermore, since the representational forms must be mental in substance and, at the very least, shaped by concepts and linguistic structures, they “represent” only the content of mind’s latent impressions produced by prior mental activities (karma) that take on particular forms due to ignorance, like appearances in a dream. Although all Yogācāra scholars agree that the external things imagined by ordinary persons are ontologically grounded in nothing but consciousness, the two main camps of Yogācāra 588

Ratnākaraśānti

epistemology disagree on what that consciousness is ultimately like, especially once its basis undergoes a fundamental transformation (āśrayaparāvṛtti) upon awakening from that dream. Among the two camps, the “Form-bound” (sākāra) scholars claim that consciousness is intentional by definition and must be ultimately “bound to” representational forms.3 Their rivals, the “Form-free” (nirākāra) scholars, argue that the consciousness of an awakened being is intentional but in a way that is different from ordinary beings and ultimately “free” from representational forms. On the surface, the “Form-bound” and “Form-free” disagreement may seem to revolve only around the ontological existence or nonexistence of representational forms at the level of a buddha. But, in fact, their epistemologies entail fundamentally divergent understandings of the nature of representational forms, and these are reflected in their rival theories of consciousness and error. For Form-bound scholars such as Dharmapāla (sixth century), the fundamental cognitive error in an ordinary person’s consciousness leads to a conceptual misinterpretation of the representational forms, which are viewed as ontologically distinct from consciousness. But a buddha’s consciousness, which eliminates this cognitive error, sees the very same representational forms that ordinary beings see, except it apprehends them in a more accurate way without superimposing externality and other conceptual proliferations. Hence, a buddha’s consciousness is intentional, like that of an ordinary person. But a buddha’s consciousness, rather than misinterpreting the nature of the representational forms, is reflexively aware that its intentional objects are nothing but mind. For Form-free scholars such as Sthiramati (sixth century), the fundamental cognitive error is a drifting (viplava) of consciousness into the false and nonexistent representational form of a grasping subject (grāhaka) and its grasped objects (grāhya). The characteristics of a grasping subject and its grasped objects are shot through with conceptualization and only imagined ( parikalpita) to be real. But since the term “representational form” refers to a moment of consciousness in the mode of grasping momentary content (ālambana), its referent is only the grasping subject, not the grasped object. In other words, the representational form of a grasping subject is simply a false appearance (vitathapratibhāsa) of real consciousness, but a grasped object, which is not a representational form, has nothing true or real about it (Middle-BeyondExtremes, 16). In any event, when the fundamental cognitive error – on which the imagined nature of the grasping subject and grasped object are dependent ( paratantra) – dissolves, then both subject and object are eliminated and consciousness’s own inherent nature is realized ( pariniṣpanna) to be free of subject-object structuring (Middle-Beyond-Extremes, 84–85, 143; Instructions, 138a4 – b7).4 Hence, unlike the Form-bound notion, the Form-free scholars hold that for a buddha at the ultimate level, consciousness is free of subject-object duality and is thus intentional only in the sense that the sheer existence of all phenomena manifests directly within a buddha’s own mind (Textual Commentary ad 3627, 6–8; Funayama 2007, 194–96).5 Skeptical critics of these two Yogācāra onto-epistemologies, like Kamalaśīla, pointed out many weaknesses and incoherencies in their explanations of what consciousness is ultimately like and how a buddha’s mind could be aware of the nature and variety of limitless objects and still communicate with ordinary people perceiving different particular conventional realities (Textual Commentary ad 535, ad 3627). In brief, they argue that the Form-bound epistemology is coherent, but its ontology fails to fully account for the transcendent qualities of a buddha’s natural body, whereas the Form-free ontology is clear, but its epistemology fails to account for the immanence of a buddha’s bodies of enjoyment and emanation. Thus, as an intervention, Ratnākaraśānti’s somewhat hybrid position, which agrees with and responds to Kamalaśīla’s critiques, can be seen as accepting the basic Form-free ontology but making some limited concessions to Form-bound epistemology so as to explain the mechanism through which a buddha interacts with ordinary persons. 589

Gregory Max Seton

Ratnākaraśānti’s philosophical intervention, contra Sthiramati, explains the grasped object and a grasping subject as two types of representational forms that appear in consciousness and argues that a buddha’s reflexive awareness is ultimately free of these. But Ratnākaraśānti also asserts that – by deliberately retaining a small amount of cognitive error – a buddha’s immanent awareness perceives the two representational forms, which she reflexively knows to be unreal and false, so that she can keep an epistemic connection to ordinary persons. He states: After the representational forms dissolve into a transcendent (lokottara) awareness, the transcendent awareness itself arises as an all-pervasive buddha free of appearances, free of the two representational forms, and free of conceptual proliferations. Even though her pure immanent (laukika) awareness has representational forms, those forms are discerned to be false and unreal, and, in this sense, that awareness is said to be free of representational forms. (Guide, 226a7) In Ratnākaraśānti’s position, when a buddha realizes the ontological reality of the dream, she does not exit to some awakened world beyond the dream where she knows the mere existence of things directly. Rather, her transcendent awareness is inseparable from an immanent awareness that compassionately manifests within the dream world, like a lucid dreamer who interacts with the dream-forms of ordinary persons in the shared dream-space of conventional reality. Thus, for Ratnākaraśānti, since a buddha chooses to participate in the dream and knows the dream-forms to be false and unreal, she can still be understood to be ultimately “free.” Although Ratnākaraśānti himself characterizes his unique viewpoint as a sub-position within the Form-free (nirākāra) school  – where he incidentally also places Kamalaśīla’s Mādhyamika position – Ratnakīrti (eleventh century) and others classify Ratnākaraśānti’s position as “False-form” (alīkākāra) and differentiate it from the paradigmatic Form-free position associated with Sthiramati and others (Guide, 226b2; Nondual Variegated, 122). Since many modern and traditional scholars incorrectly conflate these two distinct positions and use the names Form-free and False-form interchangeably, it must be emphasized here that this chapter follows Ratnakīrti in using the terms to refer to very different viewpoints (Kajiyama 1965, 426–25).6 Now that these two names have been distinguished, the following two main sections will focus on the two main pillars of his onto-epistemological argument – namely his theory of the fundamental nature of consciousness and his implied theory of error – which aimed to bridge some of the gaps between the rival ontologies of Yogācāra monists and Mādhyamika skeptics and between the two divergent representationalist epistemologies.

Ratnākaraśānti’s Ontological Account of Consciousness When arguing that all phenomena are nothing but consciousness, Yogācāra scholars often compare consciousness to a lamp that illuminates itself while illuminating other things. For Form-bound scholars, the two functions of consciousness  – namely its “illumination” that allows phenomena to appear and its “awareness” that knows them – are like two equal sides of the same coin. Hence, the analogy of a lamp also means that awareness knows itself reflexively as it knows other things. But Ratnākaraśānti’s ontological argument treats illumination ( prakāśa) as a nature of consciousness more fundamental than even its awareness, because – insofar as a lamp illuminates itself while illuminating other things – illumination is the very condition of reflexivity that makes awareness of anything possible (Instructions,145b3). In other words, Ratnākaraśānti presents illumination as the substance (dravya) of consciousness 590

Ratnākaraśānti

and awareness as its natural function (Instructions,143a5, 145b3, 150a2). By grounding mind’s knowing quality within its illuminating quality, Ratnākaraśānti suggests that the lamp-like reflexivity of illumination underlies both an ordinary person’s intentional consciousness and a buddha’s simultaneously transcendent and immanent awareness. In this way, he sets up his explanations both of how illumination remains unchanged when an ordinary person becomes awakened and of how buddhas can still communicate with ordinary beings. To explain his ontology, Ratnākaraśānti establishes what it means that all phenomena are only consciousness in four main steps. In the first step, Ratnākaraśānti argues that phenomena are ontologically grounded in awareness due to an identity (tādātmya) relation between what we might call illumination qua phenomena and illumination qua awareness. Second, he argues that if awareness grounds phenomena and if awareness is aware of phenomena, then awareness is ultimately reflexively aware of itself. Third, he argues that the identity relation between real illumination and unreal representational forms is not contradictory because the former is the “bare” nature, while the latter is something “superimposed” upon that. Finally, he explains how illumination, as an ontological substance, manifests, due to cognitive error, either in the three types of intentional consciousness of ordinary persons or in the three bodies of a buddha. Each of these five points will be expanded upon in the four following subsections.

Phenomena Are Ontologically Grounded in Awareness Ratnākaraśānti’s basic ontological argument is that illumination is the only real substance, but it can take on unreal, adventitious representational forms due to cognitive error. To begin with, he proves that all phenomena have nothing but consciousness as their inherent nature, and since this means they lack the nature of a grasping subject or grasped object, they themselves do not have their own inherent nature (Instructions, 145a6). The upshot of his argument is that phenomena may appear to be different from awareness, but through direct perception ( pratyakṣa) and inference (anumāna), phenomena can be proven to be simply representational forms made out of the substance of illumination whose nature is clear and indiscernible from awareness. For Ratnākaraśānti, the fact that phenomena are just illumination can be established first through direct perception because we directly experience illumination in the phenomena shining forth in our minds. That is to say, just like a film spectator is aware of the light within film images, an ordinary person is always subconsciously aware that phenomena are nothing but illumination in various representational forms because they are shining forth, manifesting, or appearing in awareness and are neither “dim” (  jaḍa) nor “inaccessible” ( parokṣa) to it (Instructions, 145a6). For this reason, when an ordinary person says “I see a blue form,” it is identical to saying “I see the illumination of a blue form.” Here, Ratnākaraśānti explains the blueness of the form as appearing due to mind’s latent impressions; he explains illumination’s nature as being clear – that is, lacking its own color, shape, form, and so on – because, if it were opaque, nothing would be able to appear in it and no other phenomena could ever be established through it. For Ratnākaraśānti, on the one hand, this clear illumination cannot be known unless it is the substance of awareness itself. On the other hand, an intentional object is by definition accessible and connected to awareness only through illumination that is clear, like a nonmaterial mirror that illuminates objects. That is to say, any intentional object that appears in that mirrorlike clear nature, such as a blue color patch and so on, must share a connection (sambandha) of identity (tādātmya) with that illumination (Instructions, 153a4).7 So, if one were to say, “I see the illumination of a blue thing,” this would be the same as saying “Illumination qua awareness is taking on the representational form of grasping the illumination in the blue representational form grasped.” In other words, the representational form of 591

Gregory Max Seton

a grasped blue object and the representational form of a grasping subject are both made out of illumination. In this way, Ratnākaraśānti argues, on the basis of direct perception and infe­ rence, that all phenomena have awareness as their nature because illumination is established as the substance out of which both phenomena and awareness are made.

Awareness Is Ultimately Reflexively Aware of Itself Ratnākaraśānti next defends the implication that if awareness is the ontological ground of phenomena and if awareness is aware of phenomena, then awareness must be ultimately only reflexively aware of itself. To this end, he puts forth a two-part proof that makes use of the ambiguous and multivalent Sanskrit instrumental case to insinuate simultaneous concomitance, causality, and identity. First, Ratnākaraśānti argues that, if an intentional object cannot be known except through (vinā) illumination, then that intentional object is known as or by that illumination (Instructions, D153a4). His point here seems to be that, for instance, a blind dog who knows a woman only through her scent knows her as or by her scent. Hence, there is no contradiction in saying that one can be reflexively aware of intentional objects as illumination, because it is the intentional object’s illumination alone that is known through illumination qua awareness. Second, he argues that, if an intentional object is something to be known by or as illumination, then that illumination must be awareness (Instructions, D153a5). Since only awareness can be known through reflexive awareness (svasaṃvedya), the awareness of illumination qua phenomena cannot be anything other than an identity connection with it. Thus, he concludes that the nature of pure consciousness is proven to be the reflexive awareness of illumination, and hence, the false and unreal representational forms of objects grasped by consciousness are also refuted as being external to the mind (Instructions, D153a6).

The “Bare” and “Superimposed” Natures Are Not Contradictory Next, Ratnākaraśānti defends his argument against a possible objection – that is, that an identity connection between a real thing and unreal thing is self-contradicting – by distinguishing the representational forms as a “superimposed” (āropita) false nature and their illumination qua awareness as their “bare” (anāropita, lit. “unsuperimposed”) true nature. According to Ratnākaraśānti, it is meaningful to speak of an identity between the unreality of the representational forms and the reality of their illumination, because “unreal” and “real” – insofar as they are based on (niṣṭha) different natures that are, respectively, superimposed and bare – can be considered contradictory types of properties that refer to a single locus (vastu) without negating one another (Instructions, 149a3 – a6; Textual Commentary ad 3623–24). To explain this special type of identity connection, we may compare it to Gibbard’s (1975, 187) notion of socalled “contingent identity” between an indestructible form-free lump of clay and a destroyable clay statue made from it.8 Gibbard’s idea seems to be that even though a form-free lump of clay and the statues or pots fashioned out of it can be considered identical, their identity is contingent, in the sense that the statues or pots can be destroyed by the hands, but the lump of clay – which is “form-free” in the sense of being substance without a particular form – is not destroyed by that action. Likewise, Ratnākaraśānti’s notion of sheer illumination being “bare” is comparable to the indestructible lump of clay, and his notion of representational forms being “superimposed” is comparable to the destroyable clay statues or pots fashioned out of it. In this way, Ratnākaraśānti asserts a special type of identity relationship between reflexively aware illumination and its false representational forms. By holding this real substance to be a momentary material cause in line with mainstream Buddhist doctrine, Ratnākaraśānti shifts 592

Ratnākaraśānti

the focus away from the function of awareness toward the substance of illumination in order to avoid the prior criticisms of Form-free scholars who refute any possible causal relationship between a real awareness and any unreal intentional objects. Furthermore, Ratnākaraśānti’s argument sets up his subsequent assertion that  – by retaining a small amount of cognitive error – a buddha’s nondual illumination is an immaterial substance that has the moment-bymoment power (śakti) to be aware reflexively either of its own transcendent inherent nature alone, for example, when it is like a lump of clay, or of its immanent inherent nature in the false representational forms themselves, for example, when it is like a clay statue.

Illumination’s Reflexivity as Intentional Consciousness Next Ratnākaraśānti connects the perception of buddhas to ordinary persons epistemically. To do so, he explains illumination’s inherent reflexivity as having two nondual aspects – namely a cognizing (  jñāna) and cognized aspect (  jñeya) – that appear, due to a fundamental cognitive error (bhrānti), either as the quasi-intentional immanent awareness of a buddha or as the intentional consciousness of an ordinary person. In the case of a buddha, the quasi-intentional immanent awareness arises due to cognitive error as a cognizing aspect that recognizes the inherent nature (svarūpa) of its cognized aspect as being identical to its own inherent nature (Excellent, 15; Seton 2016, 96). But in the case of an ordinary person, since intentional consciousness has conceptual proliferations in addition to the fundamental cognitive error, it does not recognize the identical nature of these two nondual aspects and instead superimposes upon them the imagined characteristics of a grasped object and grasping subject, even though these are unreal (asat). That is to say, in the ordinary person’s imagination, the cognized and cognizing aspects appear to be discrete representational forms of a grasped object and grasping subject despite their not being real discrete substances (avastu) (Excellent, 15; Seton 2016, 93–95). In this way Ratnākaraśānti demonstrates that an ordinary person’s consciousness is not perceiving the bare inherent nature of the cognized or cognizing aspects of his or her reflexively aware illumination, simply because he or she is fixated upon the unreal content in the non-discrete representational forms superimposed upon the cognized and cognizing aspects due to their conceptual proliferations. That said, to compare how illumination manifests in an ordinary person’s intentional consciousness and in the buddha’s quasi-intentional immanent awareness subsequently, it will be helpful first to describe the way that illumination is the substance of the three streams of an ordinary person’s intentional consciousness – that is, the “container consciousness” (ālayavijñāna), the “afflicted ego mind” (kliṣṭamanas), and the “actively aware consciousness” ( pravṛttivijñāna) – which arise with their own false and unreal representational forms of a grasper and grasped moment to moment. To begin with, illumination is the substance of the unselfconscious “container consciousness,” which is like the sleeping mind’s consciousness in which and out of which a dream world appears (Instructions, 139a2). This stream of consciousness is called a “container” because its primary function is to contain the latent impressions of past actions (karma) that shape our character, dispositions, and tendencies and to project them onto our mindscreen, like the reels holding the film that is projected frame by frame in the mental movie of our dream.9 Within this nonconscious awareness, illumination’s natural reflexivity – with its cognizing and cognized aspects – manifest as the animate beings and inanimate world within a dream (Instructions, 139a4–5; Five Aggregates, 91–92). This container consciousness is intentional because it operates passively in the background of all experience – even when the other two streams of consciousness cease temporarily during deep sleep, fainting, meditative cessation, and so on – as a subconscious awareness that registers moment by moment what occurs in the world, mind, and 593

Gregory Max Seton

body (Instructions, 141a1; Five Aggregates, 93).10 It also serves as the ground for the continuity of consciousness when mind alternates between its own unconsciousness and the conscious mental states – the actively aware consciousness and afflicted ego mind – which arise out of it. Second, illumination is the substance of the “afflicted ego mind,” a consciousness that arises from the container consciousness like a wave arising from the ocean. The ego-mind manifests as a self that is “afflicted” because it is painfully engrossed in self-delusion (ātmamoha), self-imaging (ātmadṛṣṭi), self-attachment (ātmasneha), and self-conceit (ātmamāna). The ego-mind is the subconscious sense of a subject grasping the container consciousness as an unseen internal personal identity with particular characteristics gleaned from the various representational forms (Five Aggregates, 108; Instructions, 139a1), like emotionally identifying with the experiencer of the dream images as a “me” with certain characteristics, such as body size, personal history, and so on. In this way, the ego-mind influences the actively aware consciousness to become engrossed in particular aspects of the dream as though they were real and related to “me” (Instructions, 139a5). Third, illumination is the substance of six types of “actively aware consciousness” ( pravṛttivijñāna), namely the mental consciousness and the five types of sense consciousness. Like further waves, the mental consciousness arises from the container consciousness moment by moment, and the five types of sense consciousness arise in sequence or simultaneously based on conditions (Instructions, 141a1–3). Each of the five sense perceptions is a grasping of different grasped intentional objects produced by latent impressions emerging from the container consciousness, like perceiving various images in a dream (Five Aggregates, 92; Instructions, 138b7). Since the mental consciousness is purely conceptual, it is a conceptual grasping of its intentional objects (artha) as possessing the characteristics of cowness or blueness common to all particular manifestations (vyakti) of cows or blue color patches. Hence, its grasped perceptual content “cow” or “blue” appears in representational forms that are false or nonveridical due to being connected with universalizing linguistic structures that cognitively exclude non-blue and non-cow.11 In modern neuroscientific terms, we might say that our mental images of “cow” or “blue” are purely our human brain’s predictive interpretations of light and that they do not correspond to any blue color or cow that exists in external reality. Thus, the illumination that appears as the representational forms is real, but their presentation ( prakhyānam) as a “cow” or as “blue” is merely a mentally constructed determination (adhyavasāya) of unreal intentional objects (Instructions, 149a3-a6), like determinations of a person’s character in a dream. Now, in order to realize the bare nature of illumination that forms the substance of the superimposed streams of consciousness, reflexive awareness must be established by valid cognition in a nondual way – that is, free of the subject-object structuring of ordinary intentional consciousness – through the following four stages of phenomenological and ontological selfreflection practice ( yogabhūmi).12 In the first stage, one must reduce all appearances to a fixed set of categories consisting of what is grasping and grasped. In the second stage, one must examine the grasping subject and the grasped object as a purely mental experience. In the third stage, one must identify illumination’s emptiness of forms as the single common characteristic (sanimittā) of these forms and then reflect upon that emptiness until one discovers the transcendent awareness that is free of subject-object structuring. Although this transcendent awareness is free of intentional content (nirālambana), it is not free of all content, insofar as it is reflexively aware of its own suchness being empty of forms (Instructions, 149b7).13 This discovery at the end of the third stage marks the point where an ordinary person first enters the so-called path of seeing (darśanamārga). Here, if she did not previously do accompanying

594

Ratnākaraśānti

Mahāyāna practices to cultivate unconditional compassion, she would become a first-phase “arhat” – that is, someone who permanently remains in a discrete body of sheer illumination never to be reborn in further physical bodies. As such, she remains unable to benefit others in the way that a buddha can. If she did previously do accompanying Mahāyāna practices, she would at this time become a first level (bhūmi) “noble bodhisattva.” In the fourth stage, one must practice resting in reflexive awareness of illumination’s emptiness qua freedom from any identifiable characteristics (nirnimittā) (Instructions, 149b7) on the so-called path of cultivation, where a noble bodhisattva alternates between her transcendent awareness  – which involves resting in nondual reflexive awareness without representational forms during meditative states  – and its after-effect, a pure immanent awareness (śuddhalaukikajñāna) during post-meditative states in order to perfect her ability to rest in the nondual reflexive awareness even while interacting with the world (Instructions, 159a4). According to Ratnākaraśānti, the post-meditative awareness itself is called “immanent” because it has strayed (’khrul ba nyid kyis; bhrāntatvena) from sheer illumination into the world of illusory appearances, and yet “pure” because it is able to discern the suchness (tathatā) of these illusions (Guide, 226a7). Finally, at the end of the fourth stage, one wakes up as a buddha with a nondual reflexive awareness in which the distinction between the meditative “transcendent” awareness and the post-meditative “immanent” awareness disappears. According to Ratnākaraśānti, once a buddha awakens, the very basis of the three streams of intentional consciousness is permanently transformed (āśrayaparāvṛtti), and the bare cognizing aspect of reflexive awareness simply recognizes the bare cognized aspect of its own illumination that was always present in the superimposed representational forms of a grasping subject and grasped subject. In the Form-free Yogācāra notion of this awakening, a buddha’s reflexive awareness is called nondual because its cognizing and cognized aspects never again take on representational forms. That is to say, the transformed consciousness will never again be intentional, because in a miraculous way it penetrates all phenomena (sarvadharmavibhutvalābhatas) of ordinary persons without a subject-object structure (Thirty Verses Commentary ad 30). However, in Ratnākaraśānti’s False-form position, such a nondual reflexive awareness free of representational forms can only result in the state of an arhat ­(Sky-like, 231). For Ratnākaraśānti, what distinguishes a buddha’s nondual reflexive awareness from that of an arhat is its ability to retain cognitive error – which provides a mechanism for interacting with the world of ordinary persons through representational forms – without being subject to its negative effects, like a lucid dreamer who is not at all subject to the dream. In order to distinguish the buddha’s nondual reflexive awareness from a bodhisattva’s pure immanent awareness, Ratnākaraśānti qualifies the amount of error retained as being “small” to suggest that it produces representational forms so translucent that their empty nature is effortlessly obvious to a buddha at all times – unlike to bodhisattvas who, in effect, must return to meditative states in order to recharge their awareness of illumination. Hence, a buddha’s nondual awareness is completely “nondual” in two senses. First, her reflexive awareness is free of the ignorant subject-object structuring of an ordinary person’s intentional consciousness. Second, she is able to simultaneously and completely recognize her own inherent nature in those representational forms, as if her dream images were so translucent she could not avoid seeing their illumination. Thus, in Ratnākaraśānti’s False-form position, a buddha’s nondual reflexive awareness is not just an ordinary person’s inferential knowledge of the reflexivity of awareness, not just an arhat’s direct perception of sheer illumination free of forms after physical death, not just a bodhisattva’s transcendent awareness of sheer illumination during meditation with its aftereffect of immanent awareness during post-meditation, but

595

Gregory Max Seton

rather a transcendent and immanent, quasi-intentional consciousness of representational forms that never loses sight of sheer illumination being the only real thing.

Ratnākaraśānti’s Implied Theory of Error Ratnākaraśānti speaks often of error without ever providing a systematic theory of error. Hence, this section will have to tease out his unique perspective from his various assertions about the three bodies of a buddha and how they function in the world. But before delving into these assertions, we must first address the fact that we have been translating the term bhrānti as “cognitive error,” but the term has a wider range of meaning in Ratnākaraśānti’s writings. First, at the level of ordinary persons, Ratnākaraśānti uses the term bhrānti to refer not only to a “cognitive error” – which causes ordinary persons to perceive a dualistic illusion of representatational forms whose nature is real illumination – but also to a “conceptual error,” which is the conceptual proliferations about a false and unreal grasping subject and grasped object.14 Second, unlike any other prior Buddhist scholar, Ratnākaraśānti claims that a buddha, out of compassion, retains a small amount of bhrānti so that she can engage with representational forms and still accurately know the content of those illusory forms to be false and unreal. By specifying only a “small amount,” Ratnākaraśānti implies that a buddha retains only the most basic cognitive error that gives rise to representational forms but not the conceptual error tied to ignorance that leads to suffering. Thus, beyond the “cognitive” versus “conceptual” distinction, Ratnākaraśānti’s notion of bhrānti might not best be translated as “error” because the buddha is not deceived by retaining it, and her own reflexive awareness remains nonerroneous and “free” of representational forms. That is to say, in Ratnākaraśānti’s system, the word bhrānti – a noun formed from the Sanskrit verbal root √bhram, whose range of meanings includes “to wander,” “to err,” or “to spread” – might be better translated as a more neutral “spreading” or “diffusion” of illumination which can be erroneous or not. Whatever English translation is generally adopted, though, it should be emphasized that Ratnākaraśānti’s idea of a buddha’s bhrānti is that it enables her transcendent awareness to appear in an intersubjective world of representational forms as an accurate, non-erroneous, immanent awareness – rather than causing her somehow to be lost in ignorance and fallible. In this way, unlike the Form-free notion of bhrānti as an all-or-nothing, purely accidental error that is eliminated upon awakening as a buddha, Ratnākaraśānti treats bhrānti more like a neutral efficient cause that makes the substance of sheer illumination – which exists moment to moment in a stream ( pravṛtti) – appear as representational forms, like a crystal through which a ray of clear illumination is diffused in order to produce visible colored forms that others can perceive. That said, in Ratnākaraśānti’s system, when sheer illumination is diffused by bhrānti, it is not just perceived by others, but rather it illuminates both itself and others in a way that it can consciously interact with other beings in an intersubjective world. In Ratnākāraśānti’s nonsolipsistic universe, distinct ordinary persons are lost in their own bubbles of self-illuminating awareness distorted by their cognitive and conceptual error into various representational forms of grasping subjects and grasped objects. But persons in different groups share common illusions that they grasp differently due to their different conceptual proliferations, like different audience members in the same theater grasping different aspects of the same unreal movie images on screen. So, the difference between a buddha’s and an ordinary person’s representational forms is that a buddha has no conceptual proliferations and recognizes the cognized and cognizing aspects of diffused illumination to be her own projections, whereas an ordinary person has conceptual proliferations and misperceives the cognized and cognizing aspects as an imaginary grasped and grasper. 596

Ratnākaraśānti

As for how a buddha can function in a false and unreal world in order to awaken other beings, Ratnākaraśānti explains the three bodies of a buddha as being both real and illusory forms of illumination free of conceptual proliferations. To begin with, a buddha’s natural body consists in real illumination reflexively aware of its own emptiness. However, as a natural outcome (niṣyandaphala) of the natural body being the pure source of qualities (dharmadhātu), it automatically diffuses itself through the prism of bhrānti to manifest a light body of enjoyment, which is its own mental construct (vikalpa). This diffusion of illumination is reflexively aware of itself but also able to be perceived directly in various forms by beings who are almost awake (bodhisattva) and indirectly, as an emanation body, by ordinary persons, who are still asleep. Hence, even though Ratnākaraśānti does not explain what representational forms are like for a buddha, his notion of bhrānti enabling her both to see the illusory world and to be seen in it might explain why Ratnākaraśānti, unlike Sthiramati, holds both the grasper and grasped to be representational forms consisting of illumination. Whatever the case, though, he makes clear that an awakened person’s natural body does not just embody its own freedom (vimuktikakāya) – like that of an arhat  – but embodies and shines forth the qualities (dharmakāya) that can participate in different illusory dimensions, like an awakened person entering into a lucid dream in order to project herself into others’ dream worlds. In this regard, Ratnākaraśānti also makes clear that the three bodies ultimately consist in the same real illumination but he treats the transcendent aspect of a buddha as the natural body – that is, a sunlike orb of reflexively aware illumination – and her immanent aspect as the sunray-like enjoyment body shining forth illusory representational forms to communicate with beings. High-level bodhisattvas, close to awakening themselves, directly perceive and enjoy the real illumination of that illusory enjoyment body in translucent representational forms, like semi-lucid dreamers with sunlight on their faces noticing the brightening of all their dream images.15 But those ordinary persons, who are light sleepers, perceive that same illumination indirectly as opaque representational forms of physical and verbal content, like ordinary dreamers imagining the sunlight on their faces to be bright dream lightbulbs shining in their dream eyes (Guide, 228b3). In both cases, it is bhrānti that allows a buddha to shine forth her own illusory bodies with a cognized and cognizing aspect as representational forms that appear to others like a grasping subject aware of what is grasped. In this way, a buddha’s bhrānti is instrumental in allowing the substance of a buddha’s reflexively aware illumination to take on the form of a real enjoyment body at the conventional level of bodhisattvas and to appear as various emanation bodies in the conventional dream world of ordinary persons, without a buddha actually entering the dreams herself. Therefore, in Ratnākaraśānti’s implied theory of error, bhrānti is more or less a diffusion that causes the cognizing and cognized aspects of sheer illumination to appear distinct, like visual floaters seeming to be objects outside an eye, whereas the conceptual proliferations create erroneous content. By retaining only a small amount of bhrānti, a buddha’s illumination is diffused just enough to produce representational forms but not enough to produce erroneous conceptual proliferations that would cloud the cognizing and cognized aspects of a buddha’s reflexive awareness. In this way, bhrānti is central to explaining the transcendent and immanent aspects of a buddha’s natural body that has the power (śakti) to communicate accurately with ordinary beings without being subject to the same ignorance and suffering that they experience.

Conclusion For Ratnākaraśānti, the Form-bound ontology cannot explain the transcendent qualities of a buddha’s natural body; whereas the Form-free epistemology cannot explain the immanence of 597

Gregory Max Seton

a buddha’s bodies of enjoyment and emanation. For this reason, he argues for a position that can be seen as agreeing with Form-free ontology but making some limited concessions to Formbound epistemology with regard to a buddha’s interaction with ordinary persons. To counter those who follow Kamalaśīla’s idea of a true dream but refute illumination, Ratnākaraśānti asserts that a buddha must realize illumination – that is, the substance of intentional consciousness – through reflexively aware valid cognition but, out of compassion for others, retain a small amount of bhrānti in order to produce physical bodies able to communicate with ordinary persons who perceive only false and unreal forms, and to teach them the path to realize the nature of their consciousness as illumination.16 By arguing that seemingly contradictory properties – such as real and illusory, superimposed and bare, and so on – can in fact coincide within a single locus, he establishes a type of contingent identity relation between illumination and forms. Based on this type of identity, we are meant to understand why these physical bodies of buddhas can have both real and illusory properties and how their nondual reflexive awareness can be intentionally conscious of representational forms while reflexively aware of their illumination. In effect, Ratnākaraśānti’s ontological assertion of sheer illumination ultimately free of representational forms (nirākāra) does not indicate that a buddha does not have any forms, but simply that a buddha’s awareness has the ability to be “free” of forms. Furthermore, a buddha engages in cognitive diffusion that is free of the conceptual error represented by conceptual proliferations. Hence, although Ratnākaraśānti classifies his own viewpoint simply as “Form-free” and although many traditional and modern scholars conflate the names False-form (rnam brdzun) and Form-Free (rnam med) and use them interchangeably, we can call his position “False-form insofar as it asserts the illumination of false forms.” Thus, in this way, Ratnākaraśānti’s assertions truly distinguish his position from all other scholars within the history of Mahāyāna Buddhist philosophy.

Notes 1 As described in the Guide’s colophon, ostensibly written by Śāntibhadra, likely a student of his. 2 See McClintock’s (2013, 4) discussion of Kamalaśīla’s agreement with the “third” position in Textual Commentary 3626–27. 3 Here, the term “bound” does not signify this position’s self-description or any particular terminology but rather a metaphorical interpretation of the word sa- (“with”) in sākāra (viz. “with representational form”). Different sākāra positions explain consciousness as bound in different ways, such as indivisibly (citrādvaita), and so on. See Textual Commentary, 3627, where a buddha’s momentary awareness is described in several different ways, for example, as arising in conjunction (upagraheṇotpatti) with representational forms. 4 Here and subsequently, where my conclusions are based upon my own reading of Ratnākaraśānti’s works extant in Sanskrit or Tibetan, which have not been published in English translation, I will cite an abbreviation of a translated English title, but the subsequent folio numbers refer to the Sanskrit or Tibetan editions, while “a” or “b” refers to front or back sides of the folios. 5 I have interpreted the phrases sattāmātreṇa and sattām anubhavati differently from Funayama 2007, 194 n. 35. 6 Kajiyama (1989) explicitly conflates the two names nirākāra (“Form-free”; Tib. rnam med) and *alīkākāra (“False-form”; rnam brdzun).” For more on classifying the position, see Seton 2016, chapters 2–4 and Luo 2018. 7 It is widely accepted among Buddhist scholars (e.g., Textual Commentary ad 3459) that there are only two types of connection between things, either an identity connection (tādātmya), meaning they share the same nature, or a causal connection (tadutpatti), meaning one arises from the other. Other epistemological idealists hold that awareness and its intentional object share a causal connection. 8 Tomlinson (2018, 4) discusses the Gibbard’s clay example in a slightly different way. 9 In modern terms, one might say that the unconscious processes during resting and digesting all rely on the identification of some sort of intentional objects, but active awareness is just unaware of them.

598

Ratnākaraśānti 10 Here, meditative cessation is a state into which meditators enter into with an intention to return to consciousness after a given period of time; that is, the basic continuity of consciousness is not lost during unconsciousness. 11 There are many Buddhist theories of exclusion. The basic idea is that the predictive mechanism of the mind pragmatically develops the abstract universal “cow” together with learned linguistic conventions – that is, the word “cow” – and it does so not by establishing all the common characteristics of “cow-ness” but by excluding the main characteristics of a non-cow. Based on this abstract universal, a conceptual cognition can then determine those unique characteristics (which are not non-cow characteristics) as belonging to particular instances of cows. 12 For a brief comparison of phenomenology in Buddhist and non-Buddhist contexts, see Tomlinson 2019, 20–26. 13 See, for example, Pearl Garland 97. 14 Ratnākaraśānti’s virtual silence on conventional “perceptual errors” – like the optical illusions of a rotating firebrand or cataracts disease – shows his general disinterest in epistemological explanations of how to act accurately in the world. 15 For Ratnākaraśānti, mental constructs exist qua substance. Cf. Guide, 228b3 or Pearl-Garland, i.37–38. 16 For more on the true dream, see McClintock 2014, 10.

Bibliography Chan, Ngan Che. 2007. “A  Study of Yogācāra Theory of the Ten Causes.” PhD diss., University of Hong Kong. Chimpa, Lama, and Alaka Chattopadhyaya. 1990. Taranatha’s History of Buddhism in India. Edited by Debiprasad Chattopadhyaya. Reprint edition. New Delhi: Motilal Banarsidass. Dowman, Keith. 1986. Masters of Mahamudra: Songs and Histories of the Eighty-Four Buddhist Siddhas. Albany: State University of New York Press. Excellent Commentary = Rinpoche, Samdhong and Vrajavallabh Dwivedi, eds. 1992. Mahāmāyātantra with Guṇavatī by Ratnākaraśānti. Rare Buddhist Text Series 10. Project edition. Sarnath, Varanasi: Central Institute of Higher Tibetan Studies. Five Aggregates Explanation  =  Kramer, Jowita, ed. 2013. Sthiramati’s Pañcaskandhakavibhāṣā:­ Critical Edition (Part 1). Vienna: China Tibetology Publishing House, Austrian Academy of Sciences Press. Funayama, Toru. 2007. “Kamalaśīla’s Distinction Between the Two Sub-Schools of Yogācāra. A Provisional Survey.” In Pramāṇakīrtih. Papers Dedicated to Ernst Steinkellner on the Occasion of His 70th Birthday, Part 1, edited by Birgit Kellner, Horst Lasic, M. T. Much, and Helmut Tauscher, 187–202. Wien: Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. Gibbard, Allan. 1975. “Contingent Identity.” Journal of Philosophic Logic 4 (2): 187–221. Guide to Adornment of the Middle Way = Rin-chen-’byung-gnas-zhi-ba (Ratnākaraśānti). dBu ma rgyan gyi man ngag (*Madhyamālaṃkāropadeśa, MAu). Translated by Śāntibhadra & Śākya-’od. Revised by Amogha & ‘O ru. Tōh. 4085. Instructions on Transcendent Wisdom = Ratnākaraśānti. Shes rab kyi pha rol tu phyin pa’i man ngag (Prajñāpāramitopadeśa, PPu). Translated by Zhi-ba-bzang po (Śāntibhadra) and’Gos [Khug pa]-lhasbtsas. Tōh. 4079. Innate Sādhana  =  Isaacson, Harunaga. 2001. “Ratnākaraśānti’s Hevajrasahajasadyoga (Studies in Ratnākaraśānti’s Tantric Works I).” In Le Parole E I Marmi: Studi in Onore di Raniero Gnoli Nel Suo 70 Compleanno, edited by Raffaele Torella, 457–87. Roma: Istituto per L’Africa e L’Oriente. Isaacson, Harunaga, and Francesco Sferra, eds. 2014. The Sekanirdeśa of Maitreyanātha (Advayavajra) with the Sekanirde-śapañjikā of Rāmapāla: Critical Edition of the Sanskrit and Tibetan Texts with English Translation and Reproductions of the MSS. Manuscripta Buddhica 2. Napoli: Università degli Studi di Napoli “L’Orientale.” Kajiyama, Yūichi. 1989. Studies in Buddhist Philosophy: Selected Papers. Edited by Katsumi Mimaki et al. Kyoto: Rinsen Book Co., LTD. [Reprint of Kajiyama, Yūichi. 1965. “Controversy between the sākāra- and nirākāra-vādins of the yogācāra school – some materials.” Journal of Indian and Buddhist Studies 14 (1): 418–29]. Kano, Kazuo. 2016. Buddha-Nature and Emptiness: rNgog Blo-ldan-shes-rab and A Transmission of the Ratnagotravibhāga from India to Tibet. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien, Universität Wien.

599

Gregory Max Seton Luo, Hong. 2018. “Is Ratnākaraśānti a gZhan stong pa?” Journal of Indian Philosophy 46: 577–619. Mahāyāna Scriptural Adornment = Bagchi, S., ed. 1970. Mahāyānasūtrālaṅkāra of Asaṅga. Buddhist Sanskrit Texts, no. 13. Darbhanga: Mithila Institute. Funahashi, Nagoya, ed. 1985. Mahāyānasūtrālaṃkāra Chapter I, II, II, IX, X. Tokyo: Kokushokankokai. Makransky, John J. 1998. Buddhahood Embodied: Sources of Controversy in India and Tibet. Albany: State University of New York Press. McClintock, Sara. 2014. “Kamalaśīla on the Nature of Phenomenal Content (ākāra) in Cognition: A Close Reading of TSP ad TS 3626 and Related Passages.” Journal of Indian Philosophy 42 (2/3): 327–37. Middle-Beyond-Extremes Commentary = Bhattacharya, Vidhushekhara and Giuseppe Tucci, eds. 1932. Madhyāntavibhāgasūtrabhāṣyaṭīkā of Sthiramati: Being a sub-commentary on Vasubandhu’s Bhāṣya on the Madhyāntavibhāgasūtra of Maitreyanātha (Part I). London: Luzac & Co. Nondual Variegated Illumination  =  Thakur, Anantalal, ed. 1957. “Citrādvaitaprakāśa.” In Ratnakīrtinibandhāvalī. Patna: Kashiprasad Jayaswal Research Institute. Pearl-garland Commentary = Tripāṭhī, Rāmaśaṅkara and Ṭhākurasena Negī, ed. 2001. Hevajratantram: Ratnākaraśāntiviracita-Hevajrapañjikā-muktāvalī-saṃvalitam. Vārāṇasī: Kendrīya Uccạ Tibbatī Śikṣāsaṃsthānam. Cf. Muktāvalī, I.viii (unpublished), ed. by Harunaga Isaacson. Proving Mere Representation = Ratnākaraśānti. Vijñaptimātratāsiddhi. rNam par rig pa tsam nyid du grub pa zhes bya ba. Tōh. 4259, D, tshad ma, zhe 306b4–309b3. Roerich, George de, and Gendün Chöphel. 1988 (1949). The Blue Annals by Gö Lotsawa. Translated by George de Roerich and Gendün Chöphel. New Delhi: Motilal Banarsidass. Reprint of Calcutta, Royal Asiatic Society of Bengal, in two volumes. Quintessential Commentary = Ācārya Ratnākaraśānti. 1979. Sāratamā: A Pañjikā on the Aṣṭasāhasrikā Prajñāpāramitā Sūtra. Edited by Padmanabh S. Jaini. Calcutta: Kashi Prasad Jayaswal Research Institute. Schmithausen, Lambert. 2005. On the Problem of the External World in the Ch’eng wei shih lun. Studia Philolgica Buddhica (Occasional Paper Series XII). Tokyo: International Institute for Buddhist Studies of the ICPBS. Seton, Gregory Max. 2016. “Defining Wisdom: Ratnākaraśānti’s Sāratamā.” DPhil diss., University of Oxford. ———. 2018. “Ratnākaraśānti.” In Brill’s Encyclopedia of Buddhism, Volume 2: Lives, edited by ­Jonathan A. Silk. Leiden: Brill. ———. 2019. “Ratnākaraśānti.” The Treasury of Lives. https://treasuryoflives.org/biographies/view/ Ratnākaraśānti/P00EGS1016642. Sky-like Commentary  =  Upādhyāya, Jagannāth, ed. 1983. “Khasama-nāma-ṭīkā.” In Śramaṇavidyā, edited by Gokul Chandra Jain. Varanasi: Sampurnanand Sasnkrit Vishvavidyalaya. Ten levels Scripture = Rahder, J., ed. 1926. Daśabhūmikasūtra et Bodhisattvabhūmi: Chapitres Vihāra et Bhūmi. Paris: Paul Geuthner Louvain, J. B. Istas. Textual Commentary on the Compendium of Reality = Krishnamacharya, Embar, ed. 1926. Tattvasaṃgraha of Śāntarakṣita with the Commentary of Kamalaśīla. Gaekwad’s Oriental Series, no. 30–31. Baroda: Central Library. Thirty Verses Commentary = Buescher, Harmut, ed. 2007. Sthiramati’s Triṃśikāvijñaptibhāṣya. Critical Editions of the Sanskrit Text and its Tibetan Translation. Vienna: Verlag der Östereichischen Akademie der Wissenschaften. Thurman, Robert. 2004. The Universal Vehicle Discourse Literature: Mahāyānasūtrālaṃkāra by Maitreyanātha/Āryāsaṅga together with its Commentary (Bhāṣya) by Vasubandhu. Translated by L. Jamspal, R. Clark, J. Wilson, L. Zwilling, M. Sweet, and R. Thurman. New York: American Institute of Buddhist Studies at Columbia University. Tomlinson, Davey. 2018. “Ratnākaraśānti’s Metaphysics of Difference: Limiting the Scope of the NeitherOne-Nor-Many Argument.” Delivered in the Buddhist Studies Unit at the Conference of the American Academy Religion, Denver, November 17–20, 2018 (unpublished). ———. 2019. “Buddhahood and Philosophy of Mind: Ratnākaraśānti, Jñānaśrīmitra, and the Debate over Mental Content (ākāra).” PhD diss., University of Chicago. Twenty Verses = Silk, Jonathan A., ed. and trans. 2018. Materials Toward the Study of Vasubandhu’s Viṃśikā (I): Sanskrit and Tibetan Critical Editions of the Verses and Autocommentary; An English Translation and Annotations. Harvard Oriental Series, no. 81. Cambridge: Department of South Asian Studies, Harvard University.

600

37 ATIŚA The Great Middle Way of Mere Appearance James B. Apple

Introduction Atiśa, also known as Dīpaṃkaraśrījñāna (982–1054), is famous as a master from the ancient Indian Buddhist land of Bengal and for his journeys in Indonesia and Nepal. In the last thirteen years of his life, he became one of the most influential Indian Buddhist masters ever to set foot in Tibet. His Mahāyāna Buddhist teachings, encompassing instructions on entering the way of bodhisattvas up through highly advanced practices of the esoteric secret way of mantras, came to influence all subsequent traditions of Buddhism in Tibet. Atiśa followed a lineage of Indian Buddhist Madhyamaka (Middle Way) that was based on the teachings of Nāgārjuna and Candrakīrti and influenced by his direct teachers, including Bodhibhadra (fl. c. 1000), a scholar-monk at Nālandā; the contemplative-monk Vidyākokila (fl. c. 1000); and the tantric yogi Avadhūtipa (fl. c. 1000). Atiśa’s Madhyamaka lineage was upheld and actively taught in Tibet among communities of his Kadampa (bka’ gdams pa) followers until it was superseded by forms of Madhyamaka infused with epistemology that developed at the famous Tibetan monastery of Sangphu Neuthok (gsang phu ne’u thog) in the twelfth century (Apple 2019a). Atiśa’s lineage of the Middle Way was contemplative in nature and did not utilize epistemological warrants ( pramāṇa) to realize ultimate reality. Still, although not advocating the employment of epistemological warrants for Buddhist meditative realization, Atiśa did utilize reasoning ( yukti) for the direct cognition of insight. He grounded his use of reasoning from scriptural authoritative texts attributed to the Buddha that demonstrate a continuity in the use of reasoning within Buddhist tradition. He denied a substance-based ontology while preserving a Buddhist soteriological program that accepts causality in the efficacy of conventional practices of moral discipline and meditation. Atiśa offers the modern reader a rare glimpse into an integrated Indian Buddhist Middle Way philosophy composed of a nominalism of “mere appearances” (snang ba tsam) in which both objects and cognitions are dependently designated and are therefore mere imputations ( prajñāptimatra). Atiśa advocates a mentalist theory of Madhyamaka in which the mind, as mere appearance, is not at all established and is a mere nominal designation free from the extremes of existence and nonexistence. In its overall orientation, Atiśa’s Middle Way outlines a program of meditative cultivation that results in a nondual awakening whereby all conceptual thought has been eliminated and not even nonconceptual wisdom exists in Buddhahood. DOI: 10.4324/9781351030908-52

601

James B. Apple

Historical Context Atiśa lived at a unique juncture in the history of India and Tibet. He resided in India during a revival of the East Indian Pāla Dynasties (760–1142) of Bihar and Bengal. During this time, Bengal was an “international” region with trade routes on land from Assam and Burma passing through ancient Buddhist pilgrimage holy sites of Magadha, including Vajrāsana, the “Diamond Seat,” where Śākyamuni Buddha attained awakening, located in present-day Bodh Gayā, India. The international environment of northeast India also included maritime routes connecting the Bay of Bengal to harbors in South and Southeast Asia (Acri 2016). The Pālas ruled over the northeastern lands of Bihar, West and North Bengal, and explicitly claimed to be Buddhist in their inscriptions: they utilized the Dharma wheel of the Buddha’s teachings on the top of inscribed copperplates and, in the colophons of manuscripts they sponsored, described themselves as “entirely devoted to the Buddha” (Sanderson 2009). Furthermore, the Pālas supported a perpetual endowment of donations for the Buddhist monastic community and established a network of major monasteries, including the prestigious centers of Vikramaśīla and Somapura. Atiśa, the Princely-Lord, came of age, studied, meditated, and taught during this buoyant time of maritime Buddhist Asia. The Pāla dynastic period was permeated with the religious practices of esoteric Buddhism, or Secret Mantra practices. This form of Buddhism, based on groups of texts called tantras, was primarily disseminated from master to disciple. Institutional monastic esoterism and the esoterism of siddhas, accomplished adepts on the margins of society, dominated the Buddhist culture of the time. By the late tenth and early eleventh century, during the lifetime of Atiśa, Buddhist monastic communities and siddha culture had undergone almost two centuries of blending and accommodation. The accommodation of esoteric Buddhism into monastic communities was influenced, in part, by a highly competitive environment between Buddhists and non-Buddhists seeking patronage and economic support. Esoteric Buddhists lived in a cultural context, where Buddhists who did not support esoteric practices, as well as non-Buddhist esoteric practitioners including the diverse groups of Śaivas (followers of Śiva), contested for prestige and authority. These groups took part in a whole range of ascetic and ritual practices. While residing in India before his journey to Tibet, Atiśa served as a monastic official at monasteries in Bodh Gayā and Vikramaśīla. One way to understand the form and content of the Mahāyāna Buddhism practiced and taught by Atiśa is to compare Atiśa with his esteemed colleagues at Vikramaśīla. Most of Atiśa’s colleagues there followed various forms of Yogācāra thought and worked with the philosophical traditions of epistemology based on the texts of the seventh-century Buddhist thinker Dharmakīrti. His colleagues at Vikramaśīla were also holders of diverse monastic ordination lineages: Atiśa’s colleague Jñānaśrīmitra, a prominent scholar who initially upheld Saindhava śrāvaka views, advocated a form of Yogācāra thought that asserts that cognitive images (ākāra) are real. Jñānaśrīmitra also formulated a sophisticated version of the exclusion (apoha) theory of language based on complex epistemological arguments (McCrea and Patil 2010). Atiśa’s senior colleague and teacher Ratnākaraśānti was ordained in the Sarvāstivāda school at Odantapurī. Ratnākaraśānti composed numerous works on subjects such as valid cognition, the Perfection of Wisdom Sūtra (Prajñāpāramitā Sūtra), Yogācāra, and Tantra, as well as Buddhist verse metrics (cansaḥśāstra) and riddles (Isaacson 2013). In terms of his viewpoint, Ratnākaraśānti articulated a middle way based on Yogācāra principles that incorporated the theory of the three natures (trisvabhāva) with an emphasis on self-awareness (svasaṃvedana) as equivalent to luminosity ( prakāśa) (Apple 2018a). Atiśa’s junior colleague Ratnakīrti followed a nondual consciousness (citrādvaitavāda) understanding of Yogācāra and articulated sophisticated arguments against the Nyāya school’s God-like being called “Īśvara” (Patil 2009). In contrast 602

Atiśa

to all of his known Vikramaśīla colleagues, Atiśa followed a lineage of Madhyamaka based on Nāgārjuna and Candrakīrti that was transmitted through his teachers Avadhūtipa, Bodhibhadra, and Vidyākokila, with whom he studied in his youth. Such differences between Atiśa and his colleagues may have been a contributing factor in Atiśa’s decision to leave Vikramaśīla for Tibet. At that time, the kings in West Tibet were seeking to rejuvenate Buddhism in order to replicate the order, ethical principles, and stability that Mahāyāna Buddhist ideals had brought to the Tibetan Empire during the seventh to ninth centuries. In seeking out Atiśa as the rejuvenator of Mahāyāna Buddhist ideals, the Tibetans likely initially considered him a descendant of the great bodhisattva-scholar Śāntarakṣita, who was also from Bengal. Śāntarakṣita was responsible for establishing the first ordained monks in Tibet and contributing to the translation of texts and the construction of Tibet’s first monastery of Samyé (Blumenthal 2004). Although Atiśa and Śāntarakṣita may have been from the same region, Atiśa lived in an era of Indian Buddhism different than Śāntarakṣita’s: while Śāntarakṣita was a Mūlasarvāstivāda monk who blended Yogācāra and Madhyamaka philosophies, Atiśa upheld the Mahāsāṃghika ordination lineage and was a follower of the Madhyamaka tradition of Candrakīrti. Atiśa was also a lineage holder of a number of Yoginī tantras, such as the Laghuśaṃvaratantra and Hevajratantra, which were only introduced into Tibetan culture in the late tenth century by Rinchen Zangpo (958–1055). Tibetans at this time were not fully familiar with these areas of monastic discipline, Madhyamaka philosophy, and esoteric Buddhism that Atiśa brought with him to Tibet.

Writings Throughout his life in India, Indonesia, Nepal, and Tibet, Atiśa composed over 100 works. Recent publication of his collected writings in Tibetan organize his compositions into those related to view, conduct, union of view and conduct, and Secret Mantra practices. Atiśa focused his writings and teachings on essential practices and integrative meditative cultivations leading to the nondual realization of full buddhahood (Miyazaki 2007a). Such essential practices consist of teachings that Atiśa distilled from the vast corpus of Buddhist scriptural and commentarial literature that was preserved at the monasteries of Vikramaśīla, Nālandā, and Somapura. He composed several works focusing on essential practice, including the Essential Summary (Garbhasaṃgraha), the Essence Clearly Explained (Hṛdayanikṣepa), Lamp for the Summary of Conduct (Caryāsaṃgrahapradīpa), and Bodhisattva’s Jewel Garland (Bodhisattvamaṇyāvalī). In these works, Atiśa emphasized cultivation of the awakening mind, moral conduct, and practices for generating merit. The Open Basket of Jewels (Ratnakaraṇḍodghāṭamadhyamakopadeśa; Miyazaki 2007b; Apple 2019a, 63–113) is one of the primary works among Atiśa’s collected writings and perhaps the most extensive of his extant writings composed in India. Within this work, Atiśa cites sūtras and tantras that provide one of the most thorough overviews on the awakening mind (bodhicitta) in Indian Buddhist literature. Atiśa’s Lamp for the Path to Awakening (Bodhipathapradīpa; Apple 2019b) is his most well-known work. The text contains sixty-eight verses outlining the integration of three forms of discipline, including the vows of the monastic disciplinary code ( prātimokṣa), bodhisattva precepts, and the precepts of the Secret Mantra Vehicle. Mahāyāna and Vajrayāna practices and cultivations are also discussed. Written in response to questions from King Jangchup Ö in West Tibet around 1042, Atiśa’s Lamp became “one of the most influential of Indian texts received by Tibetans” and was “the model for mainstream Tibetan monastic Buddhists for the next nine hundred years” (Davidson 1995, 293). The Song with a Vision for the Realm of Reality (Dharmadhātudarśanagīti; Blo bzang rdo rje rab gling 1999; Apple 2019b), which elucidates the dharmadhātu, or realm of reality, provides a summation of Buddhist and non-Buddhist philosophical views on reality in a little over 140 verses. Atiśa discusses 603

James B. Apple

the realm of reality in the first twenty-five verses, fourteen of which are based on the work of Nāgārjuna. Atiśa then summarizes Buddhist views (vv. 26–96), beginning with Madhyamaka and descending to Vaibhāṣika tenets, follows this with non-Buddhist philosophical tenets (vv. 97–142), and then provides concluding verses (vv. 143–147) (Apple 2019b). Atiśa’s Madhyamaka thought has traditionally been understood based on his Entry to the Two Realities (Satyadvayāvatāra), Special Instructions on the Middle Way (Madhyamakopadeśa), and Commentary on the Difficult Points in the Lamp for the Path to Awakening (Bodhimārgapradīpapañjikā, D. 3948; Sherburne 2000). Entry to the Two Realities and Special Instructions on the Middle Way are considered by traditional Gelukpa historians the two foremost textual teachings (gzhung) on the view (lta ba) within Atiśa’s works (Apple 2019a, 270). Entry to the Two Realities succinctly lays out in twenty-eight verses a general exposition on the two realities. Atiśa composed this work between 1012 and 1025 while residing in Sumatra and studying under Serlingpa. Serlingpa inquired about Atiśa’s philosophical views in a letter, and Atiśa composed this set of verses as a response. Atiśa wrote Entry to the Two Realities as an attempt to transform Serlingpa’s philosophical view from a Yogācāra position to a Madhyamaka one. This introductory text on Madhyamaka presents Atiśa’s understanding based on the synthesis of a number of previous Indian Madhyamaka thinkers. Special Instructions on the Middle Way is Atiśa’s advice for self-transformation through the practice of Madhyamaka, given in Lhasa at the request of Ngok Lekpai Sherap. The brief text provides instructions for how Mādhyamikas meditate, cultivating the three wisdoms of learning, reflection, and meditation within the context of meditative equipoise and postmeditative wisdom construed through the purviews of conventional and ultimate reality. A General Explanation of, and Framework for Understanding, the Two Realities (Bden gnyis spyi bshad dang/bden gnyis’jog tshul; hereafter A General Explanation) is a late eleventh-century Indo-Tibetan Madhyamaka text that records oral teachings attributed to Atiśa on the two realities (satyadvaya). The text furnishes an exposition of the Middle Way thought of Nāgārjuna based on an exegesis of conventional reality and ultimate reality within the framework of Mahāyāna path structures found in texts attributed to Maitreyanātha (Apple 2019a, 171–266). A General Explanation provides an early account of points of Madhyamaka exegesis that would go on to become polemical points of debate in later decades and centuries in Tibetan Buddhist thought. Atiśa discusses such points – whether Mādhyamikas have a thesis, an inference that is known to others (gzhan la grags pa’i rjes dpag), the object that is negated (dgag bya), the negation of self-characteristics (rang gi mtshan nyid ≈ svalakṣaṇa), and the notion that two things are “a single nature but different conceptual isolates” (ngo gcig ldog pa tha dad) – in a late eleventh-century Indo-Tibetan Buddhist historical context.1 The text emphasizes that correct conventional realities are indicated through nonimplicative negations (med dgag) and that things are mere appearances that are transactually designated without being established.

Atiśa’s Middle Way Thought Atiśa was an adherent to Madhyamaka thought and practice with a basis in the works of Nāgārjuna. For Atiśa, Nāgārjuna was a towering figure of Buddhist culture who not only had great insight and realization but also sparked innovations in other areas such as politics and medicine. Atiśa cites and comments on a number of texts attributed to Nāgārjuna – such as the Commentary on the Awakening Mind (Bodhicittavivaraṇa), Twenty Verses on the Great Vehicle (Mahāyānaviṃśīkā), Verses on the Heart of Dependent Arising (Pratītyasamutpādahṛdaya kārikā), and Stages of Meditation (Bhāvanākrama) – that he considered vital to understanding the Middle Way. These texts were not yet fully translated into Tibetan by the eleventh century 604

Atiśa

and are often not included in the exegesis of Madhyamaka by later traditional Tibetan scholars or by modern scholars. As such, Atiśa followed a commentary attributed to Nāgārjuna, No Fear from Anywhere (Akutobhayā), for his interpretation of verses found in the Fundamental Verses of the Middle Way (Mūlamadhyamakakārikā). Atiśa understood Nāgārjuna’s devotional praises, such as the Hymn to the Realm of Reality (Dharmadhātustava), as complementary to his works emphasizing reasoning, such as the Sixty Stanzas on Reasoning (Yuktiṣaṣṭikā). Nāgārjuna was thus the principal master of Atiśa’s Madhyamaka in this holistic and inclusive way of interpretation. Atiśa also considered Nāgārjuna to have lived for 600 years and to have written esoteric Buddhist works. Moreover, Atiśa described his teachers’ visionary encounters with Nāgārjuna as part of an ongoing revelatory lineage of Madhyamaka. Atiśa also considered Candrakīrti had lived for 400 years in India and been a direct disciple of Nāgārjuna. Atiśa’s Madhyamaka lineage represented a contemplative tradition that emphasized a cultivation of the resolution for awakening, the development of compassion, and the realization of emptiness leading to the inconceivable state of buddhahood. Although reasoning does have its place in Atiśa’s system, he advocated a faith-based Madhyamaka that valued the instructions of the spiritual teacher and held predictions, prayers, and meditative cultivation in high regard. Atiśa’s Middle Way synthesized the teachings of Bhāviveka and Candrakīrti, built on the basis of Nāgārjuna’s teaching. In both India and Tibet, Atiśa pedagogically utilized the works of Bhāviveka – particularly his Blaze of Reasoning (Tarkajvālā) and Jewel Lamp of the Middle Way (Madhyamakaratnapradīpa) – as an introduction to Madhyamaka and then taught advanced students Candrakīrti’s system as exemplified in the Introduction to the Middle Way (Madhyamakāvatāra). Atiśa primarily synthesized the teachings of Bhāviveka and Candrakīrti by bringing together compatible elements of their teachings for soteriological efficacy in progression on the path. He emphasized the commonality of Bhāviveka and Candrakīrti’s teachings, as both pertain to practices within conventional reality to reach the goal of realizing ultimate reality while downplaying any minor points of philosophical difference between the two thinkers. Atiśa identified this commonality between Bhāviveka and Candrakīrti as “the stairway of correct conventional reality” ( yang dag kun rdzob kyi skas), providing him with an avenue for advocating conventional practices, such as moral discipline and meditation, as the dependently arisen means that lead toward the goal of realizing ultimate reality. Atiśa’s Madhyamaka thought and practice underlie his understanding of advanced bodhisattva practices within both the Perfection and Secret Mantra Vehicles. In Mahāyāna forms of Buddhism, the term “vehicle” ( yāna) is a metaphor for the way or path that a bodhisattva takes to achieve Buddhahood. The exoteric way is the vehicle of the Perfections ( pāramitā) in which a bodhisattva practices perfections, such as morality, patience, and wisdom, as outlined in Mahāyāna scriptures (i.e., sūtras). The esoteric way is the vehicle of Secret Mantra in which a bodhisattva receives a consecration from a qualified preceptor in secret to utilize mantras, maṇḍalas, and mudrās to quickly achieve Buddhahood as outlined in Mahāyāna ritual texts (i.e., tantras). Atiśa called his understanding of Madhyamaka, or the Middle Way, “Great Madhyamaka.” Atiśa’s classification of Great Madhyamaka, in brief, represents his effort to differentiate the meaning of the Perfection of Wisdom as taught by Nāgārjuna from its meaning as taught by the Yogācāra scholar Asaṅga. Atiśa consistently upheld a Madhyamaka view over and against Yogācāra ontological tenets in all of his known works. Along these lines, Atiśa rejected the utilization of epistemological cognitions for the realization of emptiness. As Atiśa states in his Entry to the Two Realities (v. 10), Direct perception and inference are the two valid cognitions ( pramāṇas) accepted by Buddhists. The deluded whose vision is narrow say that emptiness is understood by these two. 605

James B. Apple

For Atiśa, the science of epistemology and logic was a secular science common to Buddhist and Indian non-Buddhist schools. Cognitions derived from epistemological warrants were, for Atiśa, only utilized at the level of conventional reality to refute opponents. According to Atiśa, valid cognitions are only conventional and are not in the domain of ultimate reality, nor are they able to realize ultimate reality. Along these lines, Atiśa also disparaged the practice of debate in his A Brief Treatise on the Method of Practicing the Great Vehicle Path (Mahāyānapathasādhana-varṇasaṃgraha; Sherburne 2000, 452–53): Neglecting one’s hard-to-tame mental stream while practicing argument in order to learn debate, or engaging in the explanation of the teaching in every moment of the day and night for worldly things such as fame and so forth, life quickly passes on without purpose and one degenerates from the supreme path. However, although Atiśa did not advocate the practice of debate, this does not mean that he rejected the use of reasoning ( yukti) – for Atiśa, reasoning is an essential method that leads to the direct realization of insight. Reasoning thus signifies an “internal” Buddhist form of critical analysis that is different from the “external” epistemological devices used to defend Buddhist Dharma and defeat non-Buddhist opponents. A General Explanation specifies that the object of negation of reasoning is a conceived object based on conceptualization that imputes things as either existent or nonexistent. The object negated by reasoning consists of conceptual thought that imputes an object as existing with its own character. According to Atiśa, things appear to be substantially existent but, under analysis through reasoning, are realized to subsist without essence. Atiśa refers to appearances without substance as mere appearances. However, Atiśa stresses in this work that unestablished mere appearances are not refuted by reasoning. Rather, appearances are overturned through antidotes cultivated while practicing the path, particularly the path of vision (darśanamārga) and the path of meditation (bhāvanāmārga). The vision of gnosis cultivated in Atiśa’s Middle Way meditation instructions, in brief, dissipates the cognition of appearances as real. A General Explanation offers an early distinction between objects negated by an antidote (gnyen po’i dgag bya) while implementing the path and objects negated by reasoning (rigs pa’i dgag bya) when searching out the inherent existence of something (Apple 2019a, 207–8). Atiśa states “four great reasons” (gtan tshigs chen po bzhi) for the cultivation of insight in meditation in his Commentary on the Difficult Points in the Lamp for the Path to Awakening (Bodhimārgapradīpapañjikā) and also in A General Explanation. The “four great reasons” proving emptiness for Atiśa are: (1) the reason refuting production according to the tetralemma (mu bzhi skye ’gog gi gtan tshigs; catuṣkoṭyupādapratiṣedhahetu), (2) the “diamondsplinters” reason (rdo rje gzegs ma’i gtan tshigs; vajrakaṇahetu), (3) the reason of being neither one nor many (gcig dang du ma bral ba’i gtan tshigs; ekānekaviyogahetu), and (4) the reason consisting in dependent origination (rten ’brel gyi gtan tshigs; pratītyasamutpādahetu). Atiśa’s four reasons constitute a system of analytical procedures to prove the absence of intrinsic essence in all things. For Atiśa, the (4) reason consisting in dependent arising is based on the argument that because things arise due to causes and conditions that are mutually related, they therefore lack an intrinsic essence. Atiśa claims that the reasoning of dependent arising (rten ’grel gyi rigs pa) contains all four reasonings within it. The (1) reason refuting production according to the tetralemma, or four points, is an analytical procedure which demonstrates that things lack intrinsic essence because (i) something already intrinsically existent cannot dependently arise, and, likewise, (ii) something nonexistent cannot arise. Things also cannot arise (iii) through being existent and nonexistent at the same time, 606

Atiśa

nor (iv) through being neither existent nor nonexistent. The (2) diamond-splinters reason (rdo rje gzegs ma), metaphorically like a diamond that crushes the rock of substantial views, is an analytical procedure that examines the possibility of intrinsic essence from the point of view of the cause. The (3) reason of being neither one nor many, which examines both the cause and effect in determining the possibility of intrinsic nature, argues against things having intrinsic identity or intrinsic plurality in proving the lack of intrinsic essence in things. Atiśa’s system of positing four reasons for proving emptiness is different from Kamalaśīla’s, which includes five reasons (Keira 2004, 10–13). Atiśa leaves out the reason refuting the production of existent and nonexistent things ( yod med skye ’gog gi gtan tshigs, *sadasadutpādapratiṣedhahetu) that is discussed by earlier Mādhyamikas. A General Explanation clarifies that these reasons are based on the reasoning of dependent arising and that all four reasons are accepted as consequences that nonimplicatively negate the intrinsic existence of things but do not negate the mere appearance of causes and effects. The mere appearances that arise from causes and effects are overturned through antidotes cultivated while practicing the path. Based on statements in the Kadampa Collection on the Two Realities and A General Explanation, Atiśa followed a lineage of Madhyamaka that advocated the use of consequences that exposed contradictions and employed other-acknowledged inferences (Apple 2019a, 39). In fact, A General Explanation contains the earliest extant mention of the four types of consequence used by Mādhyamikas. The four types of consequence are (1) consequences that compose contradictions (’gal ba brjod pa’i thal ’gyur, *virodhacodanāprasaṅga), (2) the inference that is known to others (gzhan la grags pa’i rjes dpag), (3) the evidence that is not established due to the equivalence with what is being established (bsgrub bya dang mtshungs pa’i ma grub pa), and (4) the equivalence of the reason. Atiśa explains these four as follows: The text of the Ācārya Nāgārjuna states (1) consequences that expose contradictions (gal ba brjod pa’i thal ’gyur); (3) a pseudo-sign similar to what is to be proven; (4) equivalence (mgo bsgre ba): “If you accept in this way, because the reason is not different, you must accept this as well.” These are bound to the opponent. (2) Inference that is known to others (gzhan la grags pa’i rjes dpag) in which one states: “If you yourself accept in this way, your own understanding is contradictory with this conclusion.” The property of the subject and the entailment are bound with the opponent and are established by their acceptance. Even though these arguments by consequence are proclaimed through four reasons, they are not different than being included within the reason of dependent-arising. (Apple 2019a, 204) Atiśa did not accept that the Madhyamaka posits a thesis (dam ’cha’ ba ≈ pratijñā), nor that a Madhyamaka concedes to the principle of common establishment. The early Kadampa commentaries provide evidence that Atiśa’s early followers accepted his position on Madhyamaka reasoning of not adhering to the principle of common establishment (ubhayasiddhatva), which became in later Tibetan scholarship the contentious issue of “commonly appearing subjects” (chos can mthun snang ba) (Apple 2019a, 39). The principle of common establishment, traceable to statements made by Dignāga (fifth to sixth century), was that the reason and subject must be established for both parties in a debate. That is, for both parties in a debate, there must be something that similarly appears (mthun snang ba) in order for logical debate and discussion to take place. Some Mādhyamikas in India accepted this principle; others did not. Śāntarakṣita, Kamalaśīla, and Jñānagarbha accepted it at least on the conventional level (Keira 2004). Candrakīrti, on the other hand, appears to be 607

James B. Apple

one of the few scholars within Indian Madhyamaka traditions not to have followed the rule of common establishment. Atiśa followed this lineage tradition (Apple 2019a, 38). Atiśa compares the reasoning process to two sticks, which, after rubbing together and generating a fire, burn up and become nonexistent. Although Atiśa does not state his textual source, he draws this example from the Chapter on Kāśyapa (Kāśyapaparivarta), which is cited in the Jewel Lamp of the Middle Way. Kamalaśīla also cites this sūtra in his Stages of Meditation (Bhāvanākrama) and Commentary on the Dhāraṇī “Entering into Non-Conceptuality” (Avikalpapraveśadhāraṇīṭīkā) as an example to illustrate that although the analysis of reality is indeed the nature of conceptual thought, it will nevertheless be consumed by the fire of correct wisdom produced by it (Apple 2018c). In his Open Basket of Jewels, Atiśa states that “the wisdom of individual analysis (so sor rtog pa’i shes rab) itself turns into clear light.” Atiśa’s Indian disciple Prajñāmukti is even clearer in his Commentary on the Special Instructions (Madhyamakopadeśavṛtti; Apple 2019a, 285–86): That the very wisdom which individually discriminates is not established either means that analytical cognition negates itself. Since wisdom is a particularity of an entity, when an entity is not established, the very wisdom itself is also not established, just like when a tree is not established, the wood and so forth are negated. As it is said, “In this regard, a fire which burns fuel, having burned its fuel does not remain.” These passages indicate that, for Atiśa and his followers, reasoning is a conventional process that dissolves itself when seeking to establish the existence of an object. Analytical reasoning which dissolves itself is preparatory for wisdom – more specifically, for non-conceptual gnosis (nirvikalpajñāna). The texts suggest a difference between discernment (  prajñā) at the level of learning and the reflection, reasoning (rigs pa’i shes rab), and non-conceptual realization that constitute gnosis (  jñāna). The numerous reasonings which appear in Atiśa’s works and his Kadampa commentaries are meditations that lead the reader through merelogical forms of analysis to dissolve the conceptual thought that reifies things and their relations. Atiśa did not differentiate Madhyamaka into separate schools. Instead, he utilized the texts of various Madhyamaka authors for pedagogical and soteriological purposes. Atiśa introduced his students to the thought and practice of Madhyamaka with the works of Bhāviveka. To support more advanced levels of understanding, Atiśa closely followed the works of Nāgārjuna, Candrakīrti, and Śāntideva. According to Atiśa’s Madhyamaka, most prominently in the General Explanation (Apple 2019b, 191–266), all appearances are based on ignorance and composed of ignorance. Appearances are not nonexistent but rather false, erroneous, and mistaken. The synonyms of the conventions of mere appearance in Atiśa’s nominalistic Madhyamaka are mere name, mere word, mere convention, and mere imputation. For Atiśa, the subjective perceiver, the appearance of the cognition, and the appearance of the mind, which is like an illusion, are only imputedly established. Thus, the cognizer, what is cognized, and the phenomenal awareness that occurs in such a conjunction of ephemeral conditions lack any substantive ontological status, either conventionally or ultimately, in Atiśa’s Madhyamaka. A key Madhyamaka classification that structures Atiśa’s thought and practice is the bifurcation of all things in the universe into the categories of two realities: ultimate reality and conventional reality. As Atiśa states in his Song with a Vision for the Realm of Reality, ultimate reality is a purified appearance of nonappearance like the center of space, known through individually intuited knowledge. For Atiśa, the “ultimate” is a conventional expression, also embodied by the phrase “realm of reality” (dharmadhātu), consisting of selfless nonappearances that are realized with nonconceptual pristine awareness (  jñāna). The realm of reality, for Atiśa, is a 608

Atiśa

naturally pure object of realization that is the ever-present real state of things that may be cognized in meditative equipoise but is not completely actualized until full buddhahood. Conventional realities are appearances that arise due to causes and conditions. Though all conventional realities are false and deceiving, they are not nonexistent. Rather, they are mere appearances, subject to the principle of cause and effect imputed through dependent arising. In Atiśa’s system, conventional realities are classified either as mistaken or correct when viewed from different perspectives in relation to the cognitive understanding of ordinary individuals or the realizations of those who have reached the path of vision. The three ways of identifying (’jogs lugs) mistaken and correct conventional realities are explained based on Candrakīrti’s Introduction to the Middle Way. Conventional realities are dependently designated in relation to the perspectives of (1) the worldly (lo ka pa), (2) philosophical tenets (grub mtha’), and (3) yogic awareness (rnal ’byor pa’i blo). A General Explanation posits correct and mistaken conventional realities based on the nature of dependent arising in relation to yogic awareness. This understanding of the dependent arising of conventional reality in correlation with its states of awareness accords with what Wangchuk (2009) has called “the relativity theory of the purity and validity of perception” in Madhyamaka works such as Candrakīrti’s Introduction to the Middle Way (6.27) and Śāntideva’s Introduction to the Practice of Awakening (Bodhicaryāvatāra) (9.3–4a). In the words of A General Explanation, “things are not higher or lower; awarenesses are higher or lower” (Apple 2019a, 180–81). Atiśa accepted the distinction between correct and incorrect conventional realities even though he considered conventional realities false and unreal. This distinction is structured within a framework of shifting perspectives as one progresses along the path in accord with Bhāviveka’s Blaze of Reasoning commentary on the Heart of the Middle Way and Candrakīrti’s Introduction to the Middle Way (Apple 2016). As outlined in the following paragraphs, Atiśa refers to this progress on the path as the “stairway of correct conventional reality.” Mistaken conventional realities are appearances of ignorance that impute impermanent and empty things as either existent or nonexistent. Mistakes are impermanent and cause suffering, and they are also deceptive and false. Mistaken appearances are like the hair that is perceived by someone suffering from eye disease. A General Explanation states that its acceptance of mistaken conventional realities is similar to how True Aspectarians (*Satyākāravadins) posit aspects, or cognitive images (ākāra). This process is comparable to Candrakīrti, who, as pointed out by MacDonald (2009, 151), skillfully adapts the Sautrāntika theory of cognition on the conventional level to justify his own views. Atiśa’s position on the status of external objects is similar to the presentation of “internal” Madhyamaka (nang gi dbu ma) in the Jewel Lamp of the Middle Way (D, 280a3–81a3; Del Toso 2014). However, Atiśa also states in his A General Explanation – in correlation with Nāgārjuna accepting appearances as mind in works such as the Stages of Meditation and in terms of the mind as mere appearance, a perspective equal to correct conventional reality – that all sentient beings are one single continuum (rgyud gcig; 712.1–713.5). All sentient beings are considered one continuum, for even though they have differences of karmic conditions, they share an undifferentiated self-nature (rang gi ngo bo la tha dad med pa) that is free from the two extremes of intrinsic essence. Atiśa also asserts consciousness to be one group (rnam shes tshogs gcig, *ekavijñānakāya) or one single flow, as opposed to the six distinct types of consciousnesses (ṣaḍvijñānakāya) found among mainstream Buddhist schools (Apple 2019a, 213, 399–400). Positing one group entails that one single consciousness moves to individual objects in dependence upon individual sense organs rather than six distinct active sensory consciousnesses that arise based on the contact of a sensory faculty with a sense object. In this way, the General Explanation presents a mentalist theory of Madhyamaka in which the mind, as mere appearance, is not at all established and is a mere nominal designation free from the extremes of existence and nonexistence. 609

James B. Apple

Correct conventional reality for the yogi, according to Atiśa, occurs only after completing the path of vision, a stage on the path when emptiness is directly cognized in a non-conceptual manner. Correct conventional realities are considered nondeceptive, nonerroneous, and trustworthy in that, from this perspective, appearances are realized to be unproduced like an illusion, and objects are cognized as essenceless entities. Although correct, they are conventional due to arising through causes and conditions and are thus considered illusions of pristine awareness. Correct conventional realities are nonerroneous illusions and are imputations conducive to purification, since the causes of purification have nondeceptive individual results. This accords with appearances of “purified worldly knowledge” mentioned by Bhāviveka in his Heart of the Middle Way and his Blaze of Reasoning. Atiśa refers to correct conventional reality, following Bhāviveka, as the stairway of correct conventional reality ( yang dag kun rdzob rnams kyi skas ≈ tathyasaṃvṛtisopānam). Atiśa’s understanding of correct conventional reality offered him an avenue for upholding the conventional practices of moral virtue that eventually lead to realizing ultimate reality, practices that some critics argued were undermined by Madhyamaka analysis. Based on these factors of Madhyamaka thought and practice, Atiśa outlines a program of bodhisattva training that integrates the cultivation of wisdom and compassion with the development of an awakening mind at both the ultimate and conventional levels. The ultimate awakening mind – the birthless, luminous, nonconceptual realm of reality equated with emptiness – is cultivated during meditative equipoise. The conventional awakening mind is cultivated during postmeditative practices. In this program of training, the two levels of the awakening mind are integrated and stabilized, having the essence of emptiness and compassion. A snapshot of instructions for this integrated cultivation is found in the Open Basket of Jewels (Apple 2019a). Atiśa states, Regarding the training, first the mind did not come from anywhere and will not go anywhere at the end. The mind does not abide anywhere and is without color and without shape. The mind does not arise from the beginning nor does it cease at the end. The mind is empty of inherent existence and has the nature of clear light. One should recall this again and again. On the other hand, one should stabilize through accustoming the awakening mind to love and compassion. One should completely purify the mind and stand firm, being continuously mindful of each moment of thought with mindfulness, awareness, thoughtfulness, and conscientiousness. The Open Basket of Jewels then discusses how the integrated awakening mind is protected and increased while the bodhisattva advances through the ten stages as outlined by the Sūtra of the Ten Stages (Daśabhūmikasūtra). In the course of the training, the bodhisattva alternates between cognizing the nonconceptual space-like realm of reality in meditative stabilization and then viewing things as illusions in the postmeditative state. The alternation, however, ceases at the stage of buddhahood. According to Atiśa, based on numerous citations from the hymns of Nāgārjuna, at the stage of buddhahood, the purified realm of reality directly and constantly fuses with the Dharma body (dharmakāya), the ultimate cosmic reality of buddhahood, without any mental element or gnosis existing at all. Atiśa emphatically advocated in his works that buddhas are completely fused with ultimate reality in a nondualistic fashion whereby all conceptual thought has been eliminated and not even nonconceptual wisdom exists. Throughout his writings, Atiśa references Mahāyāna sūtras; tantras; and the works of Nāgārjuna, Āryadeva, Candrakīrti, and Śāntideva as authoritative texts for this standpoint. In his Open Basket of Jewels, he provides an extended discussion, 610

Atiśa

based on the works of Nāgārjuna, of the point that buddhahood does not have any mind or mental factors. Related to Atiśa’s understanding of buddhahood as bereft of any mental qualities is his position that a buddha does not have a continuum of wisdom. This issue is specifically related to the interpretation of a partial stanza in Śāntideva’s Introduction to the Practice of Awakening (9.15ab) that states, “upon conditions having their continuum cut, an illusion does not arise even conventionally.” Atiśa directly repeats this portion of Śāntideva’s stanza in his Entry to the Two Realities (vv. 23cd). An early Kadampa commentary to this verse, likely following an oral tradition of Atiśa’s teachings, explains that appearances occur due to various causes and conditions, and if the continuance of the conditions for such appearances is interrupted, then appearances will no longer arise even conventionally. For Atiśa and his Kadampa followers, all appearances are due to ignorance and are composed of ignorance. Therefore, when the conditions for any type of appearance are exhausted, including wisdom or gnosis (  jnāna), then such appearances will no longer occur. In addition to this discussion within the Entry to the Two Realities, Atiśa directly addressed the status of gnosis at the level of a buddha in his Commentary on the Introduction to the Practice of Awakening (Bodhisattvacāryāvatārabhāṣya), a summary of Śāntideva’s Introduction to the Practice of Awakening. Several sections of the text are composed in the form of a dialogue with the bodhisattva Mañjuśrī to address difficult points of exegesis. Atiśa discusses the gnosis at the level of a buddha in a section describing the vision of liberating gnosis (rnam par grol ba’i ye shes mthong ba ≈ vimuktijñānadarśana) at the level of a buddha. Atiśa’s discussion in this work clearly demonstrates that he did not accept that a buddha would have a continuum of wisdom. On this point, Atiśa’s understanding is congruent with a strict Madhyamaka understanding of the nature of buddhahood found in the works of Nāgārjuna, Candrakīrti, and Śāntideva (Apple 2019a, 43–44). This understanding of buddhahood would not be followed by a majority of Tibetan scholars, particularly after the time of Chapa Chökyi Sengé (1109–1169). In his works, Atiśa outlined an undifferentiated Madhyamaka tradition. In Open Basket of Jewels, he emphasizes that his teachings focus on the lineage of Nāgārjuna and defends “the Mādhyamika followers of Nāgārjuna” as having no faults. His Commentary to the Lamp for the Path to Awakening documents a tradition that passed from Nāgārjuna through Āryadeva, Candrakīrti, Bhāviveka, and Śāntideva down to Bodhibhadra. This work also briefly describes his vision of Madhyamaka as Great Madhyamaka (dbu ma chen po) in its section on insight (shes rab). Atiśa’s General Explanation claims that Great Madhyamaka represents the definitive understanding of Nāgārjuna’s thought. Atiśa’s epithet of Great Madhyamaka, in brief, signifies his understanding of Nāgārjuna’s Middle Way as a realization free of any extremes or conceptual proliferations, a direct vision of the non-duality of appearances and emptiness. In his own writings, therefore, Atiśa did not distinguish between individual followers of Nāgārjuna, that is, Mādhyamikas (Apple 2019a, 59–60). Madhyamaka, for Atiśa, was what Ruegg (1981, 30, 57, 59) has labeled “pure” Madhyamaka, a designation that signifies a Middle Way tradition that is not differentiated based on features of individual doctrines upheld in the works of Indian Buddhist authors.

Stages of the Path Atiśa’s greatest contribution and legacy was his synthesis and reintegration of the teachings of sūtras and tantras into a coherent system of Buddhist practice. Atiśa’s well-known Lamp for the Path to Awakening and his important but lesser-known Stages for the Path to Awakening (Byang chub lam gyi rim pa ≈ Bodhipathakrama; hereafter Stages) are long-standing testaments to his innovative and dynamic system. Atiśa began to formulate teachings that integrate sūtra and mantra in his early writings on the esoteric Buddhist deity Cakrasaṃvara. His Analysis of Realization (Abhisamayavibhaṅga) articulates the “stages of the path of the 611

James B. Apple

essential meaning,” The stages of the path were initially formulated by Atiśa based on these advanced stages of esoteric Buddhist teachings. Atiśa further developed his stages of the path system under the urging of Jangchup Ö while he was in West Tibet. Atiśa’s Lamp was the result of Jangchup Ö’s patronage and inquiries, and the treatise became a model for subsequent generations of Tibetan scholars in their understanding and practice of the Buddhist path. The Lamp was an imperially sanctioned public teaching (tshogs chos) that influenced Tibetan Buddhists for over nine centuries. The Stages, on the other hand, written at the behest of Dromtönpa, was a secret teaching (lkog chos) that was transmitted among early followers of Atiśa until the time of Pakmo Drupa (1110–1170). Although the Stages has not been widely studied (or even recognized) by later traditional and modern scholars of Tibetan Buddhism, its impact was just as great as the Lamp: the Stages contains within it all the major topics found in later stages of the path literature that the Lamp does not mention. The Stages clearly outlines the practices and realizations of the individuals of small and middling capacity, including such topics as taking refuge, meditations on death and impermanence, the elimination of wrongdoing, and recollecting the sufferings of cyclic existence. Although Atiśa’s Stages is primarily devoted to the ethics of karmic cause and effect, in the context of instruction on pointing out a nonconceptual direct vision of the emptiness of one’s own mind, it also develops an account of the Middle Way between the extremes of nihilism and permanence: All dharmas are the mind, the mind itself is free from all extremes. The multiple various causes and effects of virtue and wrongdoing is unceasing, definitely free from the extreme of nihilism. Since whatever appears of the cause and effect of the round of rebirth and nirvāṇa is the nature of one’s own mind, which is not at all established, it is definitely free from the extreme of permanence. Emptiness indivisible with cause and effect is the nature of one’s own mind, free from the proliferations of extremes, the Great Middle Way. (Apple 2019b) These two stanzas situate Atiśa’s Great Middle Way between the extremes of nihilism and permanence based on the nonduality of emptiness and cause and effect. They occur in the instructions on tranquility and insight and represent the type of instruction given to disciples of advanced spiritual capacity. Later Kagyüpa scholars such as Gampopa (sgom po pa bsod nams rin chen, 1079–1153) and Pakmo Drupa ( phag mo gru pa rdo rje rgyal po, 1110–1170) described them as Mahāmudrā, or Great Seal, teachings (Apple 2019a, 52–53). The topics found in the Stages and the sequence in which they are presented impacted all subsequent stages of the path literature in Tibet.

Conclusion Atiśa was an Indian Buddhist itinerant teacher whose instruction was influenced by historical conditions of time, place, and patronage. Atiśa brought with him to Tibet an active Madhyamaka lineage, teaching the works of Nāgārjuna based on those of Candrakirti. This Madhyamaka lineage was contemplative in nature and based on faith, compassion, and resolutions to attain a miraculous form of buddhahood rather than on formal logical proof, linguistic semantics, or metaphysical speculation. Atiśa’s Madhyamaka lineage was actively taught and followed for at least fifty years after he came to Tibet. His teachings were disseminated before the rise of the early Kadampa monastery of Sangphu and its debating traditions that, particularly beginning 612

Atiśa

in the twelfth century, emphasized the merger of Madhyamaka with epistemology and came to shape all later forms of Tibetan Buddhist scholasticism.

Note 1 The intellectual history of most of these topics in Buddhist thought in general, and specifically in Indian Mādhyamika traditions, have yet to be adequately studied. Ruegg (2000, 2010) has devoted considerable attention to the topic of whether a Mādhyamika has a thesis. For discussion of the other topics in later post–fourteenth-century Tibetan Buddhist understandings, see Cozort (1998) and Ruegg (2002).

Bibliography Acri, Andrea, ed. 2016. Esoteric Buddhism in Mediaeval Maritime Asia: Networks of Masters, Texts, Icons. Singapore: ISEAS – Yusof Ishak Institute. Apple, James B. 2010. “Atiśa’s Open Basket of Jewels: A  Middle Way Vision in Late Phase Indian Vajrayāna.” The Indian International Journal of Buddhist Studies 11: 117–98. ———. 2013. “An Early Tibetan Commentary on Atiśa’s Satyadvayāvatāra.” Journal of Indian Philosophy 41: 263–329. ———. 2015. “A Study and Translation of Atiśa’s Madhyamakopadeśa with Indian and Tibetan Commentaries.” Acta Tibetica et Buddhica 7: 1–82. ———. 2016. “An Early Bka’-gdams-pa Madhyamaka Work Attributed to Atiśa Dīpaṃkaraśrījñāna.” Journal of Indian Philosophy 44: 619–725. ———. 2017. “Atiśa’s Teachings on Mahāmudrā.” The Indian International Journal of Buddhist Studies 18: 1–42. ———. 2018a. “Atiśa and Ratnākaraśānti as Philosophical Opponents with Attention to Yuktiṣaṣṭikā, Verse 34.” Critical Review for Buddhist Studies 23: 9–38. ———. 2018b. “Khu lo tsā ba’s Treatise: Distinguishing the Svātantrika/*Prāsaṅgika Difference in Early Twelfth Century Tibet.” Journal of Indian Philosophy 46: 935–98. ———. 2018c. “The Spiritual Exercises of the Middle Way: Reading Atiśa’s Madhyamakopadeśa with Hadot.” In Hadot and Buddhist Thought, edited by Luis Gómez and David Fiordalis, 105–44. Berkeley: Mangalam Research Center. ———. 2019a. Jewels of the Middle Way: The Madhyamaka Legacy of Atiśa and His Early Tibetan Followers. Somerville, MA: Wisdom Publications. ———. 2019b. Atiśa Dīpaṃkara: Illuminator of the Awakened Mind. Boulder, CO: Shambhala. Atiśa Dīpaṃkaraśrījñāna. Bodhisattva’s Jewel Garland. Bodhisattvamaṇyāvalī. Byang chub sems dpa’i nor bu’i phreng ba. Tôh. no. 3951 (also, no. 4471). Dergé Tanjur, vol. khi, ff. 294b7–296a1. Tibetan translation by the author and Tshul khrims rgyal ba. ———. Commentary on the Difficult Points in the Lamp for the Path to Awakening. Bodhimārgapradīpapanjikā. Byang chub lam gyi sgron ma’i dka’ ’grel. Tôh. no 3948. Dergé Tanjur vol. khi, ff. 241a4–293a4. Tibetan translation by Atiśa and Tshul khrims rgyal ba. Translated and Tibetan edition, Sherburne 2000. ———. Entry to the Two Realities. Satyadvayāvatāra. Bden pa gnyis la ’jug pa. Tôh. no. 3902 (also, no. 4467). Dergé Tanjur, vol. a, ff. 72a2–73a7. Tibetan translation by the author and Rgya Brtson ’grus seng ge. ———. Lamp for the Path to Awakening. Bodhipathapradīpa. Byang chub lam gyi sgron ma. Tôh. no. 3947. Dergé Tanjur, vol. khi, ff. 238a6–241a4. Tibetan translation by the author and Dge ba’i blo gros. ———. Lamp for the Summary of Conduct. Caryāsaṃgrahapradīpa. Spyod pa bsdus pa’i sgron ma. Tôh. no. 3960 (also, no. 4466). Dergé Tanjur, vol. khi, ff. 312b2–313a7. Tibetan translation by the author and Tshul khrims rgyal ba. ———. Open Basket of Jewels. Ratnakaraṇḍodghāṭa-nāma-madhyamakopadeśa. Dbu ma’i man ngag rin po che’i za ma tog kha phye ba. Tôh. no. 3930. Dergé Tanjur, vol. ki, ff. 96b1–116b7. Tibetan translation by the author, Rgya Brtson ’grus seng ge, and Tshul khrims rgyal ba. ———. The Song with a Vision for the Realm of Reality. Dharmadhātudarśanagīti. Chos kyi dbyings lta ba’i glu. Tôh. no. 2314 (also, 4475). Dergé Tanjur, vol. zhi, ff. 254b7–260b5. Tibetan translation by the author and Tshul khrims rgyal ba. See also Blo bzang rdo rje rab gling 1999, 66–214. ———. “Stages of the Path to Awakening. Bodhipathakrama. Byang chub lam gyi rim pa.” In Byang chub lam gyi rim pa, 21–63.7. Leh, Ladakh: Thupten Tsering, 1973. TBRC W1KG506. Also in Bka’

613

James B. Apple gdams gsung ’bum phyogs bsgrigs glegs bam go gcig, 21–202. Dpal brtsegs bod yig dpe rnying zhib ’jug khang, 2015. Blo bzang rdo rje rab gling. 1999. Five Treatises of Ācārya Dīpaṃkaraśrījñāna. Sarnath, Varanasi: Central Institute of Higher Tibetan Studies. Blumenthal, James. 2004. The Ornament of the Middle Way: A Study of the Madhyamaka Thought of Śāntarakṣita. Ithaca, NY: Snow Lion Publications. Cozort, Daniel. 1998. Unique Tenets of the Middle Way Consequence School. Ithaca, NY: Snow Lion Publications. Davidson, Ronald M. 1995. “Atiśa’s A Lamp for the Path to Awakening.” In Buddhism in Practice, edited by Donald S. Lopez Jr, 290–301. Princeton, NJ: Princeton University Press. Del Toso, Krishna. 2014. “Some Problems concerning Textual Reuses in the Madhyamakaratnapra-dīpa, with a Discussion of the Quotation from Saraha’s Dohākośagīti.” Journal of Indian Philosophy 43: 511–57. Hahn, Michael. 2016. Vṛttamālāstuti of Jñānaśrāmitra with Śākyarakṣita’s Vṛttamālā(stuti)vivṛti. New Delhi: Aditya Prakashan. Isaacson, Harunaga. 2013. “Yogācāra and Vajrayāna according to Ratnākaraśānti.” In The Foundation for Yoga Practitioners: The Buddhist Yogācāra bhūmi Treatise and Its Adaptation in India, East Asia, and Tibet, edited by Ulrich Timme Kragh, 1036–51. Cambridge, MA: Harvard University Press. Keira, Ryusei. 2004. Mādhyamika and Epistemology: A Study of Kamalaśīla’s Method for Proving the Voidness of All Dharmas: Introduction, Annotated Translations, and Tibetan Texts of Selected Sections of the Second Chapter of the Madhyamakāloka. Wien: Arbeitskreis für tibetische und buddhistische Studien, Universität Wien. MacDonald, Anne Elizabeth. 2009. “Knowing Nothing: Candrakīrti and Yogic Perception.” In Yogic Perception, Meditation and Altered States of Consciousness, edited by Eli Franco and Dagmar Eigner, 133–69. McCrea, Lawrence J., and Parimal G. Patil. 2010. Buddhist Philosophy of Language in India: Jñānaśrīmitra on Exclusion. New York: Columbia University Press. Miyazaki, Izumi. 2007a. “Atiśa (Dīpaṃkaraśrījñāna): His Philosophy, Practice and Its Sources.” Memoirs of the Research Department of the Toyo Bunko 65: 61–89. ———. 2007b. “Annotated Tibetan Text and Japanese Translation of the Ratnakarandodghaṭa-nāmamadhyamakopadeśa of Atiśa.” Memoirs of the Faculty of Letters, Kyoto University 46: 1–126. https://repository.kulib.kyoto-u.ac.jp/dspace/handle/2433/73130?mode=full. Patil, Parimal G. 2009. Against a Hindu God: Buddhist Philosophy of Religion in India. New York: Columbia University Press. Ruegg, David Seyfort. 1981. The Literature of the Madhyamaka School of Philosophy in India. History of Indian Literature series. Wiesbaden: Harrassowitz. ———. 2000. Studies in Indian and Tibetan Madhyamaka thought. Wien: Arbeitskreis für tibetische und buddhistische Studien, Universität Wien. ———. 2002. Two Prolegomena to Madhyamaka Philosophy: Candrakīrti’s Prasannāpāda madhyamakavṛttiḥ on Madhyamakakārikā I.1 and Tsoṅ kha pa blo bzaṅ grags pa/rgyal tshab dar ma rin chen’s dka’ gnad/gnas brgyad kyi zin bris. Studies in Indian and Tibetan Madhyamaka Thought. Wien: Arbeitskreis für tibetische und buddhistische Studien, Universität Wien. ———. 2010. The Buddhist Philosophy of the Middle: Essays on Indian and Tibetan Madhyamaka. Boston, MA: Wisdom Publications. Sanderson, Alexis. 2009. “The Śaiva Age.” In Genesis and Development of Tantrism, edited by Shingo Einoo, 41–349. Tokyo: University of Tokyo. Sherburne, Richard. 1983. A Lamp for the Path and Commentary of Atīśa. The Wisdom of Tibet Series 5. London: Allen & Unwin. ———. 2000. The Complete Works of Atīśa Śrī Dīpaṁkara Jñāna: Jo-Bo-Rje: The Lamp for the Path and Commentary, Together with the Newly Translated Twenty-Five Key Texts (Tibetan and English Texts). New Delhi: Aditya Prakashan. Wangchuk, Dorji. 2009. “A Relativity Theory of the Purity and Validity of Perception in Indo-Tibetan Buddhism.” In Yogic Perception, Meditation, and Altered States of Consciousness, edited by Eli Franco and Dagmar Eigner, 215–37. Wien: Verlag der Österreichischen Akademie der Wissenschaften.

614

38 ABHAYĀKARAGUPTA A Last Great Pandita Kazuo Kano

Life Abhayākaragupta (twelfth century CE) was one of the most influential scholars of the great monastic university Vikramaśīla.1 Born, according to a Tibetan historical account, into a Brāhmaṇa family in eastern India, in his youth, he received the instruction from a young yoginī to go to Magadha, where he became a Buddhist. Returning to Bengal, he took novice ordination and studied non-tantric teachings. From there he moved on to tantric teachings, which he studied under Ratnākaragupta, whose name is embedded in the opening verse of his Instruction on the Procedure for Self-Empowerment (Svādhiṣṭhānakramopadeśa). Eventually he established a sufficiently high reputation to gain the support of a Pāla king, very likely Rāmapāla. According to the Tibetan author Tāranātha, Abhayākaragupta served as a mkhan po2 of both Vikramaśīla and Nālandā monasteries. Many Tibetan scholars visited Abhayākaragupta in Vikramaśīla and Nālandā in order to study under him, and he supported their translations of a large number of Sanskrit Buddhist treatises into Tibetan.3 In Tibet he later came to be regarded as, as Seyfort Ruegg (1981, 115) states, “a representative of Śāntarakṣita’s Yogācāra-Svātantrika-Madhyamaka school.” In the following, we shall see that he replicates large sections of Kamalaśīla’s Illuminator of the Middle (Madhyamakāloka), thus lending weight to this doxographical attribution.

Writings Some twenty-six works are currently attributed to Abhayākaragupta (see Mori 2009, 2–3; Sinclair 2011; Luo 2020). Among them, twenty-five are preserved in Tibetan, and thirteen may be found in the Sanskrit original. They include texts which are ascribed to Abhayākaragupta but whose authorship is yet to be confirmed (e.g., no. 17 subsequently). Four of the works are classified as non-tantric and the other twenty-two as tantric works. His non-tantric works are the following: 1. 2. 3. 4.

Adornment of the Sage’s Thought (Munimatālaṃkāra) (Skt., Tib.) Flower Cluster of the Middle Way (Madhyamakamañjarī) (Skt.) Moonlight on Crucial Points (Marmakaumudī) (Tib.) Manual for Taking the Bodhisattva Vow (Bodhisattvasaṃvaragrahaṇavidhi) (Tib.)

DOI: 10.4324/9781351030908-53

615

Kazuo Kano

No. 1 (completed in 1108/1109) is his main non-tantric work. Encyclopedic in scope, it is one of the most extensive sources for determining his philosophical position, a topic to which we shall return in the following. No. 2 expounds upon Madhyamaka philosophy, seeking to establish the absence of an intrinsic nature among all phenomena and refuting other philosophical positions. This text is preserved only in Sanskrit, and an edition is currently being prepared by Hong Luo. No. 3 is an extensive commentary on the Adornment of Realization (Abhisamayālaṃkāra) on the basis of the Perfection of Wisdom in Eight Thousand Lines (Aṣṭasāhasrikā Prajñāpāramitā). No. 4 is a ritual manual for taking the bodhisattva vow and contains passages parallel to the Adornment of the Sage’s Thought. Among his 22 tantric works, the following three are commentarial literature:   5. Flower Cluster of Transmissions (Āmnāyamañjarī) (Skt., Tib.)  6. Guide by the Fearless One (Abhayapaddhati) (Skt., Tib.)  7. Moonlight on the Five Procedures (Kramakaumudī) (Skt., Tib.) No. 5 (completed in 1115/1116) is his largest work. Styled as a commentary on the Tantra of Emergence from Sampuṭa (Saṃpuṭodbhavatantra), it is in fact an encyclopedic work covering a wide variety of topics over the entire field of tantric Buddhism, including a number of philosophical subjects, such as a discussion of differences between non-tantric and tantric Mahāyāna (see Tomabechi and Kano 2008). No. 6 is his commentary on the Tantra of the Skull of the Buddha (Buddhakapālatantra) and No. 7 on the Five Procedures (Pañcakrama), which is a tantric meditation manual stemming from the Guhyasamājatantra (an edition of which is currently being prepared by Toru Tomabechi). The following three, called collectively the “Garland Trilogyˮ ( phreng ba skor gsum), are systematic compendiums of ritual or meditation manuals dealing with various tantric traditions of Indian Buddhism:   8. Garland of Vajras (Vajrāvalī) (Skt., Tib.)   9. Garland of Completion Yoga (Niṣpannayogāvalī) (Skt., Tib.) 10. Flower Cluster of Fire Oblation (Jyotirmañjarī) (Skt., Tib.) No. 8 is the key part of the trilogy, while the other two (Nos. 9 and 10) are called, by Abhayākaragupta himself, “supplements” ( parikara) to it. No. 8 treats some fifty ritual manuals, focusing on describing “maṇḍalas to be drawnˮ (lekhyamaṇḍala) for the purpose of consecrating ( pratiṣṭhā and abhiṣeka) either objects of worship or people. The Garland of Vajras had a great impact on Tibetan tantric tradition, including serving as the basis for the production of a set of scroll paintings (thang ka) depicting maṇḍalas at Ngor monastery, inspiring Tsong kha pa’s Great Gradual Paths of Esoteric Teachings (sNgags rim), and so on. No. 9 details twenty-six particular kinds of “maṇḍalas for meditationˮ (bhāvyamaṇḍala), describing visualizations of each deity within them. No. 10 is a tantric manual for the fire-offering ritual (homa, see Mori 2011). There are two works each on the Cakrasaṃvara cycle and Kālacakra cycle: 11. 12. 13. 14.

A Guide to the Realization of Śrīcakrasaṃvara (Śrīsaṃvarābhisamayopāyikā) (Skt., Tib.) Instruction on the Procedure for Self-Empowerment (Svādhiṣṭhānakramopadeśa) (Skt., Tib.) Key Points of the Kālacakra (Kālacakroddāna) (Tib.) An Entrance Gate to the Kālacakra (Kālacakrāvatāra) (Skt., Tib.)

No. 11 is a meditation manual based on Lūyīpāda’s Realization of Cakrasaṃvara (Cakrasaṃvarābhisamaya) dealing with the 62-deity maṇḍala of Cakrasaṃvara (the Sanskrit original of which was recently found by Péter-Dániel Szántó), while No. 12 deals with the visualization 616

Abhayākaragupta

leading to one’s self-transformation into the two-armed Cakrasaṃvara. Nos. 13 and 14 are short works on the Kālacakratantra, with parts of No.13 having to do with astrology (gza’ skar). The following four are miscellaneous tantric works: 15. 16. 17. 18.

Flower Cluster of Instruction (Upadeśamañjarī) (Tib.) Guide to Awakening (Bodhipaddhati) (Tib.) Short Treatise on Consecration (Abhiṣekaprakaraṇa) (Tib.) Flower Cluster of Transgressions within the Vajrayāna (Vajrayānāpattimañjarī) (Tib.)

No. 15 argues several different positions regarding tantric realizations of the generation stage (utpattikrama) and the stage of the arisen (utpannakrama) including those of Ratnākaraśānti (alluded to by the phrase “a certain master [ācārya] of Pūrvadeśaˮ), Nāgabodhi, Ānandagarbha, Jñānapāda, Indrabodhi, and Kambalāmbara. It sometimes discusses philosophical topics, such as an argument on the ontological status of visualized maṇḍala-deities, the applicability of tantric visualization to the Three Nature theory of Yogācāra, and different positions on the classification of the bodies of a buddha. Its opening verse may contain an allusion to Abhayākaragupta (mi ’jigs mgos phyag ’tshal), and the work contains a number of references to his other works (e.g., Garland of Vajras). These facts suggest that this is a work by Abhayākaragupta, though the attribution is not certain. No. 16 deals with a series of worshipful mantra recitations. No. 17 is a treatise on consecration rituals (abhiṣeka). No. 18 is a brief list of fourteen kinds of fundamental transgressions (mūlāpatti) and eight kinds of coarse but less grave transgressions (sthūlāpatti) in tantric practice. Finally, there are eight further minor works of tantric ritual manuals focusing on specific tantric deities or themes: 19. 20. 21. 22. 23. 24. 25. 26.

Accumulation of Rituals of Mahākāla (Mahākālakarmasaṃbhāra) (Tib.) Ritual Procedure for Rejoicing the Lord Kākayoni (Nāthakākayonitarpaṇavidhikrama) (Tib.) Perfection of Raktayamāntaka (Raktayamāntakaniṣpanna) (Tib.) Meditation Guide of Jñānaḍākinī (Jñānaḍākinīsādhana) (Tib.) Meditation Guide of Ucchuṣmajambhala (Ucchuṣmajambhalasādhana) (Tib.) Fire Oblation of Vajramahākāla for Harming Enemies (Vajramahākālābhicārahoma) (Tib.) Ritual Manual for Rain (Varṣāpaṇavidhi) (Skt.) Ritual Manual for the Tantric Feast (Gaṇacakravidhi) (Tib.)

These all are ritual or meditation manuals dealing with the worship of various deities or for the fulfillment of mundane wishes.

Writing Strategy One of Abhayākaragupta’s concerns appears to have been to avoid establishing a new tradition of his own, and a key strategy towards this end was to fill his text with a patchwork of passages written by other authors. He frequently draws upon such passages without mentioning the titles or authors of these sources. In this regard, Abhayākaragupta may seem to be more a compiler than an author, although such a judgment risks also overlooking his contribution. In any case, he usually reproduces the original text almost verbatim, not drastically changing the original wording or word order, and he only exceptionally omits words. Thanks to this mosaiclike patterning, we can recover a number of passages from rare Buddhist works in the original language when those Sanskrit texts have yet to be found. 617

Kazuo Kano

For instance, in his Adornment of the Sage’s Thought, passages from Kamalaśīla’s Illuminator of the Middle (Madhyamakāloka) and Candrakīrti’s Treatise of the Five Aggregates (Pañcaskandhaka), both of which are so far available only in Tibetan translations, have been recovered (see Kano and Li 2020). In his Flower Cluster of Transmissions, we likewise find passages from works of Ratnākaraśānti, Kamalanātha, Bhavabhaṭṭa, Nāgabodhi, and so forth (Isaacson and Sferra 2019, 251). These discoveries alone make his works valuable. But they are also interesting in terms of their presentation of doctrine.

Locating Abhayākaragupta’s Doctrinal Position Abhayākaragupta leans philosophically towards Madhyamaka thinkers, especially Kamalaśīla, but also Nāgārjuna, Śāntarakṣita, and Candrakīrti, among others. With regard to his understanding of the Adornment of Realization tradition, he mainly follows Āryavimuktisena, while when dealing with tantric issues, he frequently reuses passages from Nāgabodhi, Jñānapāda, Ratnākaraśānti, Kamalanātha, and others, including his master Ratnākaragupta (Isaacson and Sferra 2019, 251). His own impact is seen, for instance, in works by Daśabalaśrīmitra and Ratnarakṣita and also within the Tibetan traditions. Many of his works are yet to be studied, and future research can be expected to clarify his doctrinal position more precisely. The works that center on philosophical topics are the Adornment of the Sage’s Thought, Moonlight on Crucial Points, Flower Cluster of the Middle Way, and a number of tantric treatises, such as the Flower Cluster of Transmissions and Flower Cluster of Instruction. The Adornment of the Sage’s Thought is the most comprehensive and systematic among his non-tantric works. One might call it Abhayākaragupta’s encyclopedic overview of the entire system of non-tantric Buddhist doctrines and practices.4 Abhayākaragupta himself states his purpose in composing the work in the opening verses: The highest thought of the great sages (i.e., buddhas), namely, the own nature of this body (mūrti, i.e., dharmakāya), was revealed by the excellent noble ones (i.e., Nāgārjuna, Maitreya, etc.). But worldly people, who are confounded by darkness, do not understand this thought – therefore, the succession ( paraṃparā) of light shed on this thought (i.e., in the four chapters of this work). According to this third verse of the opening, Abhayākaragupta composed this work in order to guide ignorant persons who do not comprehend the buddhas’ thought (i.e., the “sage’s intent” or munimata) as expounded by the noble ones Nāgārjuna, Maitreya, Āryavimuktisena, and others. He defines this work as: The light from (i.e., the chapters of) this Adornment of the Sage’s Thought, which consists of the jewel of qualities, shows the indescribably correct view. The word “lightˮ or “illuminatorˮ or “elucidationˮ (āloka) in this and previous verses indicates the four chapters of the work, each of whose titles ends with the word āloka. The chapters are: Chapter 1: The Light of the Resolve to Become a Buddha (bodhicittāloka) Chapter 2: The Light of Cultivating the Resolve to Become a Buddha (bodhicittabhāvanāloka) Chapter 3: The Light of the Eight Realizations (ạṣṭābhisamayāloka) Chapter 4: The Light of the Resultant Qualities (guṇāloka) 618

Abhayākaragupta

With this as his framework, Abhayākaragupta undertakes a comprehensive, systematic treatment of Mahāyāna Buddhism, elucidating its doctrinal foundations in chapter 1, its soteriological path in chapters 2 and 3, and the fruits of its path in chapter 4. Chapter 1 delves in depth into the resolve to become a buddha (bodhicitta) by describing significant related doctrinal and practice-oriented topics of non-tantric Buddhism: the significance of compassion as a starting point of Buddhist practices; the ritual involved in taking a vow to become a Buddhist; the Six Perfections, especially the perfection of moral conduct; the cosmologies of sentient beings and the inanimate world; the Abhidharmic classifications of all phenomena into the five aggregates ( pañcaskandha), and so on; the four truths of the noble ones (caturāryasatya); the three natures of Yogācāra (trisvabhāva); the two truths (satyadvaya); and the single vehicle (ekayāna). Chapter 2 elucidates the cultivation of the twenty-two kinds of the resolve to become a buddha that are taught in the Adornment of Realization (following Jñānakīrti’s Commentary on the Entrance to the Reality (Tattvārthāvatāravyākhyā) [D 3709]; see Isoda 1998). Chapter 3 discusses the eight kinds of realization at the core of the Adornment of Realization. Chapter 4 deals with qualities resulting from the cultivation of the resolve to become a buddha. Of the four chapters, the first provides the richest insight into Abhayākaragupta’s central philosophical leanings, the two truths being by his own confession one of the most crucial doctrinal points for him (see subsequently).

Selected Topics Pertinent to Abhayākaragupta’s Philosophical Position In the following, we shall review selected topics of Abhayākaragupta’s philosophical position, mostly focusing on the Adornment of the Sage’s Thought. These topics shall be: (1) (2) (3) (4) (5)

The scriptural authority of Mahāyāna texts The two truths The two truths and an evaluation of the mind-only doctrine The illusory but beneficial mind Buddha-nature

The Scriptural Authority of Mahāyāna Texts In the middle of the last chapter of the Adornment of the Sage’s Thought, Abhayākaragupta inserts a discussion of an extra argument with the expression “now, enough of excessive argumentation.” This passage (Skt. Ms. fols. 148r1–150v2) appears in the context of the eighth of the ten great aspirations (mahāpraṇidhāna), that is, the “application of one’s single will to enter the Great Vehicle on account of not being desirous of other Vehicles,” one of the qualities in the first stage of a bodhisattva. To describe these qualities, he borrows, verbatim, sentences from Vasubandhu’s Commentary on the Sūtra of the Ten Stages (Kano 2019). The passage in question, which is probably in Abhayākaragupta’s own words, is inserted in between a series of these borrowed sentences. There, he refutes the claim to scriptural authority on the part of the rival Vehicle he designates as the Śrāvakayāna (Vehicle of the Disciples) and establishes the Mahāyāna’s superiority to this Vehicle. A summary of the argument and extractions of key points follows (note that the following is not a translation but rather a synopsis): Abhayākaragupta maintains: The Mahāyāna is great because of seven points of greatness,5 but still there are those who do not believe in the Mahāyāna. How can the Mahāyāna not be the Buddha’s words? 619

Kazuo Kano

An opponent replies: This is because the Mahāyāna is external to the traditional Three Baskets, i.e., Sūtra, Vinaya, and Abhidharma, and it does not meet the definition of the Buddha’s words: “What fits into the Sūtra Piṭaka, what is shown in the Vinaya Piṭaka, and what does not go against the reality (dharmatā) – that is the Buddha’s word.” Abhayākaragupta then insists that the Mahāyāna does meet this definition. The opponent retorts: The Mahāyāna is not what is proclaimed as the Buddha’s words in the fixed Three Baskets of the eighteen schools.6 Abhayākaragupta’s response: If the Three Baskets of the eighteen schools are the Buddha’s words, why do they contain texts that mutually contradict one another? Such texts cannot be the Buddha’s words. Furthermore, they cannot even be the words of past arhats, for arhats, who have abandoned the defilements, do not compose such mutually contradictory texts. They must have been composed by ordinary beings who have not yet abandoned their remaining traces of ignorance and doubt. Therefore, the sacred scriptures of the Mahāyāna cannot be included in such recent texts made by ordinary people. Even assuming that there were ancient Śrāvakayāna scriptures that had been lost (i.e., even though someone insists that the teaching of Mahāyāna scriptures correspond to lost Śrāvakayāna scriptures), Mahāyāna scriptures could not be included among them, for there is a huge gap between the Śrāvakayāna and the Mahāyāna. If the Mahāyāna could fit into or be included within the Hīnayāna (i.e., Śrāvakayāna), the very essence of the Mahāyāna would be lost. It is like the great ocean, which does not lose its quality of being a receptacle (āśraya) of all kinds of water, even though it cannot fit into a small hole (i.e., it does not matter that the Mahāyāna scriptures are not included in the Śrāvakayāna scriptures, and this is not a fault of the Mahāyāna, but rather a virtue). Abhayākaragupta goes on to ask how the bright, variegated Mahāyāna teachings can be useful for the blind followers of the Śrāvakayāna, who are the deserved objects of compassion? (See Kano forthcoming) So much for a summary of the passage. In the history of Indian Buddhism, the scriptural authority of the Mahāyāna has been discussed by several authors, including Nāgārjuna, Vasubandhu, Īśvara, Bhāviveka, Haribhadra, and Prajñākaramati.7 Abhayākaragupta’s argument fundamentally follows the pattern already seen in the works of these previous authors, who pointed out contradictory statements in the Śrāvakayāna scriptures. Abhayākaragupta himself seems to know these scriptures well, quoting a number of them from the different schools elsewhere in this work. On the other hand, the following two points do not seem to be found in similar discussions by previous authors and thus probably represent his own position: (1) the Śrāvakayāna scriptures are not a depository of the Buddha’s words and not even one of past arhats’ words; rather, they contain merely the words of ordinary beings, and (2) the Mahāyāna scriptures do not need to be included in the Śrāvakayāna scriptures, given the superiority of the former to the latter. This rather harsh attitude towards the followers of the Śrāvakayāna likely reflects the fact that Śrāvakayāna monks were still active and influential at the time of Abhayākaragupta.

The Two Truths In one of his closing verses of the Adornment of the Sage’s Thought, Abhayākaragupta evaluates the two truths as being the most crucial doctrine: Here (i.e., in this work), in sum, the life (i.e., crucial point) of the pair of truths (i.e., the two truths) has been presented by me from the perspective of the three methods (i.e., the 620

Abhayākaragupta

śrāvaka, pāramitā, and mantra approaches). If there is misunderstanding, please forgive me for it. As for the two truths, those who are enjoying the bhūmis are the authority.8 “Those who are enjoying the bhūmisˮ mean bodhisattvas practicing the ten stages. In the following two passages from chapter 1, he explains that the Two Truths encompass all other Buddhist doctrinal points. In his introductory statement of this Two Truths passage, he states: These very same five aggregates, twelve āyatanas, and eighteen elements are nothing but the four truths, that is, the truths of suffering, origination, cessation, and the path; the three natures, that is, the natures of conceptualization, dependence, and perfection; the two truths, that is, the conventional truth and the ultimate truth; and finally the single, nondual ultimate truth, that which is again beyond all proliferation; that which is precisely the Mahāyāna, Perfection of Wisdom ( prajñāpāramitā), and the ultimate resolve to become a buddha (bodhicitta).9 In the closing part of this same passage, Abhayākaragupta quotes the Sūtra on the Descent into Laṅkā (Laṅkāvatārasūtra) 6.5 in order to re-emphasize that these doctrinal topics encompass the whole of the Mahāyāna teaching and that all of them are in turn included in the two truths: On the other hand, a summary division of them (i.e., reality and cognition of it) was taught in the Ārya-Laṅkāvatārasūtra: “The Five dharmas, the Three Natures, the Eight Cognitions, the Two Selflessnesses – these ought to encompass the whole of the Mahāyāna.ˮ Here, the Five dharmas are name, manifestation, conceptualization, reality, and the true gnosis. Among these, name, manifestation, and conceptualization are mentioned when specifying the conventional truth, and reality and the true gnosis are mentioned, both literally and figuratively, when specifying the ultimate truth. The Three Natures are those of conceptualization, dependence, and perfection. The Eight Cognitions are the storehouse cognition (ālayavijñāna), the defiled mind, and the cognitions of the six sense faculties, i.e., those of eyes and the like. The Two Selflessnesses are the selflessness of persons and the selflessness of dharmas. Among these the conceptual nature, dependent nature, and the Eight Cognitions are mentioned when specifying the conventional truth. The perfect nature and the Two Selflessnesses are mentioned when indicating the ultimate truth.10 In sum, Abhayākaragupta subsumes the major doctrinal points under the two truths in the following manner: Instruction of the Conventional Truth

Instruction of the Ultimate Truth

Five Dharmas

name, manifestation, and conceptualization

reality and the true gnosis

Three Natures

the nature of conceptualization and the nature of dependence

the nature of perfection

Eight Cognitions

all eight cognitions

—

Two Selflessnesses

—

the two Selflessnesses

Doctrinal Topics

621

Kazuo Kano

As an adherent of Madhyamaka, he naturally takes all Eight Cognitions (including the ālayavijñāna) as falling under the conventional truth. The Eight Cognitions are a part of Yogācāra theory. They are ālayavijñāna (storehouse consciousness; the substratum of the world, which is nothing but the mind), kliṣṭaṃ manas (defiled mind; the mind which grasps ālayavijñāna as “Iˮ), and six kinds of pravṛttivijñāna (forthcoming form of the mind, such as eye-cognition).

The Two Truths and the Mind-Only Doctrine Abhayākaragupta continues his discussion with an evaluation of the mind-only doctrine (the mind-only doctrine is a part of Yogācāra doctrine that teaches all phenomena as nothing but a product of the mind).11 The central question is whether the mind is to be accepted as merely conventionally existent, and he comes down on the side of that being the case. His exposition of the mind-only doctrine is found in the same section of the Adornment of the Sage’s Thought that deals with the two truths, most of it sticking closely to Kamalaśīla’s train of thought in the Illuminator of the Middle (along with very frequent unacknowledged passages from Kamalaśīla). The relevant portion of text can be summarized as follows: Abhayākaragupta views the mindonly doctrine as the entrance gate to understanding that all phenomena are characterized by their absence of an intrinsic nature, whence he goes on to explain that externally perceived objects are only conventionally existent, to disallow any real existence to the mind and external objects, to deny the existence of a mind that has no mental image, to establish the usefulness of the mind-only doctrine for attracting ordinary beings, and to refer to a position (most likely that of Śāntarakṣita) that accepts the conventional existence of the mind but refuses to accord it to external objects. He concludes that the adherents of the mind-only doctrine are realists (vastuvādin) who teach a pseudo ( prativarṇikā) version of the Perfection of Wisdom ( prajñāpāramitā).12 Although he introduces a position that can be identified with that of Śāntarakṣita, who accepts the conventional existence of the mind but denies even that to external objects, Abhayākaragupta leaves open his final judgement on the matter, saying: Thus it does not matter whether external objects, too, exist conventionally, or just the state of mind-only; there is absolutely no Buddhist teaching at all that assuredly teaches any real intrinsic nature of phenomena.13 For him, the truth of any final judgement on the ontological status of conventional forms of existences is less crucial than the pragmatic effect, in other words, the benefit, it has. What is more important for him, that is, is how the Mind-only doctrine works for the sake of those exposed to it. In line with Kamalaśīla’s Illuminator of the Middle (D 3887, 159r6–7), he states: And the mind-only doctrine is not just talk for deluding other people, in as much as the Buddha’s instruction in general brings forth fruitful results, by bringing benefit and bliss to those undergoing training. Also, there is no delusiveness in the Fortunate Ones (i.e., buddhas), since only something that benefits others is the truth. And it was taught by the master Nāgārjuna: “The truth is something that benefits others, whereas delusion is the opposite, for it does not bring forth any benefit to others.ˮ14 Although Abhayākaragupta takes the mind-only doctrine as a provisional teaching, he insists on the significance of this doctrine’s efficacy in benefiting others, and thus, for that very

622

Abhayākaragupta

reason, considers it a truth. He continues to solidify this argument in a context that deals with the beneficial aspect of the mind, again following Kamalaśīla, as laid out in the next section.

The Illusory but Beneficial Mind Let us overview the series of arguments concerning the beneficial aspects of the illusory mind, which follow the passage mentioned just previously. The passages in question are again, in large part, silent textual borrowings from Kamalaśīla’s Illuminator of the Middle.15 An opponent first criticizes Abhayākaragupta’s statement that the selflessness of all phenomena is directly perceived through yogic perception ( yogipratyakṣa), in as much as selflessness, a state of nonexistence, should be imperceptible; that is, nonexistence cannot give rise to any perception (cf. Kano and Li 2021, 9; Ms. fol. 64r2–3). In reply, Abhayākaragupta answers that even nonexistence (for instance, a rabbit’s horn) can generate a perception and that yogins (i.e., perceivers of reality) do not see the world as really existent, but they do see its lack of a self. He then quotes the definition of direct perception from Dharmakīrti’s Commentary on the Means of Knowledge (Pramāṇavārttika): “Direct perception is free from conceptualization and from delusion” (kalpanāpoḍham abhrāntaṃ pratyakṣam). And he states that “even if there is no object, a perception does arise very vividly through the power of cultivation” (cf. Kano and Li 2021, 11; Ms. fol. 64v4) and that this perception is free from conceptualization. He thus refutes the opponent’s objection. He then explains the three kinds of direct perception (namely those of ordinary beings, yogins, and buddhas), which are free from delusion (abhrānta) with regard, respectively, to verbal expressions (vyavahārataḥ), the selflessness of persons, and every aspect of reality (cf. Kano and Li 2021, 12–13, Ms. fol. 65r1–4). With regard to the illusory nature of all phenomena, he states that “an ultimate nature or the like superimposed by ordinary people is invalidated by means of inference (or perception). Its being invalidated establishes that all phenomena are ultimately devoid of an intrinsic natureˮ (65r4). Scriptures are quoted in support of this (cf. fol. 65r4–v2). Regarding inference, he mentions two proof statements ( prayoga) and refers to the Flower Cluster of the Middle Way for further references to the relevant arguments: And regarding both the reason “being dependently arisen on the part of all phenomenaˮ ( pratītyasamutpannatva) and the reason “being devoid of the nature of oneness and manifoldness on the part of all phenomenaˮ (ekānekasvabhāvarahitatva), there are namely, (proof statement 1) “What is dependently arisen is, from the viewpoint of ultimate reality, empty by nature, just as an illusion is. And all dharmas are dependently arisen”; and (proof statement 2) “What lacks the intrinsic nature of oneness and manifoldness lacks, from the viewpoint of ultimate reality, an intrinsic nature, just as a reflection does. And all dharmas lack, from the viewpoint of ultimate reality, an intrinsic nature of both oneness and manifoldness”. The demonstration of these two proof statements is not laid out here, for I fear that those who do not know logical procedures will not listen to it, given their lack of understanding. However, it was explained by me in a summarized manner – unsurpassably so – in the Flower Cluster of the Middle Way. Therefore, the wise may reflect on it there.16 The opponent points out an undesirable consequence: If the real existence of all phenomena is refuted by inferring their selflessness, then what about direct perception, which should be

623

Kazuo Kano

the foundation for this very inference? That is, given that the basis of the inference is to be abandoned, the inference, too, should be abandoned (cf. Ms. fol. 65v3–4). Abhayākaragupta accepts this consequence but judges it no problem, because he who has the ‘fire’ of analytic gnosis (i.e., inference) burns the ‘fuel,’ that is, all these illusory phenomena, and then, having doused the ‘fire’ of analysis, realizes the excellent emancipation (nirvṛti) which has the cessation of conceptualization as its characteristic (cf. Kano and Li 2021, 15–16; Ms. fol. 65v4–66r1). The opponent then points out a further undesirable consequence: Given that even the mind is devoid of an intrinsic nature on the ultimate level, the gnosis which realizes this very absence would have to be abandoned (as being merely an illusion), and so there would be no gnosis among yogins and no gnosis among buddhas, and finally, “there would be the destruction of the entire system” of Buddhist teaching. Abhayākaragupta replies by pointing out that although the mind does not exist and is illusory on the ultimate level, it should not be abandoned. If it were both illusory and not beneficial, only then should it be abandoned. But the mind is illusory but indeed has the qualities that benefit oneself and others and so should not be abandoned. It is like a buddha’s emanation body (nirmāṇakāya), which is illusory on the ultimate level but beneficial for everyone. In order to nail down this point, he states: It is indeed not the case that existence alone is reason for accepting the mind, nor is it the case that nonexistence is reason for abandoning it. In this regard, he provides examples of the stances of śrāvakas, yogins, and buddhas stances regarding the acceptance of nonexistence: Śrāvakas: That is to say, although Śrāvakas, in embracing the idea of truly existing entities, abandon the first two truths (i.e., the truth of suffering and that of origination), they accept the third (i.e., the truth of cessation), which is not an entity. Yogins: And yogins (i.e., bodhisattvas), whose minds are filled with compassion, create various emanations – elephants and the like – for the sake of beneficial activities for other beings even though the latter are illusory. Buddhas: Likewise, it is precisely admitted, though they have already abandoned all error, that the Fortunate Ones – who have mastered the means of bringing forth infinite benefit and bliss for others by depending on great compassion, who have totally dispelled the darkness of delusion by means of the most eminent light of wisdom, and who wish benefit for others – will remain until the end of sentient beings, of their karma, of their defilements, and of space, by virtue of the power of aspiration in their past lives.17 Abhayākaragupta then adds the following comment, in which he makes reference to the pantheon of tantric buddhas such as Saṃvara and others, who continue to be active in the world. Even though in Jambudvīpa there are no disciples of the Buddha in Śākyamuni’s form (i.e., because he had gone long before), the Fortunate One remains even today in the form of Saṃvara and so on, manifests to bodhisattvas who have accumulated wholesome accumulations over many kalpas, and who are capable of teaching the profound new method (i.e., the tantric path), and he still teaches them his teachings.18 624

Abhayākaragupta

Arguing the possibility of saving infinite beings, he continues: Objection: Saving all beings is impossible because they are limitless. Answer: Precisely for this reason, buddhas’ compassionate thoughts toward them are limitless. Otherwise their activity would not accord with their affirmation (i.e., their commitment to save all beings), for they promise release from suffering for all beings at the very moment when their resolve to attain enlightenment first arises.19 He then argues that bodhisattvas’ gnosis, which is illusory but beneficial, stems from their accumulations of merit and wisdom. Likewise, the non-abiding form of nirvāṇa (apratiṣṭhānanirvāṇa) stems from a buddha’s compassion and wisdom. Given their compassion, buddhas do not completely enter nirvāṇa, and given their wisdom, they are not defiled in saṃsāra (cf. Ms. fol. 66v3–67r1). He goes on to explain that bodhisattvas bring forth great fruit because – for instance, in the case of the perfection of offering (dānapāramitā) – they understand the absence of any intrinsic nature within offered materials, such understanding removing any possible attachment to them and being supported by the realization of emptiness in a meditative state (cf. Ms. fol. 67r1–v2). From there he moves on, in the last portion of chapter 1, to discuss establishing the truth of the single vehicle. Thus a summary of the text portion, which represents one of the highlights of this work’s philosophical discourse. As with other sections of the work, in large part, Abhayākaragupta arranges and contextualizes Kamalaśīla’s passages so as to fit into his own flow of thought and occasionally interjects statements of his own into them. Among his own foci are topics that may motivate readers to take up the practice, such as the beneficial aspects of the mind, rather than proofs of selflessness or disproving wrong views. These more polemical tasks are, according to him, the concerns of another work, the Flower Cluster of the Middle Way. His choice of passages serves the specific purposes of the Adornment of the Sage’s Thought: chapter 1 deals with the doctrinal foundation for the series of cultivation practices dealt with from chapter 2 onward. He homes in on passages from Kamalaśīla that motivate readers to undertake such practices.

Buddha-Nature Another doctrine which propels one towards practice is that of Buddha-nature. Here I shall just briefly summarize Abhayākaragupta’s understanding of Buddha-nature (for the details, see Kano 2016, 108–23). To show that all sentient beings, whatever vehicle they may initially adopt, can attain the Mahāyāna awakening in line with the single-vehicle (ekayāna) theory, Abhayākaragupta states in chapter 1 of Adornment of the Sage’s Thought that the Buddha taught that “every sentient being has Buddha-nature (tathāgatagarbha)”; that is, all sentient beings are able to reach the state of perfect awakening. The term tathāgata, the first member of the compound tathāgatagarbha, refers to the sphere/element of reality (dharmadhātu), namely the natural luminosity which pervades all sentient beings in the form of the twofold selflessness (cf. Ms. fol. 69r4–5). Abhayākaragupta here takes tathāgatagarbha in the sense of dharmadhātugarbha (a term also found in the Śrīmālāsūtra), and this dharmadhātu is characterized by the selflessness of persons and phenomena (  pudgaladharmanair­ātmyalakṣaṇasya dharmadhātoḥ). These words are strikingly similar to Kamalaśīla’s statement in his Madhyamakāloka, where again the single-vehicle theory is expounded (D 3887, 242b4–7). 625

Kazuo Kano

In chapter 3 of Adornment of the Sage’s Thought, Abhayākaragupta states (cf. Ms. fol. 85v1): “The dharmadhātu has precisely as its characteristic all dharmas’ absence of an own-being.” He then roughly synonymizes the dharmadhātu with tathāgatagarbha and gotra (Ms. fol. 85v2: idam eva dharmatātmakaṃ gotraṃ tathāgatagarbha ity uktaṃ). Further on, he poses a question and answers it: Question: Now, how can the dharmadhātu, which is without an own-being, be a base ( pratiṣṭhā) that gives rise to the resolve to become a buddha (cittotpāda)? Reply: This is just like space (ākāśa), which is without an own-being, being a base for the rays of the moon and sun, and for the removal of darkness as well. (Ms. fol. 86v1) Abhayākaragupta rehashes this passage from Ratnākaraśānti’s Quintessence (Sāratamā) (D 30a4–5), with the slight modification of arūpin (gzugs med pa) into niḥsvabhāva. He claims that just as space or the sky (ākāśa), which lacks an intrinsic nature (niḥsvabhāva), functions as a base for removing darkness, so too the dharmadhātu, with no nature of its own either, serves as a base for cittotpāda.20 Abhayākaragupta does not accept the dharmadhātu (or Buddha-nature) as a cause which engenders something ( jananahetu), since it has been categorized as unconditioned (asaṃskṛta), something which cannot produce anything. Instead, he takes it as a supporting cause ( pratiṣṭhāhetu). In this way, Buddha-nature can function as a cause, even if it lacks a nature of its own for engendering something. In the Moonlight on Crucial Points (D 3805, 49b4), in the commentary on the Adornment of Realization I.37–39, we find a similar idea, Buddha-nature (synonymized with prakṛtisthagotra and dharmadhātu, again characterized by the absence of an intrinsic nature within all phenomena) being explained as a support or cause of bodhicitta (Kano 2016, 120). In chapter 1 of his tantric work, the Flower Cluster of Transmissions, Abhayākaragupta states: “If the minds of these sentient beings were not luminous by nature, they, like charcoal, would not become immaculate even when the obscuration in them is washed out” (Skt. Ms. 19r3, Tib. D 7b4–5). He then quotes Mahāyānasūtrālaṃkāra IX.37 and concludes that the mind of every sentient being is luminous by nature, tathatā pervading everything without exception. A similar idea is found in his Pañcakramatātparyapañjikā Kramakaumudī on Pañcakrama 2.1. When explaining the word kulam, “buddha-family,ˮ in Emergence from Sampuṭa I-i, again in Flower Cluster of Transmissions chapter 1, Abhayākaragupta has occasion to repeat that every sentient being has Buddha-nature. That Buddha-nature lacks any intrinsic nature is especially suggestive of the doctrinal position of Mādhyamikas, which he espouses: Here, indeed, everything has the mind as its foremost element. And sentient beings characterized by the mind continuum have a tathāgata within/have Buddha-nature (tathāgatagarbhāḥ) – have as their nature a state of absence of any intrinsic nature (naiḥsvābhāvyātma-).21 This is a passage borrowed from Ratnākaraśānti’s Commentary on the Manual of MaṇḍalaPractice of the Guhyasamājatantra (Guhyasamājamaṇḍalavidhiṭīkā),22 where the phrase “state of absence of any intrinsic natureˮ is not found, for Ratnākaraśānti does accept the real existence of the mind in the form of “illuminationˮ ( prakāśa), which has an intrinsic nature. The bold-faced words (i.e., naiḥsvābhāvyātma-) were inserted by Abhayākaragupta in order to show that everything is just illusory in accordance with his Madhyamaka position. 626

Abhayākaragupta

In sum, Abhayākaragupta utilizes the Buddha-nature doctrine to establish the single-vehicle doctrine, which teaches that all sentient beings can and will attain Mahāyāna awakening. All along, he faithfully follows Kamalaśīla’s position. His interpretation of Buddha-nature as being devoid of any intrinsic nature (niḥsvabhāva) is also paralleled in Kamalaśīla’s Madhyamakāloka. His explanation of the dharmadhātu’s causal potency is drawn from Ratnākaraśānti’s Sāratamā. He sometimes borrows passages from Ratnākaraśānti’s works that deal with Buddha-nature and related topics and carefully modifies the borrowed passages to show Buddha-nature as lacking an intrinsic nature. Abhayākaragupta’s discussion of Buddha-nature had an impact on later Indian masters, including Daśabalaśrīmitra, who utilizes the discussion of the three states of tathatā; Ratnarakṣita, who further delves into the association of hetutantra with Buddha-nature; and Jayānanda, who probably makes use of the single-vehicle discussion in the Adornment of the Sage’s Thought. The role played by Abhayākaragupta, then, was to pick up on his predecessors’ discussions of Buddha-nature, to “Madhyamakanizeˮ them, and to pass them on so altered to later generations. His analysis of Buddha-nature accords with that of the mind (see previously), in which he stresses the beneficial aspect of the mind (i.e., for the sake of others), even though the mind itself is illusory from the viewpoint of the ultimate. Buddha-nature, too, is illusory (i.e., is devoid of an intrinsic nature) at the ultimate level, but it is beneficial for guiding others to the Buddhist path and for motivating them to undertake its practices (Kano 2016, 32–33).

Notes 1 Colophons of three of his works provide three dates within his lifetime, those of the Abhayapaddhati, Munimatālaṃkāra, and Āmnāyamañjarī mentioning the dates of completion as the twenty-fifth year (1103/1104), the thirtieth year (1108/1109), and the thirty-seventh year (1115/1116) of Rāmapāla’s enthronement, respectively. The dPag bsam ljon bzang of Sum pa mkhan po gives the dates of his birth and death as corresponding to 1064 and 1125. Regarding them, Isaacson and Sferra (2019, 249) state that “these dates need not be given particular authority, but they are consistent with those given in the three works mentioned previously for their completion.ˮ Isaacson and Sferra (2019, 249–51) and Luo (2020) offer the most recent overviews of Abhayākaragupta’s life and works. 2 The term mkhan po in this passage is frequently translated as “abbot.” An exact equivalent to it, however, is not attested in the India materials; that is, there is no evidence that Buddhist monasteries in India had what could be called mkhan pos. The only references to such personages occur in histories of Indian Buddhism authored by Tibetans. See Kano 2017. 3 Erb (1997, 28) estimates approximately 135 works. 4 The lone Sanskrit manuscript consists of 202 palm leaves; its Tibetan translation, Tōhoku no. 3903 in the Derge Tanjur, covers some 220 folia. Hirofumi Isoda has studied the Munimatālaṃkāra intensively on the basis of the Tibetan translation and has published editions of parts of it. Recently, Kazuo Kano and Xuezhu Li have started to edit the Sanskrit original and already have published several portions of the text. 5 See Mahāyānasūtrālaṃkārabhāṣya on chapter 19, verses 59–60. 6 See Haribhadra, Abhisamayālaṃkārālokā 401.25–402.24. 7 See Nāgārjuna’s Ratnāvalī, Mahāyānasūtrālaṃkāra, Vasubandhu’s Vyākhyāyukti, Īśvara’s Abhidharmadīpa, Bhāviveka’s Madhyamakahṛdaya and its autocommentary (Tarkajvālā), Haribhadra’s Abhisamayālaṃkārālokā, and Prajñākaramati’s Bodhicaryāvatārapañjikā. For details of these references, see Fujita 2011, 115–16. Abhayākaragupta extracts selected passages on the scriptures of the various schools from Bhāviveka’s Tarkajvālā (D 3856, 170a6–180a2; I owe this to Vincent Tournier) and a passage on the definition of the Buddha’s words from an unknown work called skyabs gsum gyi brtag pa attributed to Vasubandhu (cf. D 3811, 270b4–271a1) in the Adornment of the Sage’s Thought (Skt. Ms. fol. 8v2–9r4 and 148v5–r1, respectively). 8 The third verse in closing verses. 9 Skt. Ms. 58r5–v1, Tib. 138r1–2. 10 Kano and Li 2014, 20–21; Ms. 69v1–3; D 150b4–151a1.

627

Kazuo Kano 11 Some citations from the Adornment of the Sage’s Intention in sections “The Two Truths and the MindOnly Doctrineˮ and “The Illusory but Beneficial Mindˮ have also been translated from the Tibetan by Kapstein 2001. 12 These topics are treated in the Sanskrit manuscript, fol. 62v5–64r2. For the Sanskrit text and a Japanese translation, see Kano and Li 2018. 13 See Kano and Li 2018, 13; Ms. fol. 63v4. 14 See Kano and Li 2018, 42; Ms. fol. 63v2–3. 15 See Isoda 1993. The summary in this section is based on the Sanskrit manuscript fol. 64r2–67v2. 16 See Kano and Li 2021, 14–15; Ms. fol. 65v2–3. 17 See Kano and Li 2021, 17; Ms. fol. 66r5–v1. 18 See Kano and Li 2021, 17; Ms. fol. 66v1–2. 19 See Kano and Li 2021, 18; Ms. fol. 66v2–3. 20 To illustrate Buddha-nature, RGV I.52–63 compares it to ākāśa as being itself a foundation which does not admit of any subdivision. 21 Skt. Ms. fol. 54r5–v1; D 1198, 7b3–6; cf. Saṃpuṭa 1.1. 22 Maṇḍalavidhiṭīkā, D 1871, 80r6–v1 on Guhyasamājamaṇḍalavidhi v. 94.

Bibliography Erb, Felix. 1997. Śūnyatāsaptativṛtti: Candrakīrtis Kommentar zu den „Siebzig Versen über die Leerheit “des Nāgārjuna [Kārikās 1–14]. Tibetan and Indo-Tibetan Studies, no. 6. Stuttgart: Franz Steiner Verlag. Fujita, Shōdo. 2011. “Daijōbussetsuron no ichi danmen – Daijō sōgon kyōron no shiten kara 大乗仏説 論の一断面―『大乗荘厳経論』の視点から.ˮ Sirīzu daijō bukkyō 1 Daijō bukkyō towa nanika シ リーズ大乗仏教1 大乗仏教とは何か edited by Jikidō Takasaki et al., 113–50. Tokyo: Shunjūsha. Isaacson, Harunaga, and Francesco Sferra. 2019. “Indian Tantric Authors: Overview.” In Brill’s Encyclopedia of Buddhism, Vol. 2: Lives, edited by Jonathan Silk et al., 228–60. Leiden and Boston: Brill. Isoda, Hirofumi. 1993. “Abhayākaragupta to Madhyamakāloka AbhayākaraguptaとMadhyamakāloka.ˮ In Indo bukkyōgaku kenkyū: Miyasaka Yūshō hakase koki kinen ronshū jō インド学密教学研究 宮 坂宥勝博士古稀記念論文集 上, 501–16. Kyoto: Hōzōkan. ———. 1998. “Haramitsu no jissen – Abhayākaragupta Munimatālaṃkāra dai ni shō 波羅蜜の実践 – Abhayākaragupta『Munimatālaṃkāra』第二章 –. ˮ In Indo mikkyō no keisei to tenkai: Matsunaga Yūkei kokikinen ronshū インド密教の形成と展開：松長有慶古稀記念論集, 287–304. Kyoto: Hōzōkan. Kano, Kazuo. 2016. Buddha-Nature and Emptiness: rNgog Blo-ldan-shes-rab and a Transmission of the Ratnagotravibhāga from India to Tibet. Wiener Studien zur Tibetologie und Buddhismuskunde, no. 91. Wien: Arbeitskreis für Tibetische und Buddhistische Studien, Universität Wien. ———. 2017. “Vikramaśīla ji no rokkenmon wo meguru shiryō to sono mondaitenヴィクラマシー ラ寺の六賢門をめぐる史料とその問題点.ˮ Indogaku bukkyōgaku kenkyū 印度学仏教学研究 65 (2): 108–14. ———. 2019. “Seshin saku juujikyōron no bonbun itsubun 世親作十地経論の梵文佚文.ˮ Indogaku bukkyōgaku kenkyū 印度学仏教学研究 67 (2): 116–20. ———forthcoming. “Abhayākaragupta no daijō bussetsuron—Muniishusōgonron daiyonshō kara— アバヤーカラグプタの大乗仏説論－牟尼意趣荘厳論第4章から－.ˮ Mikkyō bunka 密教文化 247/248. Kano, Kazuo, and Li Xuezhu. 2014. “Bonbun muniishusōgon dai isshō matsubi bubun no kōtei to wayaku—Chūgankōmyō ichijō ronshō dan no bonbun danpen no kaishū 梵文『牟尼意趣荘厳』第 １章末尾部分の校訂と和訳―『中観光明』一乗論証段の梵文断片の回収―.ˮ Mikkyō bunka 密教文化 232: 7–42. ———. 2018. “Bonbun muniishusōgon dai isshō (fol. 61r5–64r2)—Chūgankōmyō sezoku to gonzetsu oyobi yuishinsetsu hihan kasho itsubun—梵文校訂『牟尼意趣荘厳』第一章（fol. 61r5–64r2）― 『中観光明』世俗と言説および唯心説批判箇所佚文―.ˮ Mikkyō bunka 密教文化 241: 31–56. ———. 2020. “A Survey of Passages from Rare Buddhist Works Found in the Munimatālaṃkāra.ˮ In Sanskrit Manuscripts in China III: Proceedings of a Panel at the 2016 Beijing Seminar on Tibetan Studies, edited by Birgit Kellner, Jowita Kramer, and Xuezhu Li. Beijing and Vienna: China Tibetology Publishing House/Institute for the Cultural and Intellectual History of Asia.

628

Abhayākaragupta ———. 2021. “Bonbun muniishusōgon dai isshō (fol. 64r2–67v2)—Chūgankōmyō itsubun gyōja no chokkanchi to mujishōronshō—梵文校訂『牟尼意趣荘厳』第一章（fol. 64r2–67v2）―『中観 光明』佚文・行者の直観知と無自性論証―.ˮ Mikkyō bunka 密教文化 246: 5–39. Kapstein, Matthew T. 2001. “Abhayākaragupta on the Two Truths.ˮ In Reason’s Traces: Identity and Interpretation in Indian and Tibetan Buddhist Thought, 393–415. Boston: Wisdom Publications. Luo, Hong. 2010. Abhayākaragupta’s Abhayapaddhati, Chapters 9 to 14. Critically edited and translated by Luo Hong, with a preface by Harunaga Isaacson and Alexis Sanderson. Sanskrit Texts from the Tibetan Autonomous Region, no. 14. Beijing/Hamburg: China Tibetology Publishing House, University of Hamburg. ———. 2018. “The Opening and Concluding Verses of Abhayākaragupta’s Madhyamakamañjarī.ˮ China Tibetology 2 (September): 15–23. ———. 2020. “A First Investigation of Abhayākaragupta’s Madhyamakamañjarī.ˮ Indogaku chibettogaku kenkyū 24: 57–75. Mori, Masahide. 2009. Vajrāvalī of Abhayākaragupta: Edition of Sanskrit and Tibetan Versions. 2 vols. Tring: The Institute of Buddhist Studies. ———. 2011. Indo mikkyō no girei sekai インド密教の儀礼世界. Kyoto: Sekaisisōsha. Seyfort Ruegg, David. 1977. “The gotra, ekayāna and tathāgatagarbha Theories of the Prajñāpāramitā according to Dharmamitra and Abhayākaragupta.ˮ In Prajñāpāramitā and Related Systems: Studies in Honor of E. Conze, edited by Lewis Lancaster and Luis O. Gomez, 283–312. Berkeley: Berkeley Buddhist Studies Series. ———. 1981. The Literature of the Madhyamaka School of Philosophy in India, Vol. 7, Fasc. 1 of a History of Indian Literature. Wiesbaden: Otto Harrassowitz. Sinclair, Iain. 2011. “Review of Vajrāvalī of Abhayākaragupta: Edition of Sanskrit and Tibetan Versions by Masahide Mori.” Saṃbhāṣā: Nagoya Studies in Indian Culture and Buddhism 29: 94–100. Tomabechi, Toru, and Kazuo Kano. 2008. “A Critical Edition and Translation of a Text Fragment from Abhayākaragupta’s Āmnāyamañjarī: Göttingen, Cod.ms.sanscr.259b.ˮ Tantric Studies 1: 22–44.

629

PART 8

Modern Philosophers

MODERN PHILOSOPHERS Introduction to Part 8

Indian Buddhist philosophy is typically associated with early figures such as Nāgārjuna, Vasubandhu, Dignāga, and their premodern commentators. But in the nineteenth and twentieth centuries, a number of Indian scholars and social reformers took up Buddhist thought in new and intellectually creative ways. Using the language of “Buddhist Revival” – and thus emphasizing what they described as India’s glorious Buddhist past – they reformulated Buddhism as a modern way of life appropriate for democratic, egalitarian societies. Figures such as Iyothee Thass, Rahula Sankrityayana, Dharmananda Damodar Kosambi, P. Lakshmi Narasu, and B. R. Ambedkar found in Buddhist traditions a universal religion, characterized by reason and morality, that they simultaneously proposed as a national religion for India. Responding to the pressing issues of the time – such as colonialism and independence; the formation of a new republic; and religious, class, and political conflicts – they sought resources and inspiration in Buddhist traditions for developing a non-sectarian morality; advocating for freedom of thought; and critically rethinking and reconstructing society to be more equal in terms of caste, gender, economics, and politics. The philosophers whose work is explored in this section thus display two of the many ways in which Buddhist philosophy has been adapted to address the challenges of modernity. In the first chapter of this section, William Edelglass focuses on three intertwined threads of the philosophy of religion, Buddhism as a socially engaged practice, and Buddhist political philosophy in the writings of Bhimrao Ramji Ambedkar (1891–1956), the most prominent of the modern Indian Buddhist figures. Ambedkar was born into an outcaste community, the Mahars, and went on to earn multiple graduate degrees at both Columbia and the London School of Economics, eventually becoming the first minister of law in the newly independent India and chair of the Drafting Committee of the Indian Constitution. While he is primarily known as the chief architect of the constitution, Ambedkar was also a prolific author who worked in disciplines across the social sciences and humanities, including philosophy. According to Ambedkar’s analysis, the inequality that permeated Indian society was due in part to the dominant forms of Indian religion; he argued for the annihilation of caste, as well as the abandonment of the religious framework that justified caste hierarchy. However, Ambedkar believed religion was still important, as it could provide the basis for a sacred morality and responsibility to others that would bind social groups together in a democratic society. Ambedkar devoted the last decade of his life to developing a Buddhist philosophy and practice DOI: 10.4324/9781351030908-55

633

Modern Philosophers

that would explicitly center the experience of marginalized communities and lead to a more free, just, and equitable society. Jay Garfield, in the final chapter, presents an overview of the philosophy of the Dalai Lama XIV (the Ven. Tenzin Gyatso, 1935–). Characterizing the Dalai Lama XIV as a modern Indian philosopher may not be an obvious choice. However, the Dalai Lama XIV has lived his entire adult life in India, his own philosophical work is rooted in Indian Buddhist traditions as taken up in Tibet, and for decades he has been in dialogue with Indian philosophical and religious thinkers. The Dalai Lama XIV’s philosophical writings include commentaries on Indian Buddhist texts, treatises in moral philosophy seeking to develop a universal ethic, reflections on science and Buddhist metaphysics, philosophy of religion and interfaith dialogues, and works devoted to nonviolence and political thought. As Garfield shows, these texts offer a distinctive synthesis of classical Indian and Tibetan Buddhist philosophy with contemporary Indian and Western thought and make the Dalai Lama XIV one of the most significant and influential contemporary Indian philosophers.

634

39 B. R. AMBEDKAR Justice, Religion, and Buddhist Political Philosophy William Edelglass

Introduction Bhimrao Ramji Ambedkar (1891–1956) served as the first minister of law of the newly independent India and is widely known for his work as chair of the Drafting Committee of the Indian Constitution.1 Ambedkar is also known as an ex-untouchable, a Dalit2 who founded and led numerous projects to promote social justice and political rights for subordinated communities, including schools, publications, and political parties. In addition, he organized and led protest movements demanding universal access to public water and an end to prohibitions against intercaste marriage and dining. For many years, he worked as a barrister to defend the rights of low-caste Indians and negotiated with the British and the Congress Party on their behalf. Ambedkar is also famous for his extensive education. After completing his B.A. in Economics and Politics at Elphinstone College at Bombay University, he was offered a scholarship to pursue graduate work at Columbia University, where he studied philosophy, history, sociology, and anthropology, eventually earning two M.A. degrees and then a Ph.D. in economics. In 1916, he went to London, where he studied law, was admitted to the bar at Gray’s Inn, and continued his research on Indian economic history at the London School of Economics, receiving a M.Sc. and D.Sc. in economics. Following his formal education, Ambedkar continued to read widely and deeply in a range of disciplines, which informed his work as a professor, dean, barrister, politician, publisher, activist, community leader, delegate to conferences with the British leading up to independence, leader of the People’s Education Society with its colleges, theorist of caste and society, and religious reformer. Ambedkar’s prodigious learning is reflected in his many writings and speeches: his writings and speeches in English were collected and published posthumously in seventeen volumes that together surpass 14,000 pages (under the title Dr. Babasaheb Ambedkar Writings and Speeches, hereafter cited as BAWS); in addition, there is a vast corpus of his writings in Marathi. Ambedkar’s academic study and much of his writing focused on economics, sociology, history, political science, and religious studies. His work was also informed by a deep engagement with philosophy.3 During his time at Columbia, Ambedkar took three seminars with the American philosopher John Dewey, with whom he felt a close connection, and he developed a lifelong interest in American pragmatism. According to Arun Mukherjee, So deeply embedded is Dewey’s thought in Ambedkar’s consciousness that quite often his words flow through Ambedkar’s discourse without quotation marks. . . . DOI: 10.4324/9781351030908-56

635

William Edelglass

Ambedkar not only borrowed concepts and ideas from Dewey, his methodological approach and way of argumentation also show Dewey’s influence. (2009, 347–48) While in New York to receive an honorary degree from Columbia, shortly after Dewey’s death in June 1952, Ambedkar wrote in a letter to his wife: “I owe my whole intellectual life to him” (Stroud 2018, 64). In addition to Dewey, Ambedkar was particularly influenced by Rousseau, Marx, and Bergson (Kumar 2015). He read widely in British, Greek, French, and German philosophy, as well as American pragmatism (Rodrigues 2017, 101). He also read and responded to the philosophical works of his prominent contemporaries in India. As a young student in India, Ambedkar had been denied access to Sanskrit and the study of the Vedas. Later, though, he devoted considerable attention to these Brahmanical texts, the Indian epics, and classical Indian philosophy, especially Sāṃkhya, which he critically engaged as the philosophical foundation of Hinduism. Ambedkar regarded philosophy and critical thought as necessary for addressing social injustice and drew on a wide variety of thinkers in his analyses of Indian society. Thus, in a single sentence, he can recommend both communist and Catholic texts, for example, arguing in a speech at the 1943 All India Trade Union Worker’s Study Camp that everyone in the working class should study Rousseau’s Social Contract, Marx’s Communist Manifesto, Pope Leo XIII’s Encyclical on the Conditions of Labor, and Mill’s On Liberty (BAWS vol. 10, 110). Ambedkar regarded contemporary Hinduism  – and the Brahmanical traditions which gave it authority – as a primary cause of social conflict and injustice in India. Moreover, he was deeply critical of the conflation of religion and nationalism that permeated the anticolonial struggle and obscured the profound inequality and social suffering fostered by India’s caste society. Because of these views and his leading role in drafting the Indian Constitution, Ambedkar is often associated with secular constitutionalism. But his constitutionalism was informed by a conception of political faith. This political faith includes both the trust between citizens that is a necessary condition for democratic politics and a shared faith in a transcendent moral good that is ultimately, according to Ambedkar, the only viable basis for protecting the most vulnerable and marginalized in any society. While this shared faith need not be explicitly religious, over time, Ambedkar did argue for a revolutionary reconstruction of Indian society specifically grounded in religious transformation. He came to believe that while change in the political and legal sphere was necessary, such change was insufficient for the radical social transformation required to dismantle the caste structure of graded inequality that privileged some and subordinated many. In particular, after years of reading in comparative religion, he argued that Buddhism was uniquely suited to support a moral order in which social conflict and injustice would be diminished. Drawing on classical Buddhist texts and contemporary scholarship on Buddhism, as well as the work of Buddhist activists in India and European philosophers and social scientists, Ambedkar developed a Buddhist philosophy that sought to reconstruct Buddhism in ways that would address the suffering of marginalized communities to create a more just society. Ambedkar’s engagement with Buddhism culminated in two major events in twentiethcentury Indian Buddhism. First, on October 14, 1956, Ambedkar publicly converted to Buddhism by taking refuge in the Three Jewels, accepting the Five Precepts, and making twenty-two vows, including the vow to “endeavor to establish equality” (Ambedkar 1989, vol. 17, 529– 31). The second event was the publication, shortly after his death in December 1956, of The Buddha and His Dhamma (hereafter BHD). In BHD and other late works, Ambedkar develops what he calls a “neo-Buddhism,” what many Ambedkarite Buddhists call “Navayāna,” or “new vehicle,” a socially engaged Buddhism that can play a role in reconstructing the world. 636

B. R. Ambedkar

At the same time, he was developing a Buddhist political philosophy informed by the ideals of liberty, equality, and fraternity. According to Ambedkar, only in a constitutional democracy can liberty, equality, and fraternity all be present: liberty without limits undermines equality; valorizing equality without limits undermines liberty. Only a social order grounded in fraternity, or the sympathies of shared ways of life, which Ambedkar comes to call maitrī (the Sanskrit term for benevolence, one of the four divine abodes [brahmavihāras] cultivated in Buddhist practice [mettā in Pali]), can sustain the conditions for a society that embodies liberty, equality, and fraternity – or, as Ambedkar characterizes it, a just society. Today, like some of the other figures in this volume, Ambedkar  – widely referred to as “Babasaheb” by Ambedkarite Buddhists – is venerated as a bodhisattva or even a buddha.

Ambedkar’s Philosophy of Religion In his posthumously published Philosophy of Hinduism, Ambedkar argues that the discipline of comparative religion has demonstrated the arbitrariness of European hierarchies of religions. According to these hierarchies, Western “revealed religions” were superior to other religious traditions. Comparative religion thus contributes to a move beyond Eurocentric frameworks, making space for non-Western traditions to be recognized as equally valid religions. In the twentieth century, this recognition of religious relativism was part of a larger anticolonial rethinking of culture. At the same time, even as he recognizes the anticolonial power of religious relativism and the “capricious distinction between true and false religions based on purely arbitrary and a priori considerations,” Ambedkar believes that comparative religion has resulted in some misconceptions. “The most harmful one,” according to Ambedkar, is the idea “that all religions are equally good and that there is no necessity of discriminating between them” (BAWS vol. 3, 69). In contrast, Ambedkar writes at the beginning of the book, philosophy of religion is both descriptive and normative. It is descriptive to the degree that it addresses the teachings of a particular religion; “in so far as it involves the use of critical reason for passing judgment on those teachings,” however, “the Philosophy of Religion becomes a normative science” (BAWS vol. 3, 5). Ambedkar argues that analyzing and evaluating religions is necessary because like any other institution, particular religions can have beneficial or harmful impacts on individuals, groups, and societies. In defending the role of religion in reconstructing society, Ambedkar was informed by a wide array of Indian and Western thinkers. These include Indian religious reformers such as M. G. Ranade, R. G. Bhandarkar, Gopal Hari Deshmukh, Gopal Ganesh Agarkar, and especially P. Lakshmi Narasu, Mahatma Jotirao Phule, and Dharmananda Kosambi. While there were many differences between these thinkers and Ambedkar was not in complete agreement with any of them, he shared their view that a critical analysis and ultimately a transformation of religion and caste would be more effective than focusing solely on transforming political structures. He also agreed with them that religion should focus on human social relations rather than being concerned primarily with God, the soul, and life-after-death. This, he believed, was a necessary element in the construction of a more socially just and equal India. Ambedkar also engaged Western scholars of religion. He read Marx with care but makes a clear departure from the Marxist view of religion as justifying inequality. Like Weber, Ambedkar emphasizes the ways in which religion influences society; unlike Weber or Marx, he did not believe that religion would eventually wither away. Rather, as with Durkheim, he believed that religion provided a foundation for the moral order and was therefore necessary for social and political life (Omvedt 2004b). Ambedkar’s understanding of religion was deeply influenced by Durkheim’s The Elementary Forms of Religious Life, in which Durkheim proposes a kind 637

William Edelglass

of Rousseauean “civil religion,” grounded in a universal reason and morality that could bind communities together in solidarity. It was also informed by Dewey, who saw in civil religion the possibilities of forging a common foundation for sustaining and supporting democratic life and individual rights. Finally, as Aishwary Kumar has shown, Ambedkar’s mature understanding of religion was deeply influenced by Bergson’s The Two Sources of Morality and Religion (Kumar 2015), in which Bergson distinguishes between two forms of morality and religion. One is closed and static, with rigid rules that function to support the basic needs of a community. A closed morality and religion excludes other communities, so communities with a closed morality and religion are always preparing for conflict with other societies or even war. In contrast, Bergson suggests, an open morality and religion is oriented towards progress and creativity. Open morality and religion is not exclusive to one’s own community; it is universal and therefore includes other communities. It is oriented toward peace and solidarity instead of war. There is an echo of Bergson’s characterization of two forms of morality and religion in the distinction between religion grounded in principles and religion based on rules and ritual that is at the heart of Ambedkar’s philosophy of religion. In his 1936 text, Annihilation of Caste, Ambedkar writes: A religious act may not be a correct act but must at least be a responsible act. To permit for this responsibility, religion must mainly be a matter of principles only. It cannot be a matter of rules. The moment it degenerates into rules it ceases to be religion as it kills responsibility which is the essence of a truly religious act. (BAWS vol. 1, 75) His discussion of nonviolence provides a good example of what he means when he argues that a responsible act must be a matter of principle. According to Ambedkar, the Buddha articulated nonviolence as a way of life and a matter of principle, not a rule. The Buddha did not simply prohibit violence but rather prohibited violence only when it was unnecessary. Thus the prohibition against violence was not a rule to be followed blindly: “He enunciated it as a matter of principle or way of life. . . . A principle leaves you freedom to act. A rule does not” (BHD, 183). Even though the Buddha was in principle against violence, “he was also in favor of justice and where justice required he permitted the use of force” (BHD, 184). A truly religious act, then, is a responsible act because it requires freedom of thought, the freedom to discern if a principle is applicable in a particular context. And because contexts are always changing, a true religion would be one that abandoned any claim that its ideas, texts, or rituals were infallible. Instead, a true religion would always be open to rethinking its doctrines and practices based on principles. True religion, then, provides a framework for thinking for ourselves and attunes us to the moral dimension of our relations with others, helping us discern what we ought to do in any particular context. In contrast, a religion of rules – in which thoughts and actions are determined – undermines freedom of thought, blinds us to the needs and value of others, and creates an excessively powerful priestly class through its emphasis on ritual (BHD, 47). “The word ‘religion’,” Ambedkar writes, “is an indefinite word with no fixed meaning” (BHD, 167). There are, he suggests, very different forms of religion, and they need to be distinguished. For Ambedkar, true religion is moral, social, and rational and therefore belongs to the public sphere. This puts him at odds with much of European Enlightenment thinking, which characterizes religion as a private and personal affair (BHD, 168). In Enlightenment thought, the very distinction between religion and the secular was taken to be a distinction between the private, personal realm and a realm of public reason in which autonomous agents could argue based on principles. By situating religion in a transcendent realm that was complementary to but outside the realm of public reason, Enlightenment thinking allowed traditional 638

B. R. Ambedkar

religious belief and practice to be preserved as a personal matter alongside the public requirements of rationality. In Ambedkar’s view, however, a religion which is said to “transcend reason” is nothing more than a religion of rules and rituals, governed by superstition, in which moral responsibility and free thinking based on principles are weakened. In contrast to religion as private and personal, Ambedkar describes true religion, what he often refers to as “Dhamma,” as social. It is social because fundamentally it is about social relations and morality: “Morality is Dhamma and Dhamma is Morality,” Ambedkar claims (BHD, 172). Dhamma is not about performing rituals or a relationship with God; it is about a moral relationship with others: “one man, if he is alone, does not need Dhamma. But when there are two men living in relation to each other, they must find a place for Dhamma whether they like it or not” (BHD, 168). Dhamma, or true religion, is about the moral order; it is about responsibility to others. By yoking religion with responsibility and responsibility with principles, Ambedkar positions religion as also about justice and universality. Thus, according to Ambedkar, religion is not a private matter and must be evaluated in the public sphere. In “Buddha and Future of His Religion,” published in the journal of the Maha Bodhi Society in May, 1950, Ambedkar articulates four requirements for a suitable contemporary religion. It must (1) provide a morality to govern society; (2) be consistent with science; (3) recognize and affirm liberty, equality, and fraternity; and (4) not regard poverty as ennobling. In other words, it must be what Bergson would call an open religion, one which aims toward peace between social groups. Or, in Ambedkar’s terminology, it would be a religion of the principle, a religion of responsibility, a religion of morality. “So far as I know,” Ambedkar writes, “the only religion which satisfies all these tests is Buddhism” (BAWS vol. 17.2, 105).

Ambedkar’s Philosophy of Hinduism Ambedkar was born in 1891 in Mhow, Maharashtra, the fourteenth child of a military schoolteacher who belonged to the Mahar jāti.4 Mahars, the largest outcaste community in Maharashtra, were traditionally responsible for defending villages and apprehending criminals. Some, like Ambedkar’s father, were able to find employment in the British army. Thus, in the hierarchy of jātis outside the varṇa system, the Mahars were above some other groups. But they also had other “defiling” responsibilities, such as caring for corpses. Thus, while they were often regarded as inherently impure, that impurity was magnified by their polluting occupations. Their perceived impurity was regarded as contagious, and as outcastes, they were marginalized in social, economic, and political life. Stigmatized by his Mahar identity, Ambedkar’s existence was conditioned by the violence of this marginalization. At school he was unable to sit at a desk, drink from the water fountain, use the chalkboard, or even be close to his classmates; teachers would not touch his papers. Despite his accomplishments in New York and London, upon his return to India, he was still on occasion refused housing, for example, as Mahars were often denied access to the common goods of public life, such as drinking water, transportation, and accommodation, as well as education, and sometimes even public roads.5 The countless indignities and constant humiliation, exclusion, and discrimination resulted in Ambedkar’s experience of himself as outside Indian social life: “I am not a part of the whole; I  am a part apart” (BAWS vol. 2, 261). Or, as Ambedkar famously told Gandhi in 1931: “I have no homeland” (BAWS vol. 17, part 1, 53). Ambedkar drew on his wide-ranging studies of history, economics, culture, society, philosophy, and religion when analyzing Indian society, trying to understand and end the violence and social injustice that conditioned so much of his own lived experience. According to Ambedkar’s analysis, the violence of untouchability and the unequal distribution of rights 639

William Edelglass

and privileges are inherent to the varṇa system. Ambedkar characterizes this social structure as a form of “graded inequality” (BHD, 56). Men in the lowest groups in the hierarchy, for example, were denied the privilege of bearing arms or accessing education, the very resources necessary for liberation from the oppression of the caste system. Women of any caste were also denied education; thus, like lower-caste men, they were unable to follow the Brahmanic/ Hindu ideal stages of life, which begin with the stage of the student. The superior power and authority of some groups over others, then, was not a mere side effect of Brahmanism/Hinduism, Ambedkar argues, but its essence: “The inequality preached by Brahmins was its official doctrine. . . . It was opposed to equality” (BHD, 57–58). According to Ambedkar’s analysis, the graded inequality at the heart of the caste structure does not result in social harmony, freedom, or morality. Instead, it leads to “an ascending scale of hatred and a descending scale of contempt,” as the lower castes hate those who are above and the higher castes have contempt for those who are below in the social hierarchy (BHD, 58). Caste thus damages those in both the lower and the higher ranks of the social order. Ambedkar recognizes several forces that maintain this system. First, the structure of graded inequality ensures that most people, no matter how much they resent the rights and privileges of those who occupy superior positions in the hierarchy, do not want to lose their own rights, privileges, and power over those in subordinate positions. Second, regardless of the violence and suffering inherent to the caste system, many Hindus believe that it is a “divinely ordained social order” grounded in sacred Brahmanical texts. While caste hierarchy is a feature of Indian society across all religious communities – including Muslim ones, which also sometimes contain so-called “untouchable” jātis not justified by a logic of karma – Ambedkar emphasizes the way in which caste is legitimized by the Brahmanical understanding of karma, a form of theodicy that explains one’s status at birth as determined by actions in previous lives. Caste hierarchy is also codified in the Laws of Manu (Manusmṛti), which Ambedkar regards as the most authoritative Brahmanical legal text, with its horrific punishments for those outside the varṇa system who overstep the limits of their social condition. Already in Annihilation of Caste, Ambedkar argues that it is not that people are somehow wrong because they observe caste. Rather, they are simply being religious, and their religion asks them to observe caste, with all its exclusions. Thus, it is not people but their religion which is wrong (BAWS vol. 1, 68). Because Ambedkar believes caste to be at the heart of Hinduism, he argues that Hinduism cannot be reformed; it has to be rejected. Perhaps Ambedkar may have already thought as much when he publicly burned a copy of the Laws of Manu in 1927. Indian elites in the independence movement in the late nineteenth and early twentieth centuries often conflated Hinduism with nationalism. For Ambedkar, though, freedom required not just independence from the British but freedom from Hinduism itself, a religion of rules with its scriptures held to be infallible, justifying an oppressive and violent social order that set communities against each other. Ambedkar sought to show that the Indian social order, an order widely accepted as a natural given, was arbitrary and unjust and ought to be annihilated.

The Buddha and His Dhamma: Ambedkar’s Buddhism In contrast to his analysis of Hinduism as a religion of rules, Ambedkar presents Buddhism as a religion of principles committed to freedom of thought. “If a modern man who knows science must have a religion,” he writes in the preface to BHD, “the only religion he can have is the Religion of the Buddha” (BHD, xxv). Buddhism, in this account, is scientific because the Buddha makes no use of miracles to teach the Dhamma (BHD, 235). In contrast to other religions grounded in claims of divine authority or supernatural revelation, the Buddha teaches 640

B. R. Ambedkar

through rational argument, the views of one person that could and should be challenged: “He never claimed infallibility for his message. . . . It was based on universal human experience of life in the world. He said that it was open to anyone to question it, test it, and find what truth it contained” (BHD, 121). The Buddha, in Ambedkar’s account, emphasizes wisdom and rationality “because he did not wish to leave any room for superstition” (BHD, 168). Seeing Buddhism as a religion of principles, Ambedkar regards responsibility as the heart of Buddhism: “morality,” in Buddhism, “has been given the place of God” (BHD, 132). Ambedkar sees this in the Buddha’s emphasis on the Five Precepts, the Eightfold Path, and the Ten Perfections, which he regards as “the basis of his religion.” According to Ambedkar, this is what distinguishes Buddhism from other religions (BHD, 152). Ambedkar’s Buddhism, as Gitanjali Surendran shows, was preceded by numerous Indian thinkers who argued that Buddhism was universal, rational, and ethical and an appropriate civic religion for India (Surendran 2013).6 Ambedkar develops these ideas with a particular emphasis on Buddhism as especially responsive to the lives of the poor and marginalized. He rejects the idea of a stable canon understood as the true teaching of the Buddha. Instead, Ambedkar explicitly articulates three hermeneutic principles that justify and guide his interpretation, allowing him to reject doctrines he believed to be introduced by monks or Brahmins that are in tension with the rational, democratic, and moral teachings of the Buddha. Because the Buddha “was nothing if not rational, if not logical,” Ambedkar argues, we can accept what is logical and rational as the word of the Buddha. Ambedkar’s second hermeneutic principle is that “anything attributed to the Buddha which did not relate to man’s welfare cannot be accepted to be the word of the Buddha.” Finally, he insists, we need to pay attention to the distinction between those views which the Buddha articulated with certainty and those about which the Buddha only expressed his views tentatively (BHD, 184–85; Queen, 1996). These hermeneutic principles, especially rationality and social benefit, are apparent throughout BHD. In his introduction, Ambedkar presents what he views as four problems in the traditional accounts of the life and teachings of the Buddha. They are problems because, in his view, they do not appear to be rational or of social benefit. The first concerns the Buddha’s motivation to renounce the world and pursue a spiritual path. According to tradition, the Buddha’s renunciation comes about because, while on a chariot ride outside the palace at the age of twenty-nine, for the first time he saw someone who was sick, someone else who was old, and a corpse. As Ambedkar notes, the idea that as a grown man he would not have encountered sickness, ageing, and death is absurd (BHD, xxix). The story that Ambedkar tells instead is one in which Siddarth Gautama belonged to the minority anti-war view in the Sakya council, which had resolved to wage war against the neighboring Koliyas. To avoid a social boycott against his family or the confiscation of their lands, Siddarth offers to leave his home and renounce the world (BHD, 17–23). When Gautama hears that the war between the Sakyas and the Koliyas is over, he wonders if he should return home. On further reflection, he realizes that the problem of war – which originally motivated his leaving home – is part of a much larger problem of conflict that can be seen between and within nations, families, and communities. But while the conflict between nations, for example, is intermittent, “the conflict between classes is constant and perpetual. It is this which is the root of all sorrow and suffering in the world.” Recognizing that the issue that originally led to his departure from home is the much more pervasive problem of social conflict, he determines to explore for himself whether there is a solution to the “root of all sorrow and suffering” (BHD, 41). This analysis of the root of suffering in social conflict suggests a more politically engaged approach to the cause of suffering than we find in the canonical account of the Four Noble 641

William Edelglass

Truths, which Ambedkar raises as the second problem. He suggests that the Four Noble Truths – that suffering is pervasive, that this suffering arises with craving, that to let go of suffering is to let go of craving, and the Eightfold Path to letting go of craving – were not taught by the Buddha.7 Instead, it was monks who later introduced what he regards as the pessimistic teaching of the Four Noble Truths, focusing on the individual’s path rather than the social conditions of the community that, according to Ambedkar, result in the conflict that was the Buddha’s primary concern (BHD, xxx). It is pessimistic, in Ambedkar’s view, because it does not seem to offer a way to address the immediate sufferings caused by poverty and social injustice. As he argues later in the text, it is not simply greed and delusion that cause suffering: “man’s misery is the result of man’s inequity to man” (BHD, 152). Ambedkar repeatedly emphasizes that it is the social and material conditions of our lives that cause suffering and that “the Buddha’s conception of Dukkha is material” (BHD, 265). This is why, as Ambedkar frequently insists, the Buddha found poverty and destitution abhorrent and never claimed, as we hear in Jesus’s famous Sermon on the Mount, that the poor and the downtrodden are blessed. Instead, he encourages the moral acquisition of adequate wealth and justice in this world. Any account that obscures or downplays the role that material and social conditions play in human suffering – for example, by suggesting that the marginalized can achieve freedom from their suffering through individual practice  – distracts us from the actual causes of suffering and leaves them in place. Ambedkar’s critique of the canonical view of freedom from suffering does not mean that individual practice is not important. Indeed, he argues that Buddhism teaches us that instead of striving for “some imaginary heaven,” to remove our misery, “each one must learn to be righteous in his conduct in relation to others, and thereby make the earth the kingdom of righteousness.” For Ambedkar, this emphasis on moral cultivation is what distinguishes Buddhism from other religions, particularly through the Five Precepts, the Eightfold Path, and the Ten Perfections. These are, in Ambedkar’s words, “the basis” of Buddhism, “because they constitute a way of life which alone can make man righteous.” This “way of life” is a form of cultivation that begins with training the mind in order to cultivate and train our dispositions – it requires courage to stand up and do what is right (BHD, 152–53). Morality, then, is the very heart of Buddhism. We are called to cultivate and train our dispositions and virtues, but we are not responsible for the suffering caused by our social conditions. Ambedkar’s insistence on not holding individuals responsible for the suffering that arises from their social conditions is also reflected in his approach to the third problem he proposes to address: how to understand karma (BHD, xxxi). One conception of karma suggests that social conditions are determined by actions in previous lives. According to this view, which Ambedkar associates with Brahmanism/Hinduism, there is no innocent suffering. The corollary of this view is that the soul is reborn and can therefore suffer the consequences of past actions in future births (BHD, 178–82). Ambedkar distinguishes this view from what he characterizes as the Buddha’s view of karma. According to Buddhist doctrine, because there is no substantial self, there is no being who is reborn who could suffer the consequences of actions in a past life (BHD, 174).8 This does not mean that Ambedkar rejects karma or denies the importance of karma for the Buddha. Indeed, he argues, “to believe that karma is the instrument of moral order is dhamma.” For Ambedkar, the law of karma means that human actions create the moral order. Skillful actions contribute to a better moral order, and unskillful actions lead to a worse moral order. Whether there is a good or bad moral order, resulting in less or more conflict and human suffering, is up to human, not divine, beings (BHD, 131–32). The fourth problem that Ambedkar poses in his introduction is the Buddha’s understanding of the proper role of monastics. “Was the object to create a perfect man?” Ambedkar 642

B. R. Ambedkar

asks, “Or was his object to create a social servant devoting his life to service of the people and being their friend, guide, and philosopher?” (BHD, xxxi). Quoting especially from the Dhammapada, Ambedkar notes that a Buddhist monastic  – he uses the Pāli term Bhikkhu  – should cultivate virtues of temperance and wisdom. A  Bhikkhu is disciplined, has faith, integrates meditation and knowledge, lives a life of purity, and practices restraint. Still, Ambedkar is very clear: the Bhikkhu is not an ascetic. Indeed, he regards asceticism as pointless: “There is no escape from the world even for an ascetic.” With Marx’s famous thesis in mind – “the philosophers have only interpreted the world, in various ways; the point is to change it” – Ambedkar continues, “what is necessary is to change the world and to make it better.” Rather than a futile devotion to escaping suffering as an individual, the Buddha realizes that “it was his duty to return to the world and serve it, and not sit silent as the personification of inactive impassivity” (BHD, 65). And the Bhikkhu is not like the Brahmin, who is born a Brahmin and whose primary responsibility is to perform rituals. Instead, the Bhikkhu – without caste, family, or property – is primarily interested in spreading and cultivating the dhamma through teaching and conversion, thereby increasing virtue and improving the moral order. The role of the Bhikkhu is to serve the larger community (BHD, 226–40). The Buddha’s teaching was not oriented primarily toward the monk; it was “principally for the laity” (BHD, 239). In BHD, impermanence and emptiness – often regarded as the two most significant metaphysical concepts in Buddhism  – are also interpreted in an ethical framework. Analyzing Ambedkar’s employment of emptiness, Aishwary Kumar shows how he “imbues the word with an ethical substance” (Kumar 2018, 187). It is ethical because, according to Ambedkar, it describes our creaturely finitude – our bodily vulnerability, our need for sustenance and physical comfort – which he regards as the basis for the “sentiment of fraternity.” Kumar points to a passage in The Philosophy of Hinduism where Ambedkar contrasts rational exhortations to be kind to others because we are all children of God with his own focus on how fraternity arises. He argues that the “condition for the growth of this sentiment of fraternity lies in sharing in the vital processes of life. It is sharing in the joys and sorrows of birth, death, marriage and food. Those who participate in these come to feel as brothers” (BAWS vol. 3, 64). For Ambedkar, our creaturely finitude – named by the Buddhist terms “emptiness” and “impermanence” – is the basis of morality because it leads us to gather and share our lives with others and cultivate sympathy and friendship. It is also the basis of justice, which Ambedkar, like Aristotle, believes to be dependent on fraternity (Kumar 2015, 294).

Ambedkar’s Buddhist Political Philosophy Ambedkar was the chief architect of the Indian constitution, which granted equal political rights to all adult citizens. He believed equal rights enshrined in the law were necessary for justice. Ambedkar was committed to the values of a modern, liberal democracy, values he worked hard to write into the constitution. He affirmed the citizen virtues and individual rights that enable shared democratic life. And yet, while “the idea of making a gift of fundamental rights to every individual is no doubt very laudable,” Ambedkar writes, “the question is how to make them effective.” Ambedkar notes that it is widely held that rights are safeguarded as soon as they are written into law. But, he argues, experience shows otherwise: “rights are protected not by law but by the social and moral conscience of society.” If the majority of a community opposes these rights, they cannot be guaranteed by law, legislators, or judges. Of what use, he asks in 1943, are fundamental rights to Blacks in the United States or to Untouchables in India? (BAWS vol. 1, 222) Any democratic constitution, even while affirming noble ideals, 643

William Edelglass

can function to suppress the lives of minorities and the poor. Therefore, to really achieve the kind of equality necessary for justice and democratic life, one needs to pay attention to and work at the level of social relations. Ambedkar clearly states this view in a speech on All-India Radio in October 1954: “Indians today are governed by two different ideologies. The political ideal set out in the Preamble to the Constitution,” a preamble written by Ambedkar himself, “affirms a life of Liberty, Equality, and Fraternity.” The problem, though, is that “their social ideal embodied in their religion denies them” this life (BAWS vol. 17.3, 503). Earlier in that same speech, Ambedkar insists that he has not borrowed his political philosophy – at the heart of which are democracy and the ideals of liberty, equality, and fraternity – from French Enlightenment thought, as listeners might expect. Rather, he says, “My philosophy has roots in religion and not in political science. I have derived them from the teachings of my Master, the Buddha” (BAWS vol. 17.3, 503). We may be tempted to dismiss this claim, as Ambedkar had been invested in the political ideals of liberty, equality, and fraternity long before his turn to Buddhism. But, in the end, it is through Buddhism that he frames his mature understanding of how liberty, equality, and fraternity are related and how they can sustain a democratic way of life. According to Ambedkar’s account, the Buddha valued liberty and equality. However, liberty and equality are in tension. Liberty without limits, in Ambedkar’s interpretation of the Buddha, destroys equality, as we see today with the rise of massive inequality in the wake of deregulation. And equality without limits destroys liberty, as Ambedkar argues in his reflections on Marxism. While we may appreciate and support the goal of creating equality that motivated the Russian Revolution, he argues, “society cannot afford to sacrifice fraternity or liberty. Equality will be of no value without fraternity or liberty” (BAWS vol. 3, 462). But there are limits to what a constitution or the law can do to achieve equality or fraternity. In Ambedkar’s presentation of the Buddha’s views, “law had a place only as a safeguard against the breaches of liberty and equality” (BAWS vol. 17.3, 503). The law, or the constitution, could help when liberty or equality were transgressed here and there. But, quoting Edmund Burke, he observes that while the law may hold an individual perpetrator accountable, it is powerless against a multitude who choose to transgress it (BAWS vol. 1, 222). According to Ambedkar, the Buddha also “did not believe that law can be a guarantee for breaches of liberty or equality” (BAWS vol. 17.3, 503). As he writes in “Buddha and Future of His Religion,” “In all societies, law plays a very small part. It is intended to keep the minority within the range of social discipline.” The majority is constrained not by law but by morality. Thus, while the law and the constitution are important, it is morality that is necessary to actually achieve equality. “Religion in the sense of morality, must therefore, remain the governing principle in every society” (BAWS vol. 17.2, 104). This form of religion as morality, Ambedkar’s Buddhism, is often interpreted as a secular, modernist project primarily serving a political purpose.9 However, alongside Ambedkar’s discourse of religion as instrumental, secular, and thoroughly rational, he also presents Buddhism as a religion of principles and responsibility that calls for an ethical response beyond rules and social custom. Thus, Ambedkar’s Buddhism makes space for an alternative to the dominant Indian forms of modernity. As Debjani Ganguly argues, in making this space, and thus making Indian modernity less coercive, Ambedkar’s modernity “dares to speak the non-sociological, non-secular language of transcendence” (Ganguly 2002, 344). For even as Ambedkar’s Buddhism constitutes a kind of critique of Brahmanism/Hinduism – and, one could argue, of much previous and existing Buddhism as well – it is also, as Ajay Skaria has argued, a critique of “secularism, and that criticism is articulated moreover as a religion” (Skaria 2015, 451). This critique gestures toward a responsibility to others, an attentiveness to the vulnerability and need of embodied life that transcends rules, social custom, and natural relations and thus also 644

B. R. Ambedkar

instrumental rationality. As Aishwary Kumar notes, for Ambedkar, secularism as a political program or social condition does not offer a responsible and just alternative to religious fanaticism and the oppression of religious minorities. According to Kumar’s reading of Ambedkar’s critique, at [secularism’s] empty center is not the truthful equality of faiths but an unraveling project to manage the insurrectionary force of the outnumbered minority, held in its place by a peculiarly modern pact between constitutional restraint in the political realm and fanatical populism in the social. (2018, 156) Thus, while Ambedkar is deeply committed to liberal democracy, he is critical of a purely secular modernity for, as he claims in BHD, “the only way to put a stop to conflict is to have common rules of morality which are sacred to all.” Without a sacred morality, without a morality grounded in a force that is more than secular, “the privileged remain privileged. . . . This means that there can be liberty for some, but not for all. This means that there can be equality for a few, but none for the majority.” The remedy, Ambedkar argues, is fraternity, which he sometimes refers to as maitrī, the Buddhist term for benevolence, care, friendship, and kindness to others. Maitrī, or fraternity, he argues, “is another name for morality. This is why the Buddha preached that Dhamma is morality; and as Dhamma is sacred, so is morality” (BHD, 173). Because law is secular, Ambedkar suggests, there can be social norms that transgress it with impunity. “Law is secular, which anybody may break,” Ambedkar insists, but “fraternity or religion is sacred which everybody must respect” (BAWS vol. 17.3, 503). There is a necessary power to religious ideals that, Ambedkar believes, is beyond anything a purely secular ideal can offer. This is why, as Kumar argues, “it is impossible to ignore the religious foundation of his political thought and practice” (2015, 32). True religion, then, is a religion that supports fraternity, that inculcates a way of life that is attentive to the vulnerability and need of others and responsible to ideals of justice. It is a religion that therefore creates the conditions for equality and freedom. This is why Ambedkar argues, sometimes quoting Burke, that “religion is necessary for a free society” (BAWS vol. 1, 76; BAWS vol. 3, 442). By a “free society,” Ambedkar means democracy. But, a democracy is more than a form of Government. It is primarily a mode of associated living. The roots of Democracy are to be searched in the social relationship, in the terms of associated life between the people who form a society. (BAWS vol. 17.3, 519) How would such a society be described? It is one characterized by cooperation, solidarity, mutuality of sympathy, common purpose, a sense of belonging to a common community. Brahmanism/Hinduism, in Ambedkar’s view, with its caste system, obstructs the very social foundations necessary for a democracy of liberty, equality, and fraternity. The Buddha, on the other hand, Ambedkar writes, “was born a democrat and he died a democrat . . . he was a thorough equalitarian” (BAWS vol. 3, 451–52). And the Buddha’s teachings support democracy because Dhamma, in his view, “breaks down barriers between man and man” (BHD, 161). Democracy requires a society in which the minority is not excluded. A  parliamentary democracy may enshrine individual political rights in a constitution and ensure freedom of contract, but such a democracy could also maintain social and economic inequality. Instead, for Ambedkar, citizens in a democracy should have the freedom and power to make society 645

William Edelglass

more just. Contrasting his own definition with Lincoln’s famous line – that democracy is “a government of the people, by the people, and for the people” – Ambedkar defines democracy as “a form and a method of government whereby revolutionary changes in the economic and social life of the people are brought about without bloodshed” (BAWS vol. 17.3, 475).10

Conclusion During a conference in October  1935 at Yeola, located in the Nashik district of Maharashtra, Ambedkar famously declared that while he may have been born a Hindu, “I solemnly assure you that I will not die a Hindu” (BAWS vol. 17.3, 95). According to Ambedkar, freedom from untouchability required the annihilation of caste, and because caste was at the heart of Brahmanism/Hinduism, it meant an emancipatory political choice of leaving the religion into which he had been born.11 On October 14, 1956 (a day associated with Emperor Aśoka’s conversion to Buddhism), in Nagpur (a city associated with Buddhist history), Ambedkar converted to Buddhism. His conversion was witnessed by almost half a million Mahars, who, that weekend, also converted to Buddhism. Within a short time, the vast majority of Mahars had converted to Buddhism, and today there are millions of Indians who think of themselves as “Ambedkarite Buddhists.” Statues of Ambedkar can be seen in the private homes of many Dalits and in public places across India. Despite his success in the conversion of millions of Dalits, critics have pointed out that the universal religion and morality that Ambedkar sought to achieve has not been realized: while Ambedkar’s goal was a universal maitrī that could overcome divisive communal identities, Neo-Buddhism today is very much associated with Dalits and thereby seen by some as replicating casteist thinking (Fuchs 2001, 265). While Ambedkar never argued that all forms of Buddhism, or even any actually existing form of Buddhism, was a true religion of principles that could sustain a moral social order, recent Buddhist atrocities and genocide against Rohinga in Myanmar and against Tamils in Sri Lanka, the expulsion of so many Nepali-speaking Hindus in the name of a nationalist Buddhist society in Bhutan, the violent ideology of Imperial Zen in Japan, or any number of other historical instances of Buddhist violence against outgroups should give us pause in thinking that mass conversions to Buddhism would somehow create a shared sympathy across social groups that results in equality, liberty, and justice. In the end, then, we are still left with questions about shared life in democratic societies: How do we cultivate fraternity when we belong to different social groups? How do we live in a pluralistic society that is not held together by one established religion? Or, if shared religious commitments are necessary to cultivate fraternity, how do we avoid the problem of religious nationalism? Although Ambedkar’s project of conversion may not have led to the all-encompassing social transformation he envisioned, it did contribute to the increased power and voice of Dalits, often manifest by protests in parks across India named after Ambedkar. Furthermore, while we may be left with some of the same political questions with which he struggled, Ambedkar’s analysis of the limits of the law and the need for cultivating fraternity in pluralistic democracies is still relevant today. He is a capacious thinker who explores the intersecting dynamics of subordination based on gender, class, caste, and race.12 And while his social context was very much Indian society, he was a cosmopolitan thinker. As Ambedkar wrote in a 1946 letter to W. E. B. Du Bois, “There is so much similarity between the position of the Untouchables in India and of the position of the Negroes in America that the study of the latter is not only natural but necessary.”13 Ambedkar’s analysis of the limits of democratic constitutions and the law for achieving social equality and the rights of minorities drew on other social contexts and is applicable well beyond Indian caste society. 646

B. R. Ambedkar

For Buddhists in particular, in India and elsewhere, Ambedkar provides a model of a socially engaged Buddhism, rooted in tradition and addressing the needs of contemporary people. His is a model of Buddhism as a way of life that works toward a more just society and recognizes – invoking and revising Marx – that “the function of Religion is to reconstruct the world” (BAWS vol. 3, 442). This is why the slogan most closely associated with Ambedkar is taken from the labor movements in the United States and England: “educate, agitate, organize!” Regardless of whether we are persuaded by Ambedkar’s Buddhism, or his understanding of the role it can play in social life, he challenges us to think and practice ways of life that can nourish sympathy and benevolence across social differences.

Notes 1 I am grateful to Giulia Davis, Sarah Fleming, Richard Henning, Aishwary Kumar, Sara McClintock, Meg Mott, Valerian Rodrigues, Andy Rotman, Eva Seligman, and Sonia Sikka for helpful comments on a previous draft. 2 Ambedkar generally uses the term “untouchable” in his speeches and writings; he rarely uses “Dalit.” In the late nineteenth century, “Dalit” – from the Sanskrit root dal in the sense of “broken” – came to refer to groups excluded from the four Brahminic castes. While social reformers such as Savitribai and Jotirao Phule popularized the term in the nineteenth century, it was embraced by ex-untouchables across India in the twentieth century to affirm a positive cultural and political identity even as it simultaneously names the history and present of marginalization. This positive identity enacts a refusal of the condition of untouchability, with its subordination and exclusions, and thus contests the power of upper-caste and upper-class Indians across Indian society. In recent decades, some have favored the term “Bahujan,” or “Dalit-Bahujan.” “Bahujan” is intended to build coalitions between the most marginalized groups, such as Dalits and Adivasis, and the many other people in contemporary Indian political, juridical, and social life, including peasants, religious minorities, and others who find themselves subordinated to high-caste power. 3 As Valerian Rodrigues has argued, philosophical thinking permeates Ambedkar’s work, which addresses issues in metaphysics, philosophy of the person, epistemology, ethics, social and political thought, and philosophy of religion (Rodrigues 2017, 101). 4 A jāti is an endogamous group that traditionally defined social roles, occupations, identity, and status in Indian society. There are thousands of regional jātis in India in contrast to the four transregional varṇas in the Brahminic/Hindu social order: brahmins (priests), kshatriyas (warriors), vaishyas (merchants), and shudras (servants). Under the British, jātis were associated with particular varṇas, strengthening the varṇa system. Some jātis were regarded as outside the varṇa system and thus were outcastes and referred to, during Ambedkar’s time, as “untouchable.” Today, many scholars use “caste” to describe jātis and “class” to describe varṇas. 5 For more on Ambedkar’s life, see biographies by Jaffrelot (2005) and Omvedt (2004a) as well as his own notes for an unfinished autobiography, “Waiting for a Visa,” BAWS vol. 12, 661–91 6 This approach to Buddhism was also developed elsewhere in South and Southeast Asia, where Buddhism was formulated as a rational religion with a universal morality and therefore appropriate for modernity as part of a larger anticolonial movement (Edelglass 2020). 7 Ambedkar was not alone among twentieth-century Buddhist scholars and reformers either in arguing that the four Noble Truths were a later development or giving Buddhism a more explicitly engaged and political interpretation. Christopher Queen, for example, has shown how Ambedkar underlined his copy of Hermann Oldenberg’s The Buddha: His Life, His Doctrine, His Order, which also makes the argument that the four Noble Truths are not found in the earliest stages of Buddhism (Queen 2004, 140–41). 8 While the details of his argument may be different from classical Buddhist discussions, Ambedkar’s rejection of Brahmanical attempts to “naturalize” social status is consistent with much Indian Buddhist thought. See, for example, Vincent Eltschinger’s work on Buddhist critiques of realist interpretations of caste and social status from early canonical texts through Āryadeva, Vasubandhu, Dharmakīrti, Candrakīrti, Kumārila, and Prajñākaragupta (Eltschinger 2012). 9 For critiques of these readings, see Viswanathan 1996; Ganguly 2002; Gannon 2011. 10 After I wrote this chapter, Valerian Rodrigues kindly shared with me part of his forthcoming book on Ambedkar’s political philosophy. Rodrigues’s work, especially his chapter on “Religion and the

647

William Edelglass Modern Public,” addresses much of what I have discussed here in greater detail and with nuance and sophisticated analysis. 11 Religious conversion as a means for those outside the varṇa system to achieve social and political emancipation was not unique to Ambedkar; there had been such conversions to Christianity, Sikhism, Islam, and Buddhism in the nineteenth and early twentieth centuries. Ambedkar himself considered these other traditions as possibilities before committing himself to Buddhism. 12 “Ambedkar’s view that the creators of the Vedic literature should be viewed as a people rather than as a race was far ahead of his times. This view is winning approval in modern Indology” (Sharma 2005, 862). 13 www.saada.org/item/20140415-3544. Accessed January 15, 2019.

Bibliography Ambedkar, B. R. 1982–2003. Dr. Babasaheb Ambedkar Writings and Speeches (BAWS), vols. 1–17. Bombay: Government of Maharashtra, Department of Education. ———. 2011. The Buddha and His Dhamma: A Critical Edition. Edited, introduced, and annotated by Aakash Singh Rathore and Ajay Verma. New Delhi: Oxford University Press. Blackburn, Anne M. 1993. “Religion, Kinship and Buddhism: Ambedkar’s Vision of a Moral Community.” The Journal of the International Association of Buddhist Studies 16 (1): 1–23. Edelglass, William. 2020. “Aspiration, Conviction, and Serene Joy: Faith and Reason in Indian Buddhist Literature on the Path.” In Asian Philosophies and the Idea of Religion, edited by Sonia Sikka and Ashwani Kumar Peetush, 13–35. London: Routledge. Eltschinger, Vincent. 2012. Caste and Buddhist Philosophy: Continuity of Some Buddhist Arguments Against the Realist Interpretation of Social Denominations. Translated by Raynald Prévèreau in collaboration with the author. Delhi: Motilal Banarsidass Publishers. Fiske, Adele, and Christoph Emmrich. 2004. “The Use of Buddhist Scriptures in B. R. Ambedkar’s The Buddha and His Dhamma.” In Reconstructing the World: B. R. Ambedkar and Buddhism in India, edited by Surendra Jondhale and Johannes Beltz, 97–119. New Delhi: Oxford University Press. Fuchs, Martin. 2001. “A  Religion for Civil Society? Ambedkar’s Buddhism, the Dalit Issue and the Imagination of Emergent Possibilities.” In Charisma and Canon: Essays on the Religious History of the Indian Subcontinent, edited by Vasudha Dalmia, Angelika Malinar, and Martin Christof, 250–73. New Delhi: Oxford University Press. ———. 2004. “Buddhism and Dalitness: Dilemmas of Religious Emancipation.” In Reconstructing the World: B. R. Ambedkar and Buddhism in India, edited by Surendra Jondhale and Johannes Beltz, 283–300. New Delhi: Oxford University Press. Ganguly, Debjani. 2002. “History’s Implosions: A Benjaminian Reading of Ambedkar.” Journal of Narrative Theory 32 (3): 326–47. Gannon, Shane P. 2011. “Conversion as a Thematic Site: Academic Representations of Ambedkar’s Buddhist Turn.” Method and Theory in the Study of Religion 23 (1): 1–28. Gokhale, Pradeep P. 2004. “Universal Consequentialism: A Note on B.R. Ambedkar’s Reconstruction of Buddhism with Special Reference to Religion, Morality, and Spirituality.” In Reconstructing the World: B. R. Ambedkar and Buddhism in India, edited by Surendra Jondhale and Johannes Beltz, 120–31. New Delhi: Oxford University Press. Jaffrelot, Christophe. 2005. Dr. Ambedkar and Untouchability: Fighting the Indian Caste System. New York: Columbia University Press. Kumar, Aishwary. 2015. Radical Equality: Ambedkar, Gandhi, and the Risk of Democracy. Stanford: Stanford University Press. ———. 2018. “In the Void of Faith: Sunnyata, Sovereignty, Minority.” In Tolerance, Secularization and Democratic Politics in South Asia, edited by Humeira Iqtidar and Tanika Sarkar, 156–90. Cambridge: Cambridge University Press. Mukherjee, Arun P. 2009. “B. R. Ambedkar, John Dewey, and the Meaning of Democracy.” New Literary History 40 (2): 345–70. Omvedt, Gail. 2004a. Ambedkar: Towards an Enlightened India. Haryana: Penguin Books. ———. 2004b. “Confronting Brahmanic Hinduism: Dr. Ambedkar’s Sociology of Religion and Indian Society.” In Reconstructing the World: B. R. Ambedkar and Buddhism in India, edited by Surendra Jondhale and Johannes Beltz, 49–62. New Delhi: Oxford University Press.

648

B. R. Ambedkar Queen, Christopher. 1996. “Dr. Ambedkar and the Hermeneutics of Buddhist Liberation.” In Engaged Buddhism: Buddhist Liberation Movements in Asia, edited by Christopher S. Queen and Sallie B. King, 45–71. Albany: State University of New York Press. ———. 2004. “Ambedkar’s Dhamma: Source and Method in the Construction of Engaged Buddhism.” In Reconstructing the World: B. R. Ambedkar and Buddhism in India, edited by Surendra Jondhale and Johannes Beltz, 132–50. New Delhi: Oxford University Press. Rodrigues, Valerian. 2017. “Ambedkar as a Political Philosopher.” Economic & Political Weekly 52 (15): 101–7. Sharma, Arvind. 2005. “Dr. B. R. Ambedkar on the Aryan Invasion and the Emergence of the Caste System in India.” Journal of the American Academy of Religion 73 (3): 843–70. Skaria, Ajay. 2015. “Ambedkar, Marx and the Buddhist Question.” Journal of South Asian Studies 38 (3): 450–65. Stroud, Scott. 2018. “Creative Democracy, Communication, and the Uncharted Sources of Ambedkar’s Deweyan Pragmatism.” Education and Culture 34 (1): 61–80. Sumant, Yashwant. 2004. “Situating Religion in Ambedkar’s Political Discourse.” In Reconstructing the World: B. R. Ambedkar and Buddhism in India, edited by Surendra Jondhale and Johannes Beltz, 63–78. New Delhi: Oxford University Press. Surendran, Gitanjali. 2013. “ ‘The Indian Discovery of Buddhism’: Buddhist Revival in India, c. 1890– 1956.” PhD diss., Harvard University. Viswanathan, Gauri. 1996. “Religious Conversion and the Politics of Dissent.” In Conversion to Modernities: The Globalization of Christianity, edited by Peter van der Veer, 89–114. New York: Routledge.

649

40 THE DALAI LAMA XIV A Modern Indian Philosopher Jay L. Garfield

Introduction It may sound odd to some ears to represent the present Dalai Lama (the Ven. Tenzin Gyatso, 1935–) as an Indian philosopher. After all, he was born and educated in Tibet and is associated primarily with Tibetan Buddhist philosophy. But there are several good reasons to read him as an Indian philosopher, and these reasons also serve as a good introduction to his philosophical career. First, Tibetan Buddhist philosophy does represent a clear continuation of an Indian Buddhist tradition, and most of the reference points even for contemporary Tibetan philosophy – and particularly for the work of the Dalai Lama XIV – are Indian. That would hardly be sufficient, though; in that case, every Tibetan philosopher would count as Indian, and the distinctive character of Tibetan philosophy would be elided. But when we add several other reasons, the case becomes more compelling. The second reason is this: the Dalai Lama XIV has spent his entire adult life (and professional philosophical career) in India, after fleeing Tibet in 1959 amidst an uprising against Chinese occupation. And that residency in India has hardly been one of isolation within the Tibetan community; he has been in constant dialogue with Indian philosophical and religious figures and with the wider world. Third, as a contemporary Indian philosopher, one of his most important touchstones beyond the classical Indian Buddhist tradition has been Mohandas K. Gandhi, one of the most important public Indian philosophers of the twentieth century. Finally, the approach to philosophy taken by the Dalai Lama throughout his extensive corpus is utterly continuous with that of modern Indian philosophers. That is, he works self-consciously in engagement with a classical Indian tradition but also in self-conscious dialogue with contemporary philosophy derived from non-classical, non-Indian traditions, addressing philosophical problems by drawing freely from both (see Bhushan and Garfield 2017). The fact that this long-term resident of India works self-consciously in the context of recent Indian philosophy and uses a methodology consonant with that of his compatriots more than justifies reading the Dalai Lama XIV as an Indian philosopher. Indeed, the Dalai Lama XIV himself agrees with this characterization, writing in Beyond Religion: Ethics for the Entire World (2011): Sometimes I describe myself as a modern-day messenger of ancient Indian thought. Two of the most important ideas I  share whenever I  travel  – the principles of 650

DOI: 10.4324/9781351030908-57

The Dalai Lama XIV

nonviolence and interreligious harmony – are both drawn from ancient Indian heritage. Though I am of course a Tibetan, I also consider myself to be, in a sense, a son of India. Since childhood my mind has been nourished by the classics of Indian thought. . . . And since early adulthood my body, too, has been nourished by Indian fare: rice and dal. (11) The Dalai Lama XIV’s corpus is vast, comprising about 50 books and countless transcribed lectures, interviews, and teachings. It would be impossible to survey this entire body of work in a single chapter. Instead, I will discuss the outlines of his philosophical project and views and some of the respects in which he is philosophically distinctive. We can divide the Dalai Lama XIV’s published work into several broad (overlapping) genres. First, he has written substantial commentarial work on Indian Buddhist literature. Second, he has written systematic ethical treatises aimed at a global audience. Third, he has written on metaphysics and science. Fourth, he has written on the philosophy of religion, particularly on interfaith dialogue. Fifth, he has written religious homilies aimed sometimes at Buddhist audiences and sometimes at more general audiences, designed to help readers to lead happier lives. Finally, he has written autobiographical texts. Here, I will address the work comprising the first four of these rough categories.

General Philosophical Framework The overarching philosophical framework within which the Dalai Lama XIV works represents an intersection between several philosophical traditions. First and foremost, he is a follower of the Prāsaṅgika Madhyamaka tradition of Buddhism, a tradition grounded in Candrakīrti’s (seventh century CE) reading of Nāgārjuna’s (ca. second century CE) Fundamental Verses on the Middle Way (Mūlamadhyamakakārikā). Philosophers in this tradition take seriously the ultimate emptiness of all phenomena of any intrinsic nature in virtue of their interdependence. But at the same time, they take seriously the conventional reality of these interdependent phenomena. Since this is a Mahāyāna tradition, Madhyamaka moral philosophers ground ethical thought in the bodhisattva ideal – that is, in a commitment to attain awakening, conditioned by the cultivation of karuṇā, an attitude of universal care, for the benefit of all sentient beings. On the ethical side, the principal texts to which the Dalai Lama XIV refers are Śāntideva’s (eighthcentury) How to Lead an Awakened Life (Bodhicāryāvatāra) and Kamalaśīla’s (eighth-century) Stages of Meditation (Bhāvanākrama). The Dalai Lama XIV’s understanding of this tradition is mediated in several respects (even prior to the interaction with the other strains of thought that combine in his philosophical outlook). First, he reads this tradition through the extensive Tibetan commentarial tradition. Most specifically, he is influenced enormously by the commentaries and treatises of his own lineage, the dGe lugs (Geluk) school founded by rJe Tsongkhapa (1357–1419), a lineage that, more than most, emphasizes a robust realism about conventional reality and the possibility of genuine knowledge of conventional truth (Cowherds 2011; Jinpa 2019; Thakchöe 2007; Yakherds 2021). That influence is rendered less parochial, however, in virtue of the Dalai Lama XIV’s commitment to an ecumenical approach to Buddhist philosophy following the nineteenth- to twentieth-century Tibetan ris med, or non-sectarian movement, and he has received instruction from many scholars belonging to lineages other than the dGe lugs tradition, often commenting on texts from the perspective of these other lineages. Moreover, the commitment to the Prāsaṅgika Madhyamaka school is mitigated by the Dalai Lama XIV’s deep interest in 651

Jay L. Garfield

and sympathy with other strands of Buddhist thought, including that of the Theravāda and Yogācāra traditions. His own Madhyamaka thought often incorporates ideas from these other schools, something unusual in Tibetan scholarship but reflecting the kind of cosmopolitan approach to philosophy characteristic of Indian philosophers. This cosmopolitan sensibility also informs the Dalai Lama XIV’s philosophy of religion, in particular his religious pluralism. I have already noted his intra-Buddhist ecumenism deriving from the Tibetan non-sectarian movement, a tendency amplified by his constant dialogue with Buddhists of other traditions. But like Gandhi, and Ramakrishna before him, the Dalai Lama XIV has been a constant advocate for interfaith dialogue and for pluralism in religious practice and commitment. Given the historically missionary nature of Buddhism (a tendency still very much in evidence), this is a remarkable posture. Not only has the Dalai Lama XIV urged that all religions share common core commitments and that there is no single religious view appropriate to all religions and cultures, but he has actually offered commentary on the Christian gospels. This pluralist ecumenism is another respect in which his modernism and cosmopolitanism are very much in evidence. The second line of influence to note is that from Mohandas K. Gandhi, in particular Gandhi’s emphasis on ahiṃsā, or avoidance of harm. Although the Dalai Lama XIV relies principally on Buddhist teachings, while Gandhi draws inspiration from a combination of Hindu and Western pacifist teachings (Bhushan and Garfield 2017), their views converge on this point, as well as in the conviction that ahiṃsā is first and foremost an inner state, or commitment demanding self-cultivation, and only secondarily an external recusal from violent action. The Dalai Lama XIV draws heavily on Gandhi’s theorization of non-harm both in his own ethical theory and in his approach to the political campaign for greater Tibetan autonomy within the People’s Republic of China (see especially Beyond Religion, 9). Gandhi’s political thought, as many have noted, has a complex relationship to modernity. On the one hand, his theological outlook and his hostility towards technology, urbanization, and central government – ideas inspired by Thoreau, Ruskin, and Tolstoy – contribute to a trenchant critique of modernity, most explicitly in Hind Swaraj. On the other hand, Gandhi draws heavily on a discourse of rights and rule of law that depends on ideals of European modernity, arguing for Indian independence, for instance, on the grounds that colonial rule violates fundamental rights. The Dalai Lama XIV is more friendly to modern science, technology, and government but follows Gandhi on this latter route, developing a political and ethical theory that takes rights seriously in a way unprecedented in Buddhist thought. This has allowed him to become a potent actor on the world political stage and has allowed him to bring Buddhist ethical reflection into dialogue with global modernity. The Dalai Lama XIV’s modernism is also evident in his robust engagement with science and in his scientific realism. He has a long-standing interest in physics, in the biological sciences, and in cognitive science in particular. In keeping with the realistic spirit of the Geluk understanding of conventional reality, he has taken modern science to provide the best understanding of the nature of that reality and has always insisted that where science and Buddhist doctrine conflict, science wins. (Whether he always follows that dictum faithfully is a matter of some controversy, but I leave that aside here.) Through a series of dialogues among senior Tibetan scholars and scientists in which he is a regular participant, organized by the Mind and Life Foundation, and in a series of influential books inspired by those dialogues, the Dalai Lama XIV has argued that Buddhist philosophy entails a respect for science and that modern science confirms many of the most central tenets of Buddhist philosophy. He has also been instrumental in revising monastic curricula, incorporating extensive education in modern science.

652

The Dalai Lama XIV

Buddhist Philosophy Not surprisingly, a large portion of the Dalai Lama XIV’s corpus is devoted specifically to topics in Buddhist philosophy. The majority of that consists in commentarial exposition of Indian Madhyamaka texts; the remainder are independent treatises addressed predominately to popular audiences. The Dalai Lama XIV has commented on sūtras, texts addressing the graduated path to awakening, and tantric texts. These commentaries include: The Essence of the Heart Sūtra (2005); The Path to Enlightenment (1994a); Tantra in Tibet (2007); A Flash of Lightning in the Dark of Night (1994b); Sleeping, Dreaming, and Dying (1997); Dzogchen: The Heart Essence of the Great Perfection (2000); For the Benefit of All Beings: A  Commentary on The Way of the Bodhisattva (2009); and Practicing Wisdom: The Perfection of Shantideva’s Bodhisattva Way (2014). In most of these texts, his approach eschews traditional line-by-line commentary as the primary mode of explanation (although there is some of this). Instead, he favors the reconstruction of a broader philosophical picture in which fundamental metaphysical and epistemological ideas are first defended and then used as premises on the basis of which an account of cultivation and of the good life, or of salutary meditative practices, is defended. These commentaries are very much addressed to a Buddhist audience interested in advice regarding spiritual practice or illumination of difficult texts in Buddhist philosophy. The Dalai Lama XIV has returned to How to Lead an Awakened Life – a text he often teaches in public – many times. In his discussion of this text, unlike some others, the Dalai Lama XIV is self-consciously addressing both a Buddhist and a non-Buddhist audience, treating How to Lead an Awakened Life both as a cultivation text within the Buddhist tradition and as a secular text, defending an approach to ethical cultivation and conduct for everyone from premises that require no religious commitment. His books on this text include Transcendent Wisdom (1988), A Flash of Lightning in the Dark of Night (1994b), Practicing Wisdom (2004), A Profound Mind: Cultivating Wisdom in Everyday Life (2011), For the Benefit of All Beings: A Commentary on The Way of the Bodhisattva (2009), Practicing Wisdom: The Perfection of Shantideva’s Bodhisattva Way (2014), and Perfecting Patience (2018). While some of these are traditional verse-by-verse commentaries on the text, others are more discursive, focusing on a single chapter or set of verses, explaining their meaning in the text, their context in Madh­ yamaka ethical thought generally, and their importance in moral cultivation and daily life. In each case, the Dalai Lama XIV works to ground ethical ideas in metaphysical and psychological claims about the nature of the person, following the lead of Śāntideva and his predecessor Āryadeva (third century CE). And in each case, the principal link to which he draws attention is that between interdependence ( pratītyasamutpāda), the fundamentally social nature of human beings, and care (karuṇā), arguing that a true understanding of the nature of reality and of human life spontaneously gives rise to an attitude of care.

Religious Pluralism One might expect the leader of a major branch of a traditionally missionary religious movement to be a religious monist, to believe that the religious tradition that he leads is the one true religion. This is indeed the case regarding most religious leaders of this stature, reflecting the truth-claims religions make. There have been notable exceptions to this stance in the history of Indian philosophy: Jainism is pluralistic, and the nineteenth- to twentieth-century Vedānta sage Ramakrishna was a pluralist as well. The Dalai Lama XIV follows in this Indian religious tradition, advocating not only tolerance of other religious beliefs but also genuine pluralism

653

Jay L. Garfield

regarding them. He has argued that each religion can legitimately be regarded as true by its followers, that all religions have a common ethical core, that the metaphysical differences among religions are of no great import, and that different religions are best for different individuals, just as different medicines are appropriate to different patients and yet all aim at a cure. The Dalai Lama XIV has brought this pluralist spirit to interfaith dialogue, including notable dialogues with the Catholic monk Thomas Merton, with a group of Jewish leaders, and with the Anglican Archbishop Desmond Tutu. He has also defended and expressed these views in several notable books. In The Good Heart: A Buddhist Perspective on the Teachings of Jesus (1996), a volume arising from a teaching residency in a Benedictine monastery in England, the Dalai Lama XIV offers commentary on the Christian Gospels. In Beyond Religion: Ethics for a Whole World, he argues that the values advanced by religions are equally well – or even more effectively – advanced in a secular context, radicalizing his vision of religious pluralism to include irreligion as an option he also endorses. In The Good Heart, the Dalai Lama XIV introduces his project by saying I believe the purpose of all the major religious traditions is not to construct big temples on the outside, but to create temples of goodness and compassion inside, in our hearts. Every major religion has the potential to create this. The greater our awareness is regarding the value and effectiveness of other religious traditions, then the deeper will be our respect and reverence toward other religions. This is the proper way for us to promote genuine compassion and a spirit of harmony among the religions of the world. (38–39) He reinforces this strong pluralism a few pages later: People often experience feelings of exclusivity in their religious beliefs – a feeling that one’s own path is the only true path – which can create a sense of apprehension about connecting with others of different faiths. I believe the best way to counter that force is to experience the value of one’s own path through a meditative life, which will enable one to see the value and preciousness of other traditions. (41) This does not amount to a universalism, according to which all religions are taken to be basically the same, albeit articulated in different ways, but a genuine pluralism, respecting the real differences among religions, as the Dalai Lama XIV immediately emphasizes: I have always felt that we should have different religious traditions because human beings possess so many different mental dispositions: one religion simply cannot satisfy the needs of such a variety of people. If we try to unify the faiths of the world into one religion, we will also lose many of the qualities and richnesses of each particular tradition. (Ibid.) He makes the same point in the introduction to his exposition of the Heart of Wisdom Sūtra (2005, 9–13) in a discussion replete with genuine appreciation for the benefits of views foreign to Buddhism. He concludes that discussion with an admonition against proselytizing and in

654

The Dalai Lama XIV

favor of maintaining faith in one’s own tradition (13–14). He then characterizes productive interfaith dialogue as follows: At the beginning of such a dialogue, it’s important that all the participants fully recognize not only the many areas of convergence between each other’s faith traditions, but more crucially, that they recognize and respect the differences between the traditions. Furthermore, we should look at the specific causes and conditions that gave rise to the differences between the traditions. . . . Then, having clarified the differences and the origins, we look at the religions from a different perspective: becoming aware of how different religious philosophies and practices can give rise to similar results. By entering into interfaith dialogue in this way, we develop genuine respect and admiration for each other’s religious traditions. (14–15) The Dalai Lama XIV argues that the benefits from interfaith dialogue go well beyond the achievement of greater understanding of and respect for other religious traditions; he argues that traditions can actually benefit from lessons learned from other faiths: Though I don’t recommend that a person abandon his or her native religion, I believe that a follower of one tradition can certainly incorporate into his or her own spiritual practice certain methods for spiritual transformation found in other traditions. For example, some of my Christian friends, while remaining deeply committed to their own tradition, incorporate ancient Indian methods for cultivating single-pointedness of mind through meditative concentration. . . . These devout Christians, while remaining deeply committed to their own spiritual tradition, embrace aspects and methods from other teachings. This, I think is beneficial to them and wise. Buddhists can incorporate elements of the Christian tradition into their practice – for instance, the tradition of community service. In the Christian tradition, monks and nuns have a long history of social work. . . . In providing service to the greater human community through social work, Buddhism lags far behind Christianity. (17–18) In The Compassionate Life, he devotes an entire chapter (chapter 4) to an exploration of religious pluralism, developing these ideas in greater detail, and arguing directly for a serious religious pluralism that embraces difference to the benefit of all traditions. Let us now consider one example of the Dalai Lama XIV’s approach to a Christian scripture, drawn from his commentary on the beatitudes in the Sermon on the Mount. Here we see both his insistence on recognizing difference but also his genuine interest both in preserving that difference and in utilizing it to gain deeper insight into commonality. After pointing out that both Buddhist and Christian philosophers see the problem of suffering as structuring human life and that both affirm a close connection between the moral valence of one’s actions and the nature of their consequences for one’s future life, he notes: One of the most difficult concepts involved here, especially for Buddhists, is the concept of a divine being, God. Of course, one can approach this concept in terms of something which is inexpressible, something which is beyond language and

655

Jay L. Garfield

conceptuality. But one must admit that, at the theoretical level, the conceptions of God and Creation are a point of departure between Buddhists and Christians. (55) The Dalai Lama XIV then uses this “point of departure” to explore the motivations for theism and a theory of universal creation by an omniscient, omnipotent, omnibenevolent god in Christianity and the motivations for rejecting such an account in Buddhism, concluding that there are good arguments on each side. The discussion is a model of judicious pluralism in the philosophy of religion. He returns to the question of theism later in the text, explicitly grounding this pluralism in Buddhist philosophy: All [of the Buddha’s various teachings] are aimed toward sentient beings’ diverse mental dispositions, needs, and spiritual inclinations. And when I understand the truth of this, I am able to truly appreciate the richness and value of other traditions, because it enables me to extend the same principle of diversity to other traditions as well . . . Within Christianity there seem to be a number of diverse interpretations or understandings of the concept of God.  .  .  .  [Q]ualities such as compassion can also be attributed to [a] divine ground of being. Now if we are to understand God in such terms – as an ultimate ground of being – then it becomes possible to draw parallels with certain elements in Buddhist thought and practice . . . We should also be careful not to reduce everything to a set of common terms so that at the end of the day we have nothing left . . . that is distinct about our specific traditions. . . . For example, if one were to try hard to draw parallels between Buddhism and the idea of the Trinity, the first thing that might come to mind would be the idea of the three kayas, the doctrine of the three embodiments of the Buddha: dharmakāya, sambhogakāya and nirmāṇakāya. But . . . we should not push these lines of comparison too far. (72–73) We could say much more. But it should be clear that the Dalai Lama XIV is approaching the philosophy of religion through a sophisticated, cosmopolitan pluralism grounded in Buddhist philosophy, that is, as a true modernist.

Ethics The Dalai Lama XIV has written a good deal on ethics, and indeed this may be the area in which he has made the most substantive – as well as the most public – contributions to philosophy. It is also an area in which his modernism and in which his connections to Western and to Indian philosophy are particularly evident. As I noted in the previous section, the Dalai Lama XIV’s ethical thought is firmly rooted in Indian Mahāyāna Buddhist thought, particularly that of Śāntideva. As a consequence, his orientation to ethical thought is structured by several themes that run through that tradition and that are emphasized by Śāntideva. The first is that he conceives of ethics as the cultivation of modes of perception and experience and not in the first instance in terms of conduct or duties. A second is that he thinks about ethics in terms of a path to cultivation of these modes of perception and so in terms of the techniques one can use to advance ethically. Third, he conceives of the qualities one cultivates on that path in terms of the six perfections adumbrated in the context of the bodhisattva stages (generosity, attention,

656

The Dalai Lama XIV

patience, enthusiasm, mediation, and wisdom). Fourth, he conceives the goal of the path as the attainment of the four divine states (brahmavihāras) that are common to most Buddhist ethical thought – friendliness or beneficence (maitrī), care (karuṇā), sympathetic joy (muditā), and impartiality (upekṣā) – with care taking center stage. On the other hand, there are aspects of the Dalai Lama XIV’s ethical thought that come from other sources. For instance, his secularization of his arguments for moral cultivation and moral education lead him to justify his ethical vision not in terms of the goal of awakening but rather in a reasoned account of human nature and human goals. He argues that taking human nature and human goals seriously entails adopting this moral perspective. For instance, in Ethics for the New Millennium (1999), he writes: My call for a spiritual revolution is thus not a call for a religious revolution. Nor is it a reference to a way of life that is somehow otherworldly, still less to something magical or mysterious. Rather, it is a call for a radical reorientation away from our habitual preoccupation with self. It is a call to turn toward the wider community of beings with whom we are connected, and for conduct that recognizes others’ interests alongside our own. (23–24) He amplifies this idea both throughout Ethics for a New Millennium and more recently and with greater emphasis on the secular basis of his ethical thought in Beyond Religion: Ethics for the Whole World (2011), where he cites the European Enlightenment and the French Revolution explicitly (7). This approach reflects his ecumenism as well as his commitment to a universal standard of rationality and to argument based on shared premises, ideas that come both from the Indian philosophical tradition and from European modernity. There are two other distinctive features of the Dalai Lama XIV’s approach to ethics. First, there is his commitment to human rights. Second, there is his appropriation of Gandhi’s commitment to non-harm and a representational approach to action which reflects his debt to modern Indian moral and political theory. The Dalai Lama XIV’s ethical thought is developed in several books, prominently including The Power of Compassion (1995), Ethics for the New Millennium (1999), The Compassionate Life (2003), The Art of Happiness at Work (2003), The Wisdom of Forgiveness (2004), and Kindness, Clarity, and Insight (1984). Let us begin with the Buddhist roots of the Dalai Lama XIV’s ethics. Śāntideva’s How to Lead an Awakened Life is a systematic development of a Buddhist moral phenomenology, characterizing vice and suffering as depending on egocentricity grounded in a primal fear of death (Cowherds 2016; Garfield 2010/2011, 2019; Heim 2021). The path to fearlessness and moral maturity involves the accumulation of the six perfections of generosity, attention, patience, enthusiasm, meditation, and wisdom. Each of these is developed by Śāntideva and articulated by the Dalai Lama XIV as a way of experiencing oneself, the world, and other moral agents, and the progressive cultivation of these perfections leads to a more accurate, salutary, and effective mode of engagement with the world. One is transformed from an agent who self-reifies, locates oneself at the center of the moral universe, sees others as mere objects, and as a consequence takes egoism to be a rational mode of engagement into an agent who experiences herself as an impermanent interdependent participant in a vast network of beings with whom one is co-constituted, and so one who takes the interests of all to be equally motivating. In his own work, the Dalai Lama XIV places special emphasis on the cultivation of patience as an antidote to anger and aversion and the meditation on emptiness and interdependence as an antidote to egoism.

657

Jay L. Garfield

In each case, the Dalai Lama XIV argues directly that these are the only rational comportments towards the world. Anger and egoism presuppose an incoherent model of our existence as entirely independent of that of others and of agency as involving autonomy that ignores our casual dependence on our history and our environment. Egoism involves the absurd notion that I am far more important morally than anyone else. He argues that any rational understanding of the world reveals instead that people act due to countless causes and conditions for which they are not responsible, and so anger is never justified. And we are no different from others, so our own interests have no special weight in decision making. The outcome of this path of cultivation is the achievement of an understanding of ourselves as embedded, conventionally real, but ultimately empty persons, persons whose goals can only be achieved through cooperation and who are happier to the degree that they are enmeshed in bonds of friendship than if they are in competitive strife. This understanding of the world, the Dalai Lama XIV argues, is not simply that articulated in Buddhist philosophy but is supported by modern science, including the life sciences and behavioral sciences. As he puts it in Ethics for a New Millennium: [w]hen we come to see that everything we perceive and experience arises as a result of an indefinite series of interrelated causes and conditions, our whole perspective changes. We begin to see that the universe we inhabit can be understood in terms of a living organism where each cell works in balanced cooperation with every other cell to sustain the whole. (40–41) In Beyond Religion, he writes that there is now a reasonably substantial body of evidence in evolutionary biology, neuroscience, and other fields suggesting that, even from the most rigorous scientific perspective, unselfishness and concern for others are not only in our own interests but also, in a sense, innate to our biological nature. (5–6) Understanding oneself and the world in this way, he argues, leads to a spontaneous attitude of friendship to those with whom we share the world and upon whom our existence is dependent; a care for others, whose suffering we recognize as immediately motivating; happiness at success, no matter whose it is; and a sense that nobody’s value depends upon their relation to us in particular. These are the four brahmavihāras, valorized in Buddhist ethics, and argued by the Dalai Lama XIV to be the only rational attitudes towards ourselves and others. He also develops this argument in The Compassionate Life (2003, 8–10). This secular defense of his approach to ethics is one respect in which the Dalai Lama XIV is developing a modernist version of Buddhist ethics. Another is his commitment to human rights – an idea with no antecedents in Buddhist or classical Indian thought – and to an international order that guarantees them as a moral imperative. While this commitment, grounded in Western political and moral theory in a commitment to individual autonomy and a deontological ethical framework, might appear to be in tension with a Buddhist ethical framework committed to interdependence and phenomenological ethics, the Dalai Lama XIV sees the framework of human rights as having a solid foundation in his Buddhist ethical framework.

658

The Dalai Lama XIV

It is, in his view, introduced as a means to extending care to others, not as an expression of primary autonomy (see also Garfield 1998). The Dalai Lama XIV puts it this way: A sense of responsibility toward all others also means that, both as individuals and as a society of individuals, we have a duty to care for each member of our society. . . . Just like ourselves, . . . people have a right to happiness and to avoid suffering. (169) And in Beyond Religion: Ethics for the Whole World (2011), he writes: We have a universal declaration of human rights, and awareness of the importance of such rights has grown tremendously. As a result, the ideals of freedom and democracy have spread around the world, and there is increasing recognition of the oneness of humanity. (ix) Understood this way, we ascribe rights to others as a way of ensuring public policies and interpersonal behavior that supports the extension of benefits to them, prevents harm, and recognizes human equality, each manifestations of the brahmavihāras. The Dalia Lama XIV makes much the same point in Kindness, Clarity, and Insight (70 ff.) and in Perfecting Patience (2018), where he writes, “one aspect of compassion is to respect others’ rights and others’ views” (6). Mohandas K. Gandhi argued that the core of moral engagement lies in the conjunction of the commitment to avoid harm, and the achievement of self-mastery (swaraj). He also insists that public moral engagement involves the practice of representational action to demonstrate the moral status of unjust policies (satyagraha, or insistence on the truth). A practitioner of satyagraha not only behaves in a way that comports with ethical standards but does so in a public way to challenge injustice and does so nonviolently. This practice hence not only constitutes but also represents in a public forum that in which ethical conduct consists and the injustice of the institution against which satyagraha is launched. This adds an important social dimension to ethics, one missing in much of traditional Buddhist ethical thought, which focuses more on the individual perfection of one who has renounced society. The Dalai Lama XIV, as I have noted, was deeply influenced by Gandhi’s life and writings. This leads him to emphasize the social face of ethics and the importance to ethical practice of the avoidance of harm. A recognition of the importance of social engagement has led the Dalai Lama XIV to become a leader – along with the Vietnamese Zen monk the Ven. Thich Nhat Hanh and the Thai lay Buddhist activist Sulak Sivaraksa – in the Engaged Buddhism movement, a movement that has translated Buddhist ethics into a program of social reform (Hanh 1987; Queen 2000; Queen and King 1996). Buddhist ethics in this tradition positively enjoins social action on behalf of the poor and the oppressed, on behalf of social justice, and for the benefit of the environment. Engaged Buddhists reinterpret the basic Buddhist precepts, which are all expressed in negative terms – enjoining refraining from killing; from harsh, idle, or deceptive speech; refraining from intoxication; refraining from theft; refraining from sexual misconduct – in positive terms. They come to be understood as enjoining the protection of life; the use of speech for social good (including representational action); cultivating awareness of social wrongs; working actively for the redistribution of wealth and income; and protection of the vulnerable against exploitation. The Dalai Lama XIV has actively promoted this understanding of Buddhist ethics, once again confirming both his Indian sensibility and his modernist approach to ethical theory.

659

Jay L. Garfield

Engagement with Science The Dalai Lama XIV’s modernist approach to Buddhist philosophy has also been amply in evidence in his approach to science. He has been committed to taking modern science seriously for some time. The primary vehicle for his engagement with science has been the long series of dialogues sponsored by the Mind and Life Institute beginning in 1987 and continuing to the present day. These dialogues bring together groups of senior monastic scholars, including the Dalai Lama XIV, with Western natural and social scientists and a few humanities scholars to provide context. These dialogues have brought the Dalai Lama XIV into close interaction with physicists, biologists, neuroscientists, and psychologists. These conversations are not idle exchanges of interest but serious explorations of the relevance of modern science to Buddhist philosophy and doctrine and of Buddhism to problems in modern science. Many have later been distilled into books summarizing the interactions (Begley 2007; Dalai Lama 1997; Davidson and Harrington 2002; Goleman 1990, 2003; Harrington and Zajonc 2006; Hayward and Varela 1992; Houshmand, Livingston, and Wallace 1999; Kabat-Zinn and Davidson 2012; Louisi and Houshman 2009; Singer and Ricard 2015; Zajonc 2004). The Dalai Lama XIV argues that science is a partner to Buddhist philosophy, and Madhyamaka philosophy in particular, at both the conventional and the ultimate levels. Following the Geluk tradition, he is committed to the view that there is a robust sense in which conventional truth is truth and that the conventional world is real, even if only conventionally. This entails that there is a determinate truth about how things are conventionally, even if they are ultimately empty of any intrinsic nature. Moreover, in this view, all phenomena are dependently originated and so can be explained causally and have their identities only in a network of causal, mereological, and conceptual interdependence. As a consequence, an analytical investigation into the composition of natural phenomena, and into their causal relations and theoretical contexts, should lead to a deeper understanding of their conventional nature. The Dalai Lama XIV is convinced that modern science provides the best possible avenue to such analytical investigation and so into the deepest understanding of conventional truth. But this has implications for ultimate truth as well. Since, in this tradition, the two truths are regarded as extensionally identical but intensionally distinct, with emptiness taken to be identical with interdependence, any investigation of interdependence is also an investigation into the nature of emptiness, revealing just how and why things lack intrinsic identity. Moreover, the demonstration by scientific exploration that everything in the world is in fact interdependent in these ways, and so lacking in any essence, reinforces the Madhyamaka claim that all phenomena are empty of intrinsic nature. In the domain of physics, this dialogue has generally focused on three implications of quantum mechanics and general relativity. The first is thoroughgoing interdependence. This is suggested by phenomena such as quantum entanglement. The second is lack of determinate nature suggested by the probabilistic character of quantum measurement and the uncertainty principle. The third is the mind-dependence of physical reality, suggested by the dependence of the collapse of probabilistic wave functions on measurement. All of these ideas are homologous with central doctrines in Prāsaṅgika Madhyamaka metaphysics. In the life sciences, the Dalai Lama XIV has emphasized the interdependence of living things as demonstrated by ecology and the close connection between mental and physical phenomena. Neuroscience, he argues, provides deep evidence of the efficacy of meditative practice and of the link between the cultivation of awareness of one’s own cognitive states and of the nature of mental processes and the development of salutary moral properties such 660

The Dalai Lama XIV

as care, friendliness, and lack of egocentricity. He has argued that cognitive neuroscience and Buddhist meditative practices are a natural pairing as tools for understanding fundamental conscious processes and affective response. He has been particularly interested in the neuroscience of affect and has argued that results in this area demonstrate both the natural arising of spontaneous affect as a consequence of deep insight into the mind and the benefits of positive affect for other domains of human functioning and health. This extended philosophical engagement with science, and the epistemic authority the Dalai Lama XIV grants it, is yet another mark of his modernism.

Conclusion The Dalai Lama XIV is one of the most influential contemporary Indian philosophers. He draws effectively on the classical Indo-Tibetan Buddhist tradition and on modern Indian and European philosophy. The synthesis he achieves in metaphysics, the philosophy of religion, ethics, and political philosophy represents a distinctive voice in contemporary philosophy: a classically grounded modernism that brings Buddhist thought into dialogue with modern science and philosophy.

Bibliography Begley, Sharon. 2007. Train Your Mind, Change Your Brain: How a New Science Reveals Our Extraordinary Potential to Transform Ourselves. New York: Ballantine Books. Bhushan, N. and J. L. Garfield. 2017. Minds Without Fear: Philosophy in the Indian Renaissance. New York: Oxford University Press. Cowherds. 2011. Moonshadows: Conventional Truth in Buddhist Philosophy. New York: Oxford University Press. _______. 2016. Moonpaths: Ethics and Emptiness. New York: Oxford University Press. Dalai Lama XIV. 1984. Kindness, Clarity, and Insight. Boston: Snow Lion Publications. ———. 1988. Transcendent Wisdom. Ithaca: Snow Lion Publications. ———. 1994a. The Path to Enlightenment. Ithaca: Snow Lion Publications. ———. 1994b. A Flash of Lightning in the Dark of Night: A Guide to the Bodhisattva’s Way of Life. Boston: Shambhala Publications. ———. 1995. The Power of Compassion. New York: HarperCollins. ———. 1996. The Good Heart: A Buddhist Perspective on the Teachings of Jesus. Somerville: Wisdom Publications. ———. 1997. Sleeping, Dreaming, and Dying: An Exploration of Consciousness. Edited by Francisco Varela. Somerville: Wisdom Publications. ______. 1999. Ethics for the New Millennium. New York: Riverhead Books. ———. 2000. Dzogchen: Heart Essence of the Great Perfection. Ithaca: Snow Lion Publications. ———. 2003. The Compassionate Life. Somerville: Wisdom Publications. ———. 2005. Essence of the Heart Sūtra: The Dalai Lama’s Heart of Wisdom Teachings. Somerville: Wisdom Publications. ———. 2009.  For the Benefit of All Beings: A  Commentary on the Way of the Bodhisattva. Boston: Shambhala Publications. ———. 2011a. Beyond Religion: Ethics for a Whole World. Boston: Houghton Mifflin Harcourt. ———. 2011b. A Profound Mind: Cultivating Wisdom in Everyday Life. New York: Harmony Books. ————. 2014. Practicing Wisdom: The Perfection of Shantideva’s Bodhisattva Way. Somerville: Wisdom Publications. ——. 2018.  Perfecting Patience: Buddhist Techniques to Overcome Anger. Boston: Shambhala Publications. Dalai Lama XIV, and Victor Chan. 2004. The Wisdom of Forgiveness. New York: Riverhead Books. Dalai Lama XIV, and Howard C. Cutler. 2003. The Art of Happiness at Work. New York: Riverhead Books. Dalai Lama XIV, and J. Hopkins. 2007. Tantra in Tibet. Ithaca, NY: Snow Lion.

661

Jay L. Garfield Davidson, Richard J., and Anne Harrington, eds. 2002. Visions of Compassion: Western Scientists and Tibetan Buddhists Examine Human Nature. New York: Oxford University Press. Garfield, Jay L. 1998. “Human Rights and Compassion: Towards a Unified Moral Framework.” In Buddhism and Human Rights, edited by Damien V. Keown, Charles S. Prebish, and Wayne R. Husted, 111–40. Richmond, Surrey: Curzon Press. ———. 2010/2011. “What Is It Like to Be a Bodhisattva?” Journal of the International Association of Buddhist Studies 33 (1–2): 327–51. ———. 2019. “Seeing Sentient Beings: Moral Phenomenology in Śāntideva’s Bodhicāryāvatāra.” In Readings of Śāntideva’s Guide to Bodhisattva Practices, edited by Douglas Duckworth and Jonathan Gold, 192–208. New York: Columbia University Press. Goleman, Daniel. 1990. “The Psychology of Meditation.” In Proceedings of the First International Conference on Psychotherapy, Meditation and Health, edited by M. G. T. Kwee, 19–35. London: EastWest Publications. ———, ed. 2003. Healing Emotions: Conversations with the Dalai Lama on Mindfulness, Emotions, and Health. Boston: Shambhala Publications. Hanh, Thich Nhat. 1987. Interbeing: Fourteen Guidelines for Engaged Buddhism. Edited by F. Eppsteiner. Berkeley: Parallax Press. Harrington, Anne, and Arthur Zajonc, eds. 2006. The Dalai Lama at MIT. Cambridge: Harvard University Press. Hayward, Jeremy W., and Francisco J. Varela, eds. 1992. Gentle Bridges: Conversations with the Dalai Lama on the Sciences of Mind. Boston: Shambhala Publications. Heim, Maria. 2021. Buddhist Ethics. New York: Cambridge University Press. Houshmand, Zara, Robert B. Livingston, and B. Alan Wallace, eds. 1999. Consciousness at the Crossroads: Conversations with the Dalai Lama on Brain Science and Buddhism. Boston: Snow Lion Publications. Kabat-Zinn, Jon, and Richard J. Davidson, eds. 2012. The Mind’s Own Physician: A Scientific Dialogue with the Dalai Lama on the Healing Power of Meditation. Oakland: New Harbinger Publications. Kamalaśīla, and Dalai Lama XIV. 2019. The Stages of Meditation: The Buddhist Classic on Training the Mind. Boston: Shambhala. Langritangpa. 2006. Eight Verses for Training the Mind. Translated by R. Sonam and Geshe Sonam Rinchen. Ithaca: Snow Lion. Luisi, Pier L. 2019. Mind and Life: Discussions with the Dalai Lama on the Nature of Reality. Edited by Zara Houshmand. New York: Columbia University Press. Nāgārjuna. 1995. The Fundamental Wisdom of the Middle Way: Nāgārjuna’s Mūlamadhyamakakārikā. Translated by Jay L. Garfield. New York: Oxford University Press. Queen, Christopher S., ed. 2000. Engaged Buddhism in the West. Somerville: Wisdom Publications. Queen, Christopher S., and Sallie B. King, eds. 1996. Engaged Buddhism: Buddhist Liberation Movements in Asia. Albany: State University of New York Press. Śāntideva. 1995. Bodhicaryavatara (How to Lead an Awakened Life). Translated by Kate Crosby and Andrew Skilton. New York: Oxford University Press. Singer, Tania, and Matthieu Ricard, eds. 2015. Caring Economics: Conversations on Altruism and Compassion, Between Scientists, Economists, and the Dalai Lama. New York: Picador. Thakchöe, Sonam. 2007. The Two Truths Debate: Tsongkhapa and Gorampa on the Middle Way. Boston: Wisdom Publications. Tsong-ka-pa. 1977. Tantra in Tibet: The Great Exposition of Secret Mantra. Translated by Jeffrey Hopkins. London: George Allen & Unwin. Yakherds. 2021. Knowing Illusion: Bringing a Tibetan Debate into Contemporary Discourse. New York: Oxford University Press. Zajonc, Arthur. 2004. The New Physics and Cosmology: Dialogues with the Dalai Lama. New York: Oxford University Press.

662

INDEX

Abhayākargupta 615–29; analysis of two truths 620–2; on buddha-nature 625–7; critique of Yogācāra as realist 622; on inference 623–4; influences on 618; on mind-only 622–3; Munimatālaṃkāra (Adornment of the Sage’s Thought) 482, 615, 618–27; nature of the mind 623–5; non-tantric works 615–16; on perception 623; preserver of rare passages 617–18; reliance on Kamalaśīla 482, 523, 615, 618, 622, 623, 625, 627; reliance on Ratnākaraśānti 617, 618, 626, 627; on scriptural authority 619–20; tantric works 56, 616–17 Abhidhamma 143–59, 160–71; Abhidhammatthasaṅgaha (Compendium on the Meaning of the Abhidhamma) 149–52; Buddhaghosa on 329–30, 331; compared to Sarvāstivāda Abhidharma 149, 166–8, 172–84; Dhammasaṅgaṇī (Enumeration of Phenomena) 143, 146–8, 189; discovered and taught by the Buddha 144; on emptiness 153; as endless phenomenological analysis 152–6, 338; in the Milindapañha (Questions of Milinda) 107; as ontology of ultimate reality 149–52, 154, 329, 342; Paṭisambhidāmagga (The Path of Discrimination) 153, 157n37, 160, 330; Vibhaṅga (Analysis) 148–9, 151, 152; see also Kathāvatthu (Points of Discussion) Abhidharma 172–87, 188–203; Abhidharmasamuccayakārikā (Stanzas of the Compendium of Abhidharma) 191, 194, 196–7, 198–9; Candrakīrti’s use of 408–12; criticized by Bhāviveka 349–50, 351; criticized by Nāgārjuna 212, 213; as literature of sectarian Buddhism 173; Mahāsāṃghika

tradition 211; as phenomenological investigation into entities (dharmas) 409; ontological teachings of the Sarvāstivāda 180; Sarvāstivāda tradition compared to Pali Abhidhamma 149; Sthiramati on 379–80; two models for meta-cognition (svasaṃvedana) in 368; see also Mahāvibhāṣā (Large Commentary); Vātsīputrīya/Sāṃmitīya adarśana (non-seeing): Bhāviveka on 354; Jñānagarbha on 453; Madhyamaka thinkers on 415; Śāntarakṣita on 453, 456; Yogācāra versus Madhyamaka interpretation of 453 adarśanamātra (mere non-observation) 291, 296, 300n24, 300n25, 319 adhyavasāya (determination, perceptual judgment): Dharmakīrti on 311, 315; Dharmottara on 441–3; Jñānaśrīmitra on 558–61, 562, 563, 568, 570; Ratnākaraśānti on 594; Ratnakīrti on 578, 580, 582, 583–4 agent (kārakā, kartṛ): grammatical category 407–8; merely conventional or figurative 189, 273, 289; and mind-only 405, 458; and morality 91, 102, 657; non-Buddhist views of 414; reduced to causality 20, 107, 215, 273–4, 275, 276; referencing an individual cognizer 448; rejected in cognition 289, 399, 409, 435, 532; in the tantras 49, 50, 57; see also Self ākāra (aspect, image in cognition, mental image): both conceptual and nonconceptual 558; Dharmapāla on 589; four aspect theory 368–9; grasper (grāhaka) and grasped (grāhya) 289–90, 400; Haribhadra on 505–6; Jitāri on 548–9; Jñānaśrīmitra on 566–70; Ratnakīrti on 579; in Sautrāntika representationalism 367; Śubhagupta on

663

Index 434–5; in Yogācāra ontology 588; Yogācāra theory of 365, 367, 368, 588; see also nirākāravāda; sākāravāda ālambana (mental object, objective support) 36, 435, 588 ālayavijñāna (container consciousness, storehouse consciousness, store mind): Abhayākaragupta on 621–2; Asaṅga on 74, 208, 255, 256, 257, 259, 260, 261; Dharmapāla on 367, 370, 389n17; Laṅkāvatārasūtra (The Descent into Laṅka Sūtra) on 621; Ratnākaraśānti on 593, 594; Sthiramati on 378, 379, 380; Vasubandhu on 279–80 alīkākāravāda (false form, theory that images in cognition are false): in embryonic form in Śākyabuddhi 400–1; targeted by Śāntarakṣita 468–70; as a Tibetan doxographical classification 477n19 amanasikāra (cognitive disengagement, non-mentation) 132, 489 Ambedkar, B. R. 635–49; as architect of Indian constitution 635, 643; biography of 635–7; on caste 639–40, 643; conception of political faith 636, 639; conversion to Buddhism 646, 648n11; critique of Brahmanism 636, 639–40, 643; Dr. Babasaheb Ambedkar Writings and Speeches 635, 638; emphasis on moral cultivation 642; interpretation of emptiness 643; philosophy of religion 636, 637–9; political philosophy of 643–6; The Buddha and His Dhamma 640–3 anti-foundationalism 219, 310, 465 antirealism 208, 236, 313, 316–18, 469, 470 apoha see exclusion appearances (avabhāsana, nirbhāsa, pratibhāsa): as acceptable as long as not analyzed 456; as constructed 279–81; conventional truth as 460, 471, 484, 492, 544; Dharmapāla on 363–4; as dreamlike 54, 505–6, 582, 588; as false 268, 272–3, 396–7, 506, 589, 595, 609; Jñānaśrīmitra on 558, 559–61, 562–3, 568–9, 570; as mere appearances 601–13; as nondual 507, 544, 573; overturned through antidotes 606, 607; Ratnakīrti on 573–7, 581; and reflexive awareness (svasaṃvedana) 401–2; similar to illusions 504–5; as single but variegated 582–4; Sthiramati on 382–3, 284–5, 389n17; see also ākāra Arcaṭa 429–30, 553–4 Aristotle 218, 243, 331, 523, 643 arthakriyā (causal efficacy, causal function, practical activity): ambiguity of the term 395–6; appearances as having 456; as central to Dharmakīrti’s definition of pramāṇa 310, 441; imagined entities as lacking 457; as

mark of particulars 532; as only conventional 311; as qualifying mere entities (vastumātra) 456; as requiring veridical awareness 442 Āryadeva 236–51; biography of 236–7; Candrakīrti on 240–1, 243–5, 406; Catuḥśataka (Four Hundred Verses) 236–7; as co-founder of Madhyamaka 213, 236; as “Consequentialist Mādhyamika” 240; Dharmapāla on 240–2, 349, 362, 364, 365, 367; epistemic humility in 246–7, 249n24; ethics of 242–7; misogyny of 243; “no thesis” methodology of 238–40; quietism of 238–42; on scripture and what can be known 244–6; tantric author 248n6 Asaṅga 252–65; Abhidharmasamuccaya (Compendium of the Abhidharma) 208, 253, 255, 256–8, 261, 378, 379; Abhisamayālaṃkāra (Ornament of Realization) 65, 196, 253, 497, 616; Bodhisattvabhūmi (Stage of the Bodhisattva) 29, 39, 256, 384; on the eight consciousnesses 255; as a founder of Yogācāra 459, 587; Mahāyānasaṃgraha (Summary of Mahāyāna) 208, 253, 256, 258–61; on mental verbalization (manojalpa) 261; on mind only 255, 458; on non-seeing 453; on “nothing but presentation” (vijñaptimātratā) 255, 256, 257, 259, 260, 261; on the potential (gotra) for Buddhahood 70–1; questions of authorship 65, 74, 254–7; as receiving inspiration from Maitreya in Tuṣita 64, 65, 66, 252, 253; relation to Vasubandhu 65, 252, 253, 267–8; on store mind (ālayavijñāna) 255, 256, 257, 258, 259; as systematizer of Yogācāra 256; on the three natures (trisvabhāva) 255, 256, 257, 259; works attributed to 254–5; Yogācārabhūmi (Foundation of the Yoga Practitioners) 85, 253, 261, 355, 367 Aśvaghoṣa 83–96; association with early Sarvāstivāda school 83–4; critique of metaphysical entities 92–3; critique of rituals and asceticism 86–7; critique of the philosopher-teacher Arāḍa 87–90; critique of Vaiśeṣika 89–90, 93; philosophy and apologetics 84–6; presenting a preclassical Sāṃkhya path to liberation 88; refutation of the self 90–2; on the role of rational analysis 87–90; Saundarananda (Handsome Nanda) 83, 84, 85, 86, 87, 92; treatment of Sāṃkhya 83, 87–8, 89, 90, 93; Śāriputraprakaraṇa (Drama of Śāriputra) 83, 87 Atiśa 601–14; advocating faith-based Madhyamaka 605; on debate 612; Entry to the Two Realities (Satyadvayāvatāra) 604, 605, 611; four great reasons proving emptiness 606–7; Great Middle Way 601, 664

Index 605, 611, 612; importance of reasoning 606; influence of Candrakīrti on 605, 607, 609; influence of Nāgārjuna on 604–5, 610; Lamp for the Path to Awakening (Bodhipathapradīpa) 603, 610; path of bodhisattva training 610–12; on reasoning 606–9; as synthesizing teachings of Bhāviveka and Candrakīrti 605, 608 ātman see Self atoms (paramāṇu): in the Abhidhamma 147, 158n39, 329, 338; in the Abhidharma 257, 409–11; borderline between matter and mind 379; Dharmakīrti on 316, 532; refuted by Dharmapāla 363, 365; refuted by Dignāga 288; refuted by Jñānagarbha 460; refuted by Śāntarakṣita 467–8, 471–2, 473, 477n32; refuted by Sthiramati 379–80; refuted by Vasubandhu 270, 467; refuted in the tantras 49, 52; Śubhagupta’s proof as substances 432–3, 345; and the three natures 299n8; as ultimately real 507 authors: and Abhidharma texts 174; buddhas as 4, 9–10; commentators as philosophical 499–500; consistency as a modern criterion for attribution 27–8; and Foucault’s author function 3; and innovation 381–2, 500; names as designating not persons but workshops 326; as related to authority 4, 174; texts as taking on the function of 4; as texts themselves 10; traditional attributions 27–8 Avalokiteśvara 278, 280, 346

Kamalaśīla on 487, 490; as one’s own mind 54; Śāntideva on 512–14; ultimate 610; as the ultimate nature of phenomena 53 bodhisattva: Abhayākaragupta on 621, 625; Asaṅga on 256, 258, 260–1; Aśvaghoṣa on 84–90; Atiśa on 603, 605, 610, 611; Candrakīrti on 405; Dalai Lama XIV on 651, 656–67; development of ideal 22; Dignāga on 305; Haribhadra on 502, 503; Jñānagarbha on 451, 459; Kamalaśīla on 483, 488, 492; path 184, 237, 242, 278; piṭaka (collection) 29; Ratnākaraśānti on 588, 595, 597; Śāntideva on 512–25; Sthiramati on 387 Brahmanism: critiqued by Ambedkar 636, 639–40, 642, 644–5, 647n8; critiqued by Aśvaghoṣa 84, 86; critiqued by Bhāviveka 348, 352–3; critiqued by Dharmakīrti 307–8, 318; critiqued by Dignāga 289, 313; critiqued by Śāntarakṣita and Kamalaśīla 428; critiqued by Śubhagupta 431–2; critiqued by tantric buddhas 48–9 Buddha/buddhas: as free from appearances 611; as having a mind or no mind after awakening 367; as omniscient, 46; as philosophers 9–10; tantric manifestations of 10, 46–7, 53–4; temporality of 30; as ultimate truth or reality in the tantras 10, 46, 47, 53; see also Gotama Buddha; Siddhārtha Gautama; Tantric Buddha Buddhaghosa 328–45; on Abhidhamma 141, 144, 145, 148, 189; on the aggregates 338, 341; on the Buddha’s omniscience 333; as codifying Pali commentarial literature 152; on the conventional and ultimate meaning of the Buddha’s teachings 153; hermeneutic of contextualization 329–30; on joy and delight of reading 331, 332; on nidāna (context) 332; rhizomatic presentation of the path 341; on types of understanding 340–1; on understanding 340–2; understanding of emptiness 153; use of the term sabhāva 154, 189; Visuddhimagga (Path of Purification) 117, 328 Buddha Nature: Abhayākaragupta on 625–7; Maitreya on 65, 66, 69–72; nine similies for 71; not restricted to buddhas 40; Tathāgatagarbha 74, 212, 625–7; Yogācāra on 130–1 Buddhapālita 213, 348, 373n24, 476n4 buddhasaṃjñā (idea of the buddha) 34, 36, 38, 40 buddhavacana (speech/word of the/a Buddha): as Buddhist philosophy 4, 10; as containing multiple voices 33; distinct from question of provenance 30, 39–40; as inspired speech (pratibhāna) 34–5, 38, 40; and Mahāyāna 212; questioned by Buddhist traditions 28–9; as what is well spoken (subhāṣita) 34–5, 38

Bergson, H. 636, 638, 639 Berkeley, G. 317, 477n20, 477n30, 507, 531 Bhartṛhāri 285, 287, 298, 389n20 Bhāviveka 346–60; classification as a Svātantrika 347–8; criticized by Candrakīrti 405, 414; critique of Buddhapālita 213; critique of pudgala 197; on debate as an essential element of spiritual practice 347; Dharmapāla’s response to 240, 365–7; doxographical method 349–53; Madhyamakahṛdayakārikā (Verses on the Essence of the Middle Way) 346, 349, 350–6, 384, 475; medical metaphors in 350–2; Prajñāpradīpa (Lamp of Wisdom) 227n25, 229n42, 229n50, 346, 348, 454, 458; response to Yogācāra critique by Dharmapāla 365–7; Tarkajvālā (Blaze of Reasoning) 197, 346, 354–5, 384, 605, 627n7; three connotations of ultimate reality 354, 453–4; twofold nature of ultimate reality 355; use of autonomous (svatantra) inference 347–9; works attributed to 346, 458, 475 bodhicitta (mind of awakening) 605; Abhayākaragupta on 618, 619, 621, 626; as epithet for the buddha in the tantras 47; 665

Index Candrakīrti 404–20; as aligned with Buddhapālita 348, 406; on buddhas as free from appearances 611; Catuḥśatakaṭīkā (Commentary on the Four Hundred Verses) 240–1, 243–5, 406; classified as Prāsaṅgika-Madhyamaka 347, 476n4; on conceptual dependence 416; on conventional truth 474, 457; criticized by Ratnākaraśānti 587; critique of Bhāviveka 405–6, 454; critique of Dignāga 417–18; critique of Yogācāra 405, 406, 458; Madhyamakāvatāra (Introduction to Madhyamaka) 405, 412–13, 455; use of Nyāya epistemology 417–18; on the path as gradual 458; Prasannapadā (In Clear Words) 405–6, 407, 455; on the principle of common establishment in debate 607–8; on the qualifier “ultimately” 454; reference to non-Mahāyāna texts 190, 412, 418; and Sanskrit grammar 407–8, 418; on spatio-temporally extended entities 409–10, 416–17; on the two truths 412–17; on two types of negation 130, 408; use of Abhidharma, 406, 409–11; use of Sautrāntika theory of cognition 609; on what is acknowledged by the world 245, 458; works 405–7; Yuktiṣaṣṭikāvṛtti (Commentary on the Sixty Verses on Reasoning) 406, 411 Carruthers, P. 369–70, 373n31 Cārvāka (Lokāyata): critiqued by Bhāviveka 349; critiqued by Candrakīrti 414–15; critiqued by Dharmakīrti 305–6; critiqued by Dharmapāla 365; critiqued by Śāntarakṣita 465; critiqued in the Vimalaprabhaṭīkā (Stainless Light Commentary) 51 caste: Ambedkar on 639–40, 643; Aśvaghoṣa on 85; Buddhist critiques of 647n8; in Jitāri’s biography 547–8; Jitāri’s critique of brahmins 552; Saraha’s rejection of 126 Cāttaṉār 113–23; on Buddhist logic 115–16; Cilappatikāram (The Story of the Anklet) 113–14; exemplifying philosophy as narrative 117–20; exemplifying philosophy as poetry in 120–1; on interdependent origination in 116–17; Maṇimēkalai Tuṟavu (The Renunciation of Maṇimēkalai) 113–22; as philosopher 115–22 catuṣkoṭi (tetralemma): analyzed as non-classical logic 231n66; deployed in Pali discourses 214–15, 228n29, 228n31; one of four great reasons proving emptiness 606; and the Madhyamaka “no thesis” view 239–40; as key to Nāgārjuna’s project 219–22; Nāgārjuna’s use of 214–16 causal efficacy see arthakriyā causality: backward causation 529–31, 552–3; basis for inference 270; Candrakīrti’s analysis of 408–9, 412; and dependent arising 215;

mark of existence 270–1; Nāgārjuna’s critique of 217; in Sarvāstivāda 177–8; see also dependent arising citta see mind cittamātra see mind only compassion (karuṇā): on the bodhisattva path 513, 514, 521, 595, 610; Buddha’s authority based on 305, 308, 320n5; in the early discourses 12, 21–2; evoked through literature in the Maṇimēkalai Tuṟavu 121; indivisible from nondual gnosis 57; limitlessness of 625; motivation for philosophy 405, 492; nonduality of emptiness and 47, 51, 135; objectless 51, 515; perfected over lifetimes 305–8, 534; provoked by suffering 520; required for a spiritual teacher 134; root of buddhahood 483, 619; stimulated through the cultivation of wisdom 523–5 concepts (kalpanā, vikalpa; conceptual construction) 56; Dharmapāla on 367; Dignāga and Dharmakīrti on 314–15; Haribhadra on 502, 505; Kālacakratantra on 50, 59; Ratnākaraśānti on 597; ten kinds according to Asaṅga 260; Vasubandhu on 280 consciousness (vijñāna, viññāṇa): analyzed in the Vibhaṅga (Analysis) 148; Asaṅga on the Buddha’s transformation of 261; as dependently arisen 389n17, 409; eight forms or modes of 255, 385; as one of five aggregates 180, 268; higher-order theories of 369–70; inanimate objects and trees lacking 51; as insubstantial 507; as merely a convention 567; non-referential 90; in Sarvāstivāda Abhidharma 180; as one of six elements 182; six sense consciousnesses 180, 269–70, 273, 275–7, 279–80, 409, 416; transformation of 276–7, 367, 382–4, 389n17, 390n22, 390n24, 390n28; as one of twelve links of dependent arising 176, 214–15; in Yogācāra ontology 275, 279, 566–7, 588; see also ālayavijñāna; five aggregates; mind-only conventional truth (saṃvṛtisatya; conventional reality, customary truth): agreeable as long as not analyzed 456, 470, 474; Āryadeva’s revisionism concerning 242–4, 245; Bhāviveka on 352, 474, 475; Candrakīrti on 245, 409–12, 413–16, 455, 474, 609; Dharmakīrti on 313; Dharmapāla as restricting to the imagined nature 241; Jñānagarbha on 455–8, 460, 474; Nāgārjuna on 212, 220, 224, 225; relationship to ethics 242; Sthiramati’s critique of 386–7; Śāntarakṣita on 45–56, 470–2, 474–5; threefold criterion for 470; true versus false conventional 455, 456, 457, 461n4; as “valid fictions” 416; as what is acknowledged in the world 245, 457, 460 666

Index critical examination (parīkṣā): as analysis of reality (tattva) 86; decisive to the development of Buddhist philosophy 85, 347; as a sine qua non of liberation 93; in Śāntarakṣita’s Tattvasaṃgraha 485

links of 20, 23, 116, 117, 176–7, 185n7, 223, 224; see also causality; dependent nature dependent nature (paratantrasvabhāva) 240; Asaṅga on 256, 257; as ceasing at Buddhahood 366–7; Dharmapāla on 241–2, 365–6; and Dignāga’s typology of cognition 285–6, 299n8; Kamalaśīla on 484; Sthiramati on 387; see also three natures Devendrabuddhi 393–403; on arthakriyā (causal functionality) 395–8; on causal order 530; criticized by Jñānagarbha 452; criticized and relied on by Kamalaśīla 484, 485; mediated and unmediated effect 395–8; as part of the “philological school” of commentators 528; relationship to Śākyabuddhi 393, 400; Śubhagupta influenced by 433 Dewey, J. 635–6, 638 Dhammapada 110n8, 197, 643 dhamma/dharma (element, phenomenon) 145, 189; analyzed in terms of human experience 148; as conditioned 176–8; defined in terms of particularities 153; definition of 146; as dependently arisen 274; identified with sabhāva 151; list of fifty-six 146, 147; organized according to diverse modalities 147; as primary existent object 149; psychological 146; as topic of Abhidharma 172, 189, 266, 379; understood as ultimate elements of existence 152 dharmadhātu (realm of phenomena) 54, 58, 260, 597, 603, 608, 625–7 Dharmakīrti 303–21; antirealism of 312, 313, 316–18; apologetic purpose 85, 312; argument against the Self 94n31, 307, 308; on causal efficacy (arthakriyā) 311, 314; on conceptual construction 315, 502; critique of empiricism 291–2, 295, 298; critique of physicalism 305–7; definition of means of knowledge (pramāṇa) 309–12; on exclusion (apoha) 312–14; on inference 291–2, 295, 313, 315, 318–19, 529; Nyāyabindu (The Drop of Logic) 304, 549; on particulars 314, 441; on perception 287, 314–16; Pramāṇavārttika (Commentary on the Means of Knowledge) 317–18, 394–5, 398–402; Pramāṇavārttikasvavṛtti (Auto-Commentary) 304, 309, 312–14, 394–5, 531; Pramāṇaviniścaya (Ascertainment of the Means of Knowledge) 304, 314, 315, 317, 433, 493n13, 549, 567; on reflexive awareness (svasaṃvedana) 317–18, 398–402; relation to Dignāga 291, 295, 304–5; on religious authority 245–6, 305, 308–9; responding to Kumārila 285, 290–1; theory of causation 529; on universals as conventional 312–14, 441 dharmanairātymya (selflessness of dharmas) 409, 621, 625

Dalai Lama XIV 650–62; engagement with Christianity 652, 654; engagement with modern science 652, 660–1; engagement with Prāsaṅgika-Madhyamaka tradition 651, 660; engagement with thought of Mohandas K Gandhi 650, 652, 659; as leader in engaged Buddhism movement 659; interpretation of Śāntideva 653, 656–9; modernist version of Buddhist ethics 656–9; as philosopher 5, 650–1; religious pluralism of 653–6; works 651 Daśabalaśrīmitra 194, 195, 197, 198–200, 618, 627 deathless state (amata) 12, 13–14, 17, 21, 23, 24 debate (vāda): Bhāvaviveka on 365; Dharamakīrti on 295; Dharmapāla on 365; Dignāga on 285, 295, 365; pan-Indian rules of 348; post-Vasubandhu developments in 364–5; principle of common establishment 456, 607–8; represented in the Maṇimēkalai Tuṟavu (The Renunciation of Maṇimēkalai) 121; sociohistorical context of 347 definitive meaning (nītārtha) 46, 70, 343n5, 412, 451, 458 dependent arising (pratītyasamutpāda, paṭiccasamuppāda): according to Aśvaghoṣa 93; according to Nāgārjuna 213–15, 217–19, 225, 406, 470, 604; as an antidote to anger 522; in Buddhist epistemology 290, 433, 435, 530, 570; as the cause of the world in the tantras 49–50; cessation of as the eradication of ignorance 23; and compassion 653, 658–69; and conventional truth (saṃvṛtisatya) 456, 460, 470, 605, 609; as a correlate to no-self 24, 190; and dependent designation 601; and the dependent nature 69–70, 365; as discerned by Gotama Buddha 20; and emptiness 53, 216, 218–22, 350, 416–17, 473, 523, 606–7, 623, 660; as foundational principle for the Buddha’s teachings 20, 274; and Indian philosophical argumentation 530; and no-self 268; as taught in Maṇimēkalai Tuṟavu (The Renunciation of Maṇimēkalai) 114–15, 116–17, 119, 120; as taught in Abhidhamma/Abhidharma 145, 148, 155, 176–7, 200, 214–15, 416; as taught in the Śālistambasūtra (Rice Seedling Sūtra) 412, 488; as taught in the tantras 49–50; as taught in Yogācāra 69–70, 275, 278, 281, 289, 384–6; and the transformation of consciousness 383, 384–5, 389n17; twelve 667

Index Dharmapāla 361–75; in the Cheng weishi lun (A Treatise for the Establishment of Consciousness-Only) 361, 363; commentary on Āryadeva’s Catuḥśataka (Four-Hundred Verses) 240–1, 244, 245, 349, 362, 364, 365, 367; commentary on Dignāga’s Ālamabanaparīkṣā (Investigation of Cognitive Objects) 362, 364; commentary on Vasubandhu’s Viṃsikā (Twenty Verses) 361, 362, 363–4, 371; criticism of non-Buddhist schools 365; criticism of Bhāviveka 240–1, 361, 365–7; as holding opposite views to Paramārtha 370; influence on East Asian Buddhism 361, 362, 370–1; on meta-cognition (svasaṃvedana) 362, 367–70, 373n25; as original and creative or conservative thinker 362–3, 370; and theories of higher-order consciousness 369–70; theory of the cognition of self-cognition (*svasaṃvitti-saṃvitti) 368–70; on the three natures (trisvabhāva) 240, 365–7, 240–1; use of Dignāga’s logic for the interpretation of Yogācāra 361, 363–4 *Dharmaśreṣṭhin 180–3, 184 Dharmottara 439–49; association with Arcaṭa and Śubhagupta 429–30; cognizer-relative shift in epistemology 443–6; commentary on Dharmakīrti’s Nyāyabindu (The Drop of Logic), 439, 441, 444–6; critique by Jitāri 548–9; critique by Ratnākaraśānti 587; determination of a continuum (santāna) 442; doubt and certainty in inference-for-oneself 446–7; divergence from Dharmakīrti 440–3; inference as involving determination of a particular 443; influence on Jñānaśrīmitra and Ratnakīrti 440, 443, 559–61; life and works 439; non-apprehension of the perceptible (dṛśyānupalabdhi) 443–6; perception and inference as having two objects 441–3; perception as involving conceptual construction 441–3 dhātu (sense realm, entrance) 269, 411, 458 Dignāga 284–302; as accepting only two pramāṇa 284–5, 310; Ālambanaparīkṣā (Investigation of the Cognitive Object) 288, 362, 364–5; on antinomy 295; apologetic purpose 85, 285; Bhāviveka’s reliance on 348, 349, 354, 356; on conceptual construction 315; critiqued by Candrakīrti 408, 417–18; definition of pramāṇa 314; Dharmapāla’s reliance on 361, 362, 363–4, 371; on exclusion of what is other 295–8, 313; Hetucakraḍamaru (Two-Headed Drum of the Wheel of Reasons) 288; on inference 291–5, 318, 319, 363; influence of Bhartṛhari on 285; influence on Dharmakīrti 304; Nyāyamukha (Gateway to Logic)

284, 288, 299n5; on perception 285, 286; Pramāṇasamuccaya (Compendium of the Means of Valid Cognition) 295; on the principle of invariable concomitance 291–2; reliance on and critique of Vasubandhu 285; on self-awareness (svasaṃvedana/ svasaṃvitti) 269, 288–91, 368, 369, 370, 373n25; Traikālyaparīkṣā (Investigation of the Three Times) 285, 287; on two types of cognition 285–7; unifying epistemology, dialectics, and semantics 285 dilemmas: in Aśvaghoṣa 90; in Milindapañha (Questions of Milinda) 101–2; in Śāntarakṣita 466, 467, 469, 470; see also catuṣkoṭi doubt (vicikitsā) 168, 178, 182, 245, 296, 311, 440, 446–8, 472, 574 doxography 2–3, 5, 326, 356n3; Āryadeva’s Jñānasārasamuccaya (Compendium on the Essence of Knowledge) 247n4, 548; Bhāviveka’s Madhyamakahṛdayakārikā (Verses on the Essence of the Middle Way) 346–56; Cāttaṉār’s list of ten schools in the Maṇimēkalai Tuṟavu (The Renunciation of Maṇimēkalai) 113, 115, 119–20; Dharmapāla’s criticism of non-Buddhist schools 365; Jitāri’s Sugatamatavibhaṅgakārikā (Verses on the Classification of Buddhist Systems) 548, 551, 554–5; Kamalaśīla’s Tattvasaṃgrahapañjikā (Commentary on the Compendium of Realities) 428, 434, 484–6; pudgalavāda as presented in 191, 197–8; Śāntarakṣita’s Tattvasaṃgraha (Compendium of Realities) 428, 464–5, 484–6; tantric Buddha as doxographer in Kālacakratantra (Wheel of Time Tantra) 47–53; Tibetan 211, 459, 476n4, 490–1, 549, 615 Durkheim, É. 637 ekānekaviyogahetu see neither-one-nor-many argument ekayāna see single vehicle emptiness (śūnyatā, suññata): Abhayākaragupta on 625; Ambedkar on 643; Āryadeva on 245–6, 248n9; Asaṅga on 257; Atiśa on 605, 606, 607, 610, 611, 612; Bhāviveka on 349–50, 353; Buddhaghosa on 153, 157n37; Candrakīrti on 406, 409–10, 412, 414, 416–17; conscious (ajaḍaśūnyatā) 48; Dalai Lama XIV on 651, 657, 660; as dependent arising 229n51, 416; Dharmakīrti on 316; Dharmapāla on 244, 361, 366–7; emptiness of 230n54, 230n57; and epistemic foundationalism 221; Haribhadra on 499, 504–6; Jñānaŕīmitra on 567; Kamalaśīla on 483–4, 487, 489; Nāgārjuna on 214, 218–22, 223, 229n50; Ratnākaraśānti on 402, 668

Index 594, 595, 597; Śāntarakṣita on 465–6, 470; Śāntideva on 515, 523; Saraha on 129–31, 134, 135; in the tantras 46, 53–5; Vasubandhu on 277–9 error (bhrānti): Buddha’s retaining a small amount of 595, 596–7; distinction between error and aspect 505–6; and the identification of a self 271–2; inherent in dualistic cognition 400; as lacking causal efficacy 457; as mark of the conventional 456–7; non-conceptual 397; Ratnākaraśānti on 593 ethics: and Ambedkar’s understanding of Buddhism 638–9; Āryadeva on 242–7; brahmavihāra (divine abiding, immeasurable) 148, 637, 657, 658; Dalai Lama XIV on 654, 656–9; Dharmakīrti on the possibility of moral transformation 308; ethical implications in Milindapañha (Questions of King Milinda) 102; moral cultivation 642, 643, 656–7; Nāgārjuna on 208, 212, 213, 232n79; relationship to conventional truth 242; role of ethics in early Buddhist thought 23; Śāntideva on relationship to metaphysics 511, 523–5, 526n2; Saraha on 128 exclusion (apoha, anyāpoha): Dharmakīrti on 313–14; Dignāga on 284, 295–8, 313; Jitāri on 552; Jñānaśrīmitra on 448n7, 560, 561–4, 602; Ratnakīrti on 448n7, 573, 578–82, 584; Śubhagupta on 430, 431 externalism (bāhyārthavāda) 432–3

transformation of 261; Buddhaghosa on 338; Dharmapāla on the lack of externality of 363–4; Jñānagarbha on the appearance of as not negated 455, 460; occurring always with other dhammas 147–9; as one of four main types of phenomena 151; Sthiramati on 379; see also five aggregates formations (saṃskāra, saṅkhāra; volitional constructions, conceptual patterns): absent in discriminating gnosis 58; analyzed in the Vibhaṅga (Analysis) 148; Asaṅga on the Buddha’s transformation of 261; disassociated from thoughts (cittaviprayukta) 180; as impermanent 199; see also five aggregates Foucault, M. 3, 27, 490 four noble truths: in the Abhidhamma 144–5, 148; Bhāvviveka on 352; critiqued by Ambedkar 641–2; Dharmakīrti on 305, 309, 315, 317, 533; and eightfold path 23–4; explained by Cāttaṉār 116, 117; as means to abandon heterodox views 175; as a means to eliminate defilements 176, 178, 180; as medical diagnostic scheme 22, 24; Nāgārjuna on 217–18; as a practice of diagnosing experience 155; as taught by Gotama Buddha 12, 22–4; understood on the path of vision 182–3; see also suffering Gadamer, H.-G. 335–8 Gandhi, M. K. 639, 650, 652, 657, 659 Gautama Buddha see Siddhārtha Gautama God see Īśvara Gorampa 248n15, 460, 481 Gotama Buddha 11–26; construction of personal identity 19–20, 21; decision to teach 21–2; figure as close as possible to the historical Buddha 11; immaterial attainments 16; night of awakening 16–17; pursuit of ascetic practices 17–19; recollection of past lives 19–20; as represented in early discourses 11–12, 24n1 gradualism 415, 451, 458–9, 483, 492 grahaka (apprehender, subject) 53–4, 56, 289, 400, 589 grāhya (apprehended, grasped object) 53, 56, 59, 289, 400 Guhyasamājatantra (Secret Assembly Tantra) 47, 53–4, 56, 450, 487, 547–8, 626 guṇa (quality) 59, 89, 90

feeling (vedanā; feeling tone): according to the “Many Types of Feeling Sutta” 154; analyzed in the Vibhaṅga (Analysis) 148; Asaṅga on the Buddha’s transformation of 261; and craving (taṇhā/tṛṣṇā) 23; as an “ultimate” (paramattha) dhamma 151; see also five aggregates five aggregates (skandha, khanda): analyzed as empty 153, 218; analyzed by Sthiramati 378–9; Asaṅga on the Buddha’s five kinds of mastery (vibhutva) relating to 261; associated with the truth of suffering 144; as the basis for the false construction of a self 268–73, 278, 338; as an ever-changing continuum 268; examined in terms of modalities in the Analysis (Vibhaṅga) 148–9; taught as part of a gradual path 458; in the Pañcaskandhakavibhāṣā (Commentary on the Treatise on the Five Constituents) 378; Nāgārjuna on 218; as part of the skandha-dhātu-āyatana framework 189; in Sarvāstivāda Abhidharma 180; in the tantras 53, 59; as teaching appearance and emptiness 278 form (rūpa; matter): analyzed in the Vibhaṅga (Analysis) 148; Asaṅga on the Buddha’s

Habermas, J. 339 Hadot, P. 136, 509n9 Haribhadra 497–510; commentaries on the Ornament of Realizations (Abhisamayālaṃkāra) 499–504; as exegete 497, 508; as follower of Śāntarakṣita 464; and Kamalaśīla 434; as metaphysician 506–8; as 669

Index philosopher of the path 497, 501–4; questions of authorship 499–500; as specialist in prajñāpāramitā literature 497, 498; theory of illusion 504–6 Hegel, G. W. 500, 509n4 Heidegger, M. 136 hetu see logical reason Hevajratantra 47, 55, 56, 57, 58–9, 60, 61, 603 Hume, D. 477n20

interpretable meaning (neyārtha; provisional meaning) 46, 70, 268, 274–6, 343n5, 412, 451, 458 intersubjectivity 275, 387–8, 596 Īśvara (God): Aśvaghoṣa’s rejection of 92–3; Bhāviveka’s refutation of 455; Dharmakīrti’s rejection of 305; Dharmottara on 444, 445; Jitāri’s refutation of 551; Jñānaśrīmitra’s refutation of 558; in Nyāya-Vaiśeṣika 533; Ratnakīrti’s arguments again 602; refuted in the tantras 49–50; Śubhagupta’s refutation of 429, 430–1 Īśvarasena 291, 295, 298, 319

imagined nature (parikalpitasvabhāva): as defilement 72; Dharmapāla on 366; as existing only nominally 69; including three kinds of conceptual cognition according to Dignaga 286–7; Jñānagarbha’s critique of 453, 456; Ratnākaraśānti on dissolution of 589; see also three natures imagination of the unreal (abhūtaparikalpa) 69, 241, 459 impermanence (anityatā, aniccata): Ambedkar on the ethical significance of 643; Cāttaṉār on the liberating insight into 116; Śāntideva on tension between bodhisattva vow and 515; Vasubandhu on impernance as conditions 280–1 ineffability/inexpressibility: of the awakened state 135; of causal facts 390n24; of the dependent and perfect natures 240–2; of the “indeterminate self” (pudgala) 190, 274; of properties 563; of reality 57, 268, 275, 276, 277; of ultimate truth 452–5; of underlying ontology in Yogācāra 386 inference (anumāna): adopted by Bhāviveka 347, 348, 349, 355; autonomous (svatantrānumāna) 347, 349, 450, 476n4, 491; and the Buddha’s omniscience 534; Dharmakīrti on 446, 529–30; Dharmakirti’s skepticism of scripturally-based 246; Dharmottara’s theory of two objects of 441, 443; Dignāga on 291–5, 363; examples (dṛṣṭānta) in 245, 284, 363; inference-foroneself (svārthānumāna) 209, 284, 304–5, 431, 439, 446–7, 528, 536, 549; inference-for-others (parārthānumāna) 209, 284, 299n23, 304–5, 423, 439, 446, 447, 528, 549; inferential reasoning as used by Cāttaṉār 115–16, 119; governed by a natural connection (svabhāvapratibandha) 295, 319; Jñānaśrīmitra on 558, 559–61; in the Maṇimēkalai Tuṟavu (The Renunciation of Maṇimēkalai) 121; scripturally-based (āgamāśritānumāna) 245, 310, 488; similar and dissimilar examples (dṛṣṭānta) in 291–2, 293–5; threefold nature 363; Vasubandhu on the inference of sense organs 270; see also logical reason interdependent origination see dependent arising

Jaimini 119, 285 Jains: Bhāviveka critique of 349; in contrast to Buddhism 110n1; critiqued by Āryadeva 349; critiqued by Dharmapāla 365; critiqued by Jitāri 550, 554; critiqued by Kamalaśīla 485; critiqued by Śāntarakṣita 465; mainstream meditation practices of 87; as pluralistic 653; represented as opponents of Buddhism 110n7, 120, 127; represented as using dilemmas in debate 101; theory of the soul (jīva) critiqued in the tantras 52; Vidyānandin’s critique of Prajñākaragupta 528; view of the pudgala as material and non-sentient 191 Jayanta 320n3, 528, 537n1 Jayanta Bhaṭṭa 439, 448n1 Jinendrabuddhi 284 Jitāri 547–57; affiliation with Madhyamaka 548; affiliation with Yogācāra 554; as author of short digests 427, 550; on backward causation 531, 552–3; on Dharmakīrti as a Mādhyamika 555n5; life and works of 547–9; tantric practice 547–8; Vādasthānāni (Topics of Debate) 549–5 Jñānagarbha 450–62; on reflexive awareness (svasaṃvedana) 460; on conventional truth (saṃvṛtisatya) 455–8, 460; as developing Bhāviveka’s ideas 450, 453–4; as embracing a gradual path 458–9; on imagined versus mere entities 456–7; on mind-only as conventionally valid 458–9, 460; on the principle of common establishment in debate 456; on the qualifier “ultimately” 454–5; Satyadvavibhaṅgavṛtti (Commentary on the Distinction Between the Two Truths) 459–60; as synthesizing Yogācāra and Madhyamaka 450; tantric works attributed to 451; Tibetan doxographical classifications of 459–60; on the two truths 452–4, 464 Jñānaśrīmitra 558–72; on appearance and determination 559–61, 570; on conventional reality 564–6; on exclusion (apoha) 561–3; influenced by Dharmottara 440, 561; life and works 558–9; similarity to 670

Index Prajñākaragupta 528, 533; struggle with Ratnākaraśānti 567–70; synthesis of Buddhist epistemological thought 570

lists (mātṛkā, mātikā; matrix): as closed 150; in the Enumeration of Phenomena (Dhammasaṅganī) 146–8; as generative 152; as inherently incomplete 147, 155; as a modal method of analysis 147, 152, 155; as a view-from-nowhere 151; as reflecting oral/ aural context of teachings 172 logical reason (hetu, liṅga; evidence, inferential reason): Dharmakīrti on 529–30; Dharmottara on 443, 446–7; Dignāga on 446; effect-evidence (kāryahetu) 529, 530; identity-evidence (svabhāvahetu) 529, 559; nonperception (anupalabdhi) 443–6, 531–3; pseudo-reason (hetvābhāsa) 294, 295, 447, 574, 575; triple conditions for 446–7 logico-epistemological (pramāṇa) tradition 303, 361, 427, 428–9, 433, 573 Lokāyata see Cārvāka luminosity (prabhāsvara, prakāśa; illumination): of awareness 486; Dharmakīrti on the mind as naturally luminous 308; in Dharmakīrti’s definition of reflexive awareness 390–1, 402; Jñānaśrīmitra on 567–8; as the nature of the mind in Abhayākaragupta 625, 626; Śākyabuddhi’s account of contentless luminosity 402; as synonym for reflexive awareness in Ratnākaraśānti 402, 581, 590–1, 602, 626; in the tantras 54, 55, 56; of the ultimate awakening mind 610

Kālacakratantra (Wheel of Time Tantra): Abhayākaragupta’s commentaries on 616–17; Buddha as philosopher and doxographer in 47–55; on the mind as inseparable from enlightened awareness 58; on nirvāṇa as signless 57; on the world as constructed by karma 59 Kamalaśīla 480–96; acceptance of Yogācāra theories 484, 486; doxographical classification of 481, 485–6, 490–2; embodied dimension of authorship 481; Extensive Commentary (Pañjikā) 484–6; gradualism of 480–1, 483, 492; historical accounts of 480, 481; influence on Abhayākaragupta 615; influenced by Śubhagupta 428–9; Madhyamakāloka (Light of the Middle) 482–4, 615; neither-one-nor-many argument in 466; provisional acceptance of mind-only 475; questions of authorship 486–7; on reflexive awareness (svasaṃvedana) 486; Stages of Cultivation (Bhāvanākrama) 481, 482–3, 651; on term false conventional (mithyāsaṃvṛti) 461n4 Kant, I. 81, 226n12, 249n24 karma/karman (action): according to Tājiks (Tāyins) 51; Ambedkar on 641, 642; as bondage 86, 90; denied by Cārvākas 51; as incompatible with a self 90, 91–2; as mental 588; as not due to God 49; in preclassical Sāṃkhya 88–9; as taught through narrative 118–20, 188; theorized in Abhidharma 177, 180, 188; without an agent 189 Karṇakagomin 246, 303 Kathāvatthu (Points of Discussion) 160–71; citing the nun Śaila on “being” (satva) 190; distinction between conventional and ultimate teachings 152; method of argumentation 161–8; scripture and reason as implicitly accepted as pramāṇas in 170; see also Abhidhamma; Moggaliputta Tissa Kātyāyanīputra 172–87; as addressing a monastic audience 178, 179; as compiler of the *Abhidharmāṣṭagranthaśāstra (Abhidharma Treatise in Eight Sections) 175; as founder of the Sarvāstivāda 180; see also Mahāvibhāṣā (Large Commentary) Kuiji (Master Ji) 368, 370 Kumārajīva 190 Kumārila: critique of the Buddha’s omniscience 533; critique of Dignāga 285, 290; definition of pramāṇa 310; responded to by Śāntarakṣita 465

Madhyamaka: Āryadeva as co-founder of 213, 236; Bhāviveka as systematizer of 348; criticized by Yogācāra philosophers 384–7; critique of realist semantics 239; Dharmapāla’s three-nature reading 240–2; as the elimination of views 218, 221, 350, 352, 354; and ethics 212–13, 242–7, 523–5; dependent origination in 213–15, 217–19, 225, 406, 470, 604; Nāgārjuna as founder of 213, 405; as rejecting any thesis 221–2, 239, 364, 607; synthesized Yogācāra 450, 472–5, 570; as teaching notational as well as existential dependence 416; as therapy 350; see also emptiness mahābhūta (four great elements) 148, 179, 410 Mahāsāṃghika 368 Mahāvibhāṣa (Large Commentary) 172–87; Aśvaghoṣa as possible scribe for 83; on conventional cognition as ignorance 457; and Kātyānanīputra 183–4; and Sarvāstivāda 174, 177–8; as a source for Mahāyāna developments 184 Mahāyāna sūtras 27–40; Abhayākaragupta on the authority of 619–20; Acintyabuddhaviṣayanirdeśasūtra (Sūtra of Instruction on the Inconceivable Sphere of 671

Index Buddhas) 38–9; Adhyāśayasañcodanasūtra (Sūtra which Incites Resolve) 32, 34, 35, 39; Akṣayamatinirdeśasūtra 453; Bhadramāyākāravyākaraṇasūtra (Prophecy for the Magician Bhadra) 39; as buddhavacana (word/speech of the/a Buddha) 29, 31–40; Dharmasaṅgītisūtra 453; Dāśabhūmikasūtra (Ten Stages Sūtra) 393, 610, 619; echoed in the tantras 53; as featuring multiple voices 33; Gayāśīrṣa Sūtra (Summit of Mount Gayā Sūtra) 393; Kāśyapaparivarta (Chapter on Kāśyapa) 377, 381, 489, 608; Laṅkāvatārasūtra 389n19, 457, 621; Mahāvairocanasūtra (Discourse on the Enlightenment of Mahāvairocana) 451; Maitreya Buddha as teaching 252, 253; Śālistambasūtra (Rice Seedling Sūtra) 412, 488; Śrīmālādevīsūtra (Discourse for Queen Śrīmālā) 71, 625; Tathāgatagarbhasūtra (Discourse on Buddha Nature) 71; Ugraparipṛcchāsūtra (Questions of Ugra) 517; see also Prajñāpāramitāsūtra; Saṃdhinirmocanasūtra (Scripture Clarifying the Intent) Mahīśāsaka school 253 Maitreya Buddha 64–77; as author 65; cult of 64; as founder of Yogācāra 64; relation to Maitreyanātha 74; synthesis of Yogācāra and Tathāgatagarbha thought 72–4; as teacher of Asaṅga 64, 252, 253, 268 manas see mind Mañjuśrī 38, 548 Manorathananandin 303 manovijñāna see mental perception Marx, K. 636, 637, 643, 644, 647 materialism see Cārvāka; physicalism mātṛkā/mātikā see lists matter see rūpa means of valid cognition see pramāṇa meditation (bhāvanā; cultivation, habituation): for abandoning mental afflictions 516; on awakening mind (bodhicitta) 618, 619; calm (śamatha, samatha) 16, 150; on compassion 651; on the disgusting or repulsive (aśubhabhāvanā) 153, 182, 243, 248–9n18; on the divine abodes (brahmavihāra) 637; insight (vipaśyanā, vipassanā) 16, 150, 409; as giving rise to a vivid appearance 623; as habituation 503; jhāna/dhyāna 14–16, 148, 169; nondual 601, 603, 606, 610; path of (bhāvanāmarga) 183, 287, 595, 606; philosophy and ritual as types of 515; recollection of the Dhamma 336; salutary practices of 653; stages of described by Aśvaghoṣa 85, 87, 89; tantric 59–60, 616; wisdom born of (bhāvanāmayīprajñā) 356, 508; see also mindfulness

memory, 65, 269, 271, 287, 311, 315, 368, 369 mental perception (manovijñāna): Asaṅga on 255, 261; Dharmakīrti on 315; Vasubandhu on 269–70, 279–80; see also perception mereological (part-whole) analysis 316, 409–10, 467, 471, 551, 660; see also ekānekahetu (neither-one-nor-many argument) metaphor (upacāra) 274–81, 289–91, 297, 381–8 Milindapañha (Questions of Milinda) 97–112; and the art of questioning 108–9; appeal to narrative 103–5; dramatic movement of 105–7; example of lion in a gilded cage 107–8; heterogeneity of 98; and impersonalism 105–7; invitational quality of 109; Menander and Nāgasena as dramatic characters 98; Menander as a historical figure of 97; and question of relevance 108–9; as representative of a genre 97; on understanding 340; use of dilemmas 101–2; use of illustrations in 99–101 Mīmāṃsā 119, 314, 351, 353, 464 mind (manas, citta; awareness, thought): distinguishing good and bad 146; manas (mental organ) 269–70, 279, 285, 379, 534; defiled (kliṣṭaṃ manas) 256, 259, 260, 380, 593, 622; defined as nirvāṇa in the Kālacakratantra (Wheel of Time Tantra) 57; eliminated in the transcendent (lokuttara) mind 279; natural (nijacitta) mind 125; in the Pali Abhidhamma 146–7, 150, 151, 189; as primordially pure according to Saraha 129–31; in Sarvāstivāda Abhidharma 179, 180; twofold appearance of in tantras 56; see also mental perception mindfulness (sati; smṛti): on the bodhisattva path 610; Buddhagosa on 146–7; four applications or foundations of 21, 144, 148, 181, 182; as an intervention against craving (taṇhā/tṛṣṇā) 23; in the Pali texts 12, 18, 19, 21, 24, 150, 155; in tantric texts 56, 132; as a virtue 512 mind-only (cittamātra; consciousness-only): Asaṅga on 255, 256, 257, 259, 458; Abhayākaragupta on 622–3; Āryadeva on 240; Bhāviveka on 458; Candrakīrti on 405–6, 458; Dharmakīrti on 291, 306, 316–17; Dharmapāla on 363–4; Dignāga on 285–6, 288; Jñānagarbha on 459, 460; Jñānaśrīmitra on 566; Kamalaśīla on 429, 475, 486; in Mahāyāna Sūtras 256, 263n38; Maitreya on 72; Nāgārjuna on 406; ontological versus epistemological interpretation 276; Śākyabuddhi on 402; Śāntarakṣita on 429, 472–5; Sthiramati on 458, 459; Śubhagupta as refuting 432–6; Vasubandhu on 268, 274–7, 289; see also vijñaptimātra 672

Index Moggaliputta Tissa 160–1, 166, 170, 184; see also Kathāvatthu (Points of Discussion) Mokṣākaragupta 447, 448 momentariness (kṣaṇikatva): arguments for in relation to a self 91–2; as characteristic of particulars 441; Dharmakīrti’s proof of 314, 535; Dharmottara on 442, 443; Jitāri’s proof of 552, 553–4; Jñānaśrīmitra’s proof of 565–6; and perception 273, 442; Ratnakīrti’s reconciliation of with universals 578–80; Ratnakīrti’s proofs of 573, 574–8, 583, 584n7; and Sarvāstivāda 177; as tied to appearances 281; Vasubandhu’s proof of 270–1 Mūlasarvāstivāda Vinaya 84, 490; see also Sarvāstivāda

387; of the Buddha 352; as characteristic of perception 442; see also perception nonduality: of emptiness and compassion in the tantras 51; knowledge as nondual 500–1, 506; in nondual meditation 601, 603, 606, 610; of saṃsāra and nirvāṇa in the tantras 55–7; variegated (citra-advaita) 469, 545; in Yogācāra 280–1 nonexistence (abhāva) 271, 314, 623–4 nonperception (anupalabdhi): Dharmakīrti on 300n25; Dharmottara on 443–6; Dignāga on 291, 298; as a logical reason 531–2, 537n3; of the pervader 552; of the perceptible 443–6 no-self (anattā, anātman; non-self, selflessness): according to Aśvaghoṣa 84, 85, 90; as realized by Gotama Buddha 19–20; taught in the Abhidharmakośabhāṣya (Treasury of Abhidharma Commentary) 268–70; taught in the tantras 50; unique teaching of the Buddha 20, 188 Nussbaum, M. 523 Nyāya school 464, 465; pramāṇa theory 310, 417; theory of the whole (avayavin) 316; view of self 90

Nāgārjuna 211–35; authorship defined by analytic works 211–12; Bhāviveka’s interpretation of 346; Candrakīrti’s interpretation of 405–6, 407; as critical of Abhidharma 212, 213, 215–17; compared to Skeptics 221; critique of the epistemic project 221–2; on dependent arising (pratītyasamutpāda) 213–19; on emptiness 218; and ethics 212–13; as a fundamentalist reformer 213; verses of praise 213–14; and Mahāyāna 212; Mūlamadhyamakakārikā (Root Verses of the Middle Way) 213–22; Ratnāvalī (Precious Garland) 213, 466, 223–4; tantric author 248n6; on the two truths 223, 230n57; Vigrahavyāvartanī (Dispeller of Disputes) 211; works 211, 226n9; Yuktiṣāṣṭika (Sixty Verses on Reasoning) 211, 238, 406, 411, 477n26, 605 narrative 4, 9, 81–2, 103–5, 107, 117–21 neither-one-nor-many argument (ekānekaviyogahetu): in Abhayākaragupta 623; against material simples 467–8; against mental simples 468–9; in Kamalaśīla 466, 483; in Nāgārjuna 466; in Śāntarakṣita 465–70; in Śrīgupta 465, 466 Neoplatonism 506, 507 Nietzsche, F. 4, 81, 124, 129, 135 Nikāya Buddhism 172 nirākāravāda (theory that cognition is not endowed with an image, form free): Dignāga’s critique of 289; Jitāri as adherent of 549; Śākyabuddhi’s account of 402; Śubhagupta’s defense of 430, 432; see also sākāravāda nirvāṇa: and attainments 16; defined as the mind of awakening (bodhicitta) 55; in Kamalaśīla’s Stages of Cultivation 483; non-abiding 625; ultimate truth as condition for 452 nonconceptual cognition (nirvikalpajñāna): awareness obtained subsequent to it (tatpṛṣṭalabdhajñāna) 387; of bodhisattvas

object (ālambana, viṣaya) 36; as absent in nirvāṇa 57; of the ālayavijñāna 380; as already perceived 532; external 290; final cognitive state as lacking (anālambana, nirālambana) 505, 594; mental 36; as momentary perceptual content 588, 589; as pure 55, 59; six sensory 268; Śubhagupta on cognition’s need for 433–5; two kinds as basis for two kinds of pramāṇa 441; as ultimate 454; Vasubandhu on 36, 268–9, 275–6, 279 omniscience (sarvajña, sarvavid): Bhāviveka on 353; Jitāri on 547, 552; Jñānagarbha on 453; mantras as evidence of the Buddha’s 432; Prajñākaragupta on 531, 533–5; Śubhagupta on 428, 430, 431–2; of tantric buddha 46, 47 other-emptiness (gzhan stong) 131, 240 Pali discourses (sutta): Aṅguttara Nikāya 13, 42n25, 145, 243, 335; Dīgha Nikāya 16; Khuddaka Nikāya 109n1, 160, 199; Majjhima Nikāya 14–20, 101, 110n13, 333; manuscript tradition 12–13; Saṃyutta Nikāya 22, 214, 216, 217, 220; Suttanipāta 108, 214, 215, 228, 350; as sources for Abhidhamma 143–56, 162 paradox: Nāgārjuna on 230n57; Saraha on 129–30, 135 Paramārtha 64, 173, 184, 194, 196, 252, 366, 370–1 paramārthasatya see ultimate truth particular (svalakṣaṇa, bheda; particular characteristic, defining trait): Candrakīrti’s 673

Index critique of Dignāga on 408; as dependent concepts 410; as existing ultimately 441; as having causal efficacy (arthakriyā) 532; as the object of perception 311, 441, 532; realist semantic theory of 296; see also universal paryudāsa (affirming negation) 130, 408 path of vision (darśanamārga) 176, 178, 181, 183, 594, 606, 609–10 perception (pratyakṣa, upalabdhi): Dharmakīrti’s definition of 623; Dharmotta on the two objects of 441–3; Dignāga’s definition 285, 315; four types according to Dharmakīrti 315–16; Jñānaśrīmitra on 559; mental perception (mānasapratyakṣa) 286; non-conceptual versus conceptual 310, 442; Prajñākaragupta’s theory of existence and 531–3, 534; types according to Dignāga 285–6; Yogācāra understanding of sense perception 275–6, 279–80 perfected nature (pariniṣpannasvabhāva) 69, 240, 365, 484; see also three natures Perfection of Wisdom Sūtras see Prajñāpāramitāsūtra person see pudgala Phule, J. 637, 647n2 physicalism 305–7; see also Cārvāka Plato 4, 9, 81, 211, 231n61, 231n62, 231n64 prajñā see wisdom Prajñākaragupta 528–39; compared to Berkeley 531; critique of Dharmottara 529; on the Buddha’s omniscience 533–5; on Dharmakīrti 320n3, 395, 528–39, 558, 559, 567, 569;theory of backward causation 529–31; theory of existence as defined by perception 531–3 Prajñākaramati 34, 199, 405, 475, 620 Prajñāpāramitāsūtras (Perfection of Wisdom Sūtras): Aṣṭasahāsrikaprajñāpārmitā (Perfection of Wisdom in Eight-Thousand Verses) 54, 227n16, 499, 501, 504, 616; Candrakīrti on 405; and development of Mahāyāna 212, 348–9; Haribhadra on 497, 500–1; Kamalaśīla’s commentaries on 489; Prajñāpāramitāhṛdaya (Heart Sūtra) 278, 489, 653, 654; Vajracchedikā Prajñāpāramitā (Diamond Cutter Perfection of Wisdom) 43n35, 66, 254, 489 prajñaptisat (nominally existent) 69, 286 prakāśa see luminosity prakṛti (pradhāna, Nature, primordial matter) 88, 90, 92–3, 455, 456 pramāṇa (means of knowledge, means of valid cognition, source of knowledge): Bhāviveka on 348–9, 364; Brahmanical theory of 289; Dharmakīrti on 309–12, 441; Dharmapāla’s use of 364; Dignāga on 284–5, 289–90; in Indian philosophy 311, 347; Jñānaśrīmitra on 559; Kamalaśīla on 485; Nāgārjuna as

undermining 220–1; nonexistence (abhāva) as 300n24; as non non-deceptive 457, 533; in Nyāya 310; Prajñākaragupta on omniscience as the ultimate pramāṇa 534; in Sāṃkhya 310; scripture and reason as implicitly accepted in the Katthāvatthu 170; see also inference; perception pramāṇaphala (result of a means of valid cognition, instrumental effect): Devendrabuddhi on 395–8; Dharmakīrti’s revision of Dignāga’s analysis 290–1; Dignāga on 289–90, 373n27 prapañca (conceptual proliferation) 213–14, 219, 410, 452, 588–90, 593, 596–8 prasajya-pratiṣedha (non-affirming negation) 130, 241, 408 Prāsaṅgika-Madhyamaka: Āryadeva characterized as 240; Candrakīrti characterized as 230n57, 232n77, 347–8, 476n4, 481; Dalai Lama as a follower of 651, 660; as a Tibetan doxographical distinction 491–2 pratijñā see thesis pratītyasamutpāda see dependent arising pratyakṣa see perception pudgala (person): associated with term “Pudgalavādin” by opponents 142; canonical texts on 189–90; compared to Brahminical understandings of ātman (self) 191; criticized by various Buddhist authors 196; Jain understanding as material and non-sentient 191; as neither the same nor different from the elements of being 189; tantric critique of 50–1; Vasubandhu on 274; Vātsīputrīya/Sāṃmitīya views of 191–2; see also pudgalavāda pudgalanairātmya (personal identitylessness, selflessness of persons) 50, 409, 410, 621, 623, 625 pudgalavāda (doctrine that asserts a person): critiqued by diverse Buddhist philosophers 197; as a pejorative label 191; refuted by Vasubandhu 274; refuted in the Vimalaprabhāṭīkā (Commentary on Stainless Light) 51; resources for study of 192, 193; see also Vātsīputrīya/Sāṃmitīya quietism: in Āryadeva 238–42, 247; in Nāgārjuna 223, 225, 227n15, 232n77 Ratnākaraśānti 587–600; doxographical classification of 590; on the four reliances 41n13; Instruction on the Perfection of Wisdom (Prajñāpāramitopadeśa) 286; ontological account of consciousness 590–5; on the purification of error 55; theory of error 587–9, 595, 596–7 Ratnakīrti 573–86; A Proof of Exclusion (Apohasiddhi) 578–82; A Theory of 674

Index sākāravāda (theory that cognition is endowed with an image): and the sopalambhaniyama argument 433; Dignāga’s adoption of 289; Jitāri on 554; Ratnakīrti on 582; terminology of 477n19, 598n3 Śākyabuddhi 393–403; on the causal account of perception 532; criticized by Jñānagarbha 452; criticized and relied on by Kamalaśīla 484, 485; on faith-based reasoning 246; relationship to Devendrabuddhi 393–5, 400; on mediated and unmediated instrumental effect 395–8; as part of the “philological school” of commentators 528; on reflexive awareness 398–402; Śubhagupta influenced by 433; see also Devendrabuddhi sāmānya/sāmānyalakṣaṇa see universal Saṃdhinirmocanasūtra (Scripture Clarifying the Intent): commentary attributed to Asaṅga 253, 254, 255, 256; Commentary on the Maitreya Chapter of 451; denigration of Madhyamaka implied in 367; and the Mahāyānasaṃgraha (Summary of the Mahāyāna) 259; presaging the disagreement between Dharmapāla and Bhāviveka 367; on the three turnings of the wheel of Dharma 277 saṃjñā see recognition Sāṃkhya 347, 464; Aśvaghoṣa’s critique of 83, 87–8, 89, 90, 93; Bhāviveka’s critique of 349, 352; Dharmapāla’s critique of 365; Vasubandhu’s critique of 267 saṃsāra: actions binding to 91, 176; awakening as transcendence of 21; the bodhisattva remaining in 212, 513, 515, 522; as fueled by grasping 280; Kamalaśīla on the sufferings of 490; nonduality of nirvāṇa and 55–7, 128–9, 218, 219, 230n61 saṃskāra see formations saṃvṛtisatya see conventional truth saṃyojana (fetter) 148, 176, 178 Śaṅkarasvāmin 115, 299n5, 549 Śāntarakṣita 463–79; Compendium of True Principles (Tattvasaṃgraha) 92, 463–5, 472; incorporation of Yogācāra ontology 464; influenced by Śūbhagupta 428–9; on Jñānagarbha 451, 453, 454, 456, 459; on the neither-one-nor-many argument 465–70; on non-seeing as the seeing of reality 456; Ornament of the Middle Way (Madhyamakālaṃkāra) 463, 465–7; synthesis of Madhyamaka and Yogācāra 472–5; tantric works attributed to 476n7; threefold criterion for conventional truth 470–2; use of scales of analysis 465 Śāntideva 511–27; Bodhicaryāvatāra (Introduction to the Practice of Awakening) 511, 512–16, 518–25, 653, 656–9; on bodhicitta and the bodhisattva’s vow 512–16;

Multifaceted yet Non-Dual Appearance (Citrādvaitaprakāśavāda) 578, 582–4; idealism of 582–4; influenced by Dharmottara 440; Proofs of Momentariness 574–8; on the three natures 286 Ravigupta 528, 529, 537n1 realism: antimetaphysical 226n12; critical 152; direct 465, 471, 477n19; metaphysical 236, 238, 246; naïve 506, 507; non-Buddhist 311, 364, 559; ontological 27, 243; representational 468–9, 470, 471, 477n19; robust 651; Sarvāstivāda 364; Sautrāntika 460; semantic 231, 238, 239, 296–8, 564, 571; scientific 652; theory of atoms 363; Vaibhāṣika 465; Yogācāra 622; see also antirealism recognition (saṃjñā, saññā; perception, classification): analyzed in the Vibhaṅga (Analysis) 148; Asaṅga on the Buddha’s transformation of 261; based on inherited patterns 272; of the Buddha (buddhasaṃjñā) 34–8, 40, 43n30; Sthiramati on 36; wrong (mithyāsaṃjñā) 182; see also five aggregates reductio ad absurdum arguments 142, 240, 349, 357n10, 406, 476n4, 544 reflexive awareness (svasaṃvedanā, svasaṃvitti; self-awareness, meta-cognition): Dharmakīrti’s revision of Dignāga’s theory 290–1, 373n28; Dharmapāla on 367–70; Dignāga on 285, 288–91, 369; Jñānagarbha as accepting conventionally 459; Kamalaśīla’s defense of 486; Mahāsāṃghika view on 369; Prajñākaragupta as reducing all perception to 532–3; Ratnākaraśānti on 592; Śākyabuddhi on 398–402; Sarvāstivāda Abhidharma view on 368; Vasubandhu as not embracing 269; see also luminosity religious authority: Ambedkar’s critique of divine authority 640; the Buddha as a means of knowledge (pramāṇabhūta) 227n22, 305, 533; Buddhaghosa on 331, 334–5; Buddhist skepticism about 246; Gotama Buddha as a 174; Maitreya as a 66; Prajñākaragupta on 533–5; Śāntarakṣita and Kamalaśīla on 487; Siddhārtha Gautama as a 308–9; tantric buddhas as 46, 52; see also buddhavacana; scriptural authority Ricoeur, P. 334 Rousseau, J.-J. 636, 638 rūpa see form śabda see verbal testimony sahopalambhaniyama (invariable co-apprehension): Dharmakīrti on 317, 400, 433; Jitāri’s critique of Dharmotta’s interpretation of 548–9; Jñānaśrīmitra on 567; Kamalaśīla’s defense of 429, 430, 433, 435; Śubhagupta on 433 675

Index contributions to moral philosophy 525; on generosity 516–18, 525; influence on Dalai Lama 653, 656–9; on patience 518–23, 525; on relationship between emptiness and the pāramitās 523; on relationship between metaphysics and ethics 523–5; on relationship between no-self and virtue 524, 525 Saraha 124–37; as anti-philosopher 124, 126–9; on four buddha bodies 134–5; influence of buddha nature discourse on 135; as mahāsiddha 124–6; as mystic philosopher 129–35; and Nietzsche 124, 129; on non-recollection (asmṛti) 132; rejection of caste 126; on subitism and gradualism 131–4; as teacher of mahāmudrā 125; on two truths 130–1 Sāriputta/Śāriputra 110n12, 144, 160, 200, 464, 488 Sarvāstivāda 142; *Abhidharmāṣṭagranthaśāstra (Abhidharma Treatise in Eight Sections) 173–5, 180–3, 276; Aśvaghoṣa’s association with 83–4; eight Abhidharma topics 176–9; emergence as a distinct group 173; Jñānaprasthāna (Source of Knowledge) 83, 173–5, 184; and the path of vision (darśanamārga) 176; position on past, present, and future entities 166–8, 173–4; rejection of the pudgala (“person”) 190; and the second council 183–4; Vijñānakāya (Consciousness Group) 166, 167, 170 satyadvaya see two truths satyākāravāda (theory that cognition is endowed with a true image): Jitāri as adherent of 549; targeted by Śāntarakṣita 468–9; as a Tibetan doxographical classification 477n19, 609; see also alīkākāravāda Sautrāntika: Aśvaghoṣa as anticipating 84, 93; basic doctrines of 286; as emerging from the Sarvāstivādins of Bactria and Gandhāra 175; on perception 289; provisional adoption by Śāntarakṣita 465 Sautrāntika-Madhyamaka 459–60, 491 Sautrāntika-Svātantrika-Madhyamaka 347, 476n4 scriptural authority: Abhayākaragupta on Mahāyāna Sūtras and 619–20; Buddhist philosophers skeptical of 246; Mahāyāna philosophers on 627n7; Tibetan “triple analysis” (dpyad pa gsum) method for 245; see also inference, scripturally-based; religious authority Self (ātman; soul): arguments against in the tantras 50; Aśvaghoṣa’s arguments against 90–2; fundamental question of philosophy 188; Nyāya view of 90; Saṃkhya view of 88, 90; Vaiśeṣika view of 89–90; Vasubandhu on 271–3; view that there is a self (satkāyadṛṣṭi) 178

semantics: Madhyamaka critique of realist 239; negative 298; realist theories of 296–7; unifying with epistemology and dialectics 291 Siddhārtha Gautama 27–45; as distinct from Gotama Buddha 10; life according to Aśvaghoṣa 85–87 single vehicle (ekayāna) 70, 484, 619, 625, 627 skandha/khanda see five aggregates Socrates 9, 107, 110n15 Spinoza, B. 506 Śrīgupta 470, 463, 464, 470, 474 Sthiramati 376–92; commentary on Vasubandhu’s Triṃśikāvijñaptikārikā (Thirty Verses on Appearance) 381–8, 459; critique of Madhyamaka 384, 385–6; figurative theory of language 382–8, 389n20; on the grasping of signs 36; as innovative commentator 377–8, 381–2; interpretation of imagination (rtog pa) 459; reading of Nāgārjuna 240 storehouse consciousness see ālayavijñāna Strawson, P. F. 242 Śubhagupta 427–38; as author of short digests as debate manuals 427, 429; demonstration of objects external to the mind 432–6; embrace of nirākāravāda 427, 432, 433; externalism rebutted by Śāntarakṣita and Kamalaśīla 429, 433; on omniscience 429; ontological externalism (bāhyārthavāda) of 427; quoted by Kamalaśīla 433, 434; refutation of Vijñānavāda 432–6; reliance on Dharmakīrti 431, 436; Śāntarakṣita and Kamalaśīla’s debt to 428–9; and Sautrāntika views 434–5; understanding of the image (ākāra) in awareness 435 substance (dravya) 90, 218, 286, 407, 433 suffering (duḥkha, dukkha): Ambedkar on social suffering 636, 640, 641–3; Aśvaghoṣa on relationship between suffering and the self 87, 91, 92, 94n31; Dalai Lama XIV on relationship to compassion 657, 658–9; Dharmakīrti on relationship to desire 305, 307–8, 319; in early Buddhist teachings 23; Kamalaśīla on relationship to compassion 483, 490; Kātyāyanīputra on 178, 180–3; Mahāyāna understanding of 588; in the Milindapañha (Questions of Milinda) 103, 104; Nāgārjuna on 214–15, 225, 228n40; Ratnākaraśānti on ignorance and 588, 596, 597; Śāntideva on causes and elimination of 513–15, 518–25, 526n2; Saraha on 130; the tantric buddha on 50–2, 59 śūnyatā see emptiness svabhāva/sabhāva (self-nature, own nature, intrinsic nature, particular nature): connection to emptiness 67, 219, 241, 351; critiqued in Sarvāstivāda Abhidharma 177–8, 180; lacking in what is dependently arisen 676

Index 218; in Mahāyāna philosophy 236; in Pali Abhidhamma 149, 151, 153, 189 svasaṃvedanā/svasaṃvitti see reflexive awareness Svātantrika-Madhyamaka: Maitreya’s Abhisamayālaṃkāra (Ornament of Clear Realization) classified as 67; Bhāviveka classified as 347–8, 476n4; Jñānagarbha classified as 450; Kamalaśīla classified as 481, 491–2; Śāntarakṣita classified as 463, 476n4, 491–2, 615; as a Tibetan doxographical distinction 491–2; see also Prāsaṅgika-Madhyamaka

two truths (satyadvaya): Abhayākaragupta on 620–3; Bhāviveka on 354–5; Candrakīrti on 405, 409–11, 412–17; Jñānagarbha on 450–8; Kamalaśīla on 484; Nāgārjuna on 230n57; Śāntarakṣita on 465–7, 472–5; in Saraha 130; see also conventional truth; ultimate truth Uddyotakara 90, 285, 318, 465 ultimate meaning (paramattha) 151–2 ultimate truth (paramārthasatya; ultimate reality): Bhāviveka on 352, 354–5, 366, 450, 453–4; Candrakīrti on 413, 415, 416–17; as ineffable (aparyāya) and inconceivable (acintya) 275, 453; Jñānagarbha on 450, 452–4; particulars as existing ultimately 311, 441; Ratnākaraśānti on 286; see also two truths; ultimate meaning universal (sāmānya, sāmānyalakṣaṇa, jāti; universal characteristic): Dharmakīrti on 312–14, 441; as conventionally existent (saṃvṛtisat) 311; Jitari’s critique of 550, 551, 553; as the object of inference 311; realist semantic theory of 296 universal momentariness 270, 440, 573

tantra 46–63; as contributing factor to the growth of apologetics 347; defined as the cause, result, and method for attaining buddhahood 47; focus on bliss and the body 46, 59–60; self-generation as a deity in 50; subtle body (sukṣmaśarīra) 125, 127, 132, 133, 135; Vimalaprabhāṭīkā (Stainless Light Commentary) 46, 48, 50–2, 58, 61 Tantric Buddha 46–62; having the world as their body 46; indistinct from one another 39; neither male nor female 46; as philosopher and doxographer 47–52; remaining active in the world 624 Tathāgatagarbha see Buddha Nature tetralemma see catuṣkoṭi Theravāda: Abhidhamma 141, 143, 149, 160, 161–9, 196; account of Vātsīputrīya/Sāṃmitīya 195; canon 199; and Cāttaṉār 117–18; Early European scholars of 192; influence of Buddhaghosa on 325, 328; as rejecting the doctrine of pudgala 192; see also Buddhaghosa thesis (pratijñā): Āryadeva’s rejection of any 239; Atiśa on the Madhyamaka rejection of 607; Dharmapāla on 364; Dignāga on 364; Nāgārjuna as eschewing 221–2; Vasubandhu as establishing through refutation of opponents 364–5 three natures (trisvabhāva): Abhayākargupta on 621; Āryadeva on 240; Asaṅga on 255, 256, 257, 259; Bhāviveka’s critique of 366; Dharmapāla’s response to Bhāviveka’s critique of 365–7; as framework for understanding Dignāga’s epistemology 285–8; Jñānagarbha on 453; Kamalaśīla on 484; Maitreya on 75n25; Ratnākaraśānti on 286, 602; three forms of naturelessness (niḥsvabhāvatā) 277–8, 484; and the three turnings of the wheel of Dharma 277; Yogācāra interpretation of non-seeing (adarśana) in light of 453 trisvabhāva see three natures Tsongkhapa 199, 242, 475, 476n5, 482, 616, 651

Vaibhāṣika: on atoms 433; Dharmapāla’s critique of 365; on perception 289; Śāntarakṣita’s critique of 465; subgroup of the Sarvāstivāda 184; Vasubandhu’s critique of 183, 270, 433; see also Mahāvibhāṣa; Sarvāstivāda Vaiśeṣika: Āryadeva’s critique of 236, 238; Aśvaghoṣa’s critique of 81, 89–90, 93; Bhāviveka’s critique of 346, 349, 353; Candrakīrti’s critique of 410, 414; Cāttaṉār’s critique of 119, 120; Dharmakīrti’s critique of 312; Dharmapāla’s critique of 364, 365; Dignāga’s critique of 285, 288; Śāntarakṣita’s critique of 465; Śubhagupta’s critique of 430 vāsanā (habitual propensity, imprint, latent impression): appearances arising from 389n17, 588, 591, 594; as contained by the ālayavijñāna 593; as contributing to the construction of a continuum (santāna) 442; of expressions 259; of hearing the Buddhist scriptures 259; of the intention to refrain from wrong deeds 379–80; in the tantras 52, 55, 62; of the view of the self 259 Vasubandhu 266–83; Abhidharmakośabhāṣya (Treasury of Abhidharma Commentary) 183, 266–8, 434; arguments against a self 268–70; commentary on works of Maitreya 68–70; on dependent origination 274; as engaging multiple inherited traditions 266–8; explanatory simplicity in 270; on language as figurative 274–81; as a Mahāyāna advocate 275; Pañcaskandhaprakaraṇa 677

Index Weber, M. 637 wisdom (prajñā, paññā; understanding): Bhāviveka’s definition of 354, 355; born of hearing or study (śrutamayīprajñā) 355, 508; born of meditation (bhāvanāmayīprajñā) 356, 508; born of reflection, or thought (cintamāyā, cintāmayīprajñā) 341, 355, 508; Buddhaghosa on 337–43; Śāntideva on the relationship between ethics and 523–5; as virtuous practice for a happy life 223 Wittgenstein, L. 231n64, 276 Wŏnch’ŭk 366

(Treatise on the Five Aggregates) 36, 378; as not providing a positive thesis 364; on reality as inconceivable 276, 280; on scriptural interpretation 273–4; on store consciousness 279–80; Vādavidhi (Method for Argumentation) 284, 285, 292, 299n2; Viṃśikā (Twenty Verses) 266, 276, 362–4, 433, 467, 476n7, 588; Vyākhyāyukti (Proper Mode of Exposition) 41n11, 70, 273, 489; on the world as mind only 274–6 Vatsīputrīya/Sāṃmitīya 188–203; and dependent arising 200; described by Xuanzang and Yijing 193; doctrines presented by Bhāviveka 194, 197; doctrines presented by Daśabalaśrīmitra 197, 198–9; doctrines presented by Vasubandhu 194; methodological issues in the study of 192–4; status as Buddhist or non-Buddhist 199–200; theory of the pudgala rejected by Theravāda 192; as umbrella name for diverse affiliated schools 191; see also pudgalavāda Veda: Bhāviveka’s reference to 353; Dharmakīrti’s critique of the authority of 309, 314; Jitāri’s critique of the authority of 551; Jitāri’s proof of human authorship of 553; Śubhagupta’s critique of the authority of 430, 431–2; tantric critiques of the authority of 48–9 vedanā see feeling Vedānta 349, 350, 353, 464, 465, 485, 653 verbal testimony (śabda): authority critiqued by some Buddhist philosophers 246; Jitāri’s refutation of 553; reduced to inference in some Buddhist epistemology 209, 284–5, 291, 464; see also scriptural authority Vibhajjavāda 154 views (dṛṣṭi): erroneous views (mithyādṛṣṭi) in Sarvāstivāda 179; as illusory in Yogācāra 277; Madhyamaka as the elimination of 218, 221, 350, 352, 354; Nāgārjuna on 218, 220–2; non-Buddhist speculative as tainted 350 vijñāna/viññāṇa see consciousness vijñapti (information, representation): Asaṅga on 255–6, 259–61; Sthiramati’s innovative understanding of 378–9 vijñaptimātra (representation only): Asaṅga on 260–1; Dharmakīrti on 316; Dharmapāla on 363; Jitāri’s proof of 551; Sthiramati on 459; Vasubandhu on 268, 275–7, 280; Yogācāra doctrine 588–9; see also mind-only Vimuktisena 196, 197, 501, 503, 618 viṣaya see object vyavahāra (everyday activity) 311, 395, 396, 414, 435, 533

Xuanzang: and Dharmapāla’s legacy in East Asia 370; disciples of 364, 365, 367; Records of the Western Regions 252; as source for studying Buddhist Vinaya schools 193, 194, 197; translator 185n12, 240, 252–3, 255, 299n5, 362, 361, 362–3, 388n3 Yāmari 528, 537n1, 558, 529, 572 Yaśomitra 434 Yijing 193, 194, 197, 361 Yogācāra: antirealist influence on Dharmakīrti 304; Asaṅga’s contributions to 255–61; Aśvaghoṣa and 84, 93; Atiśa as critic of 605; Candrakīrti’s critique of 405; criticized by Bhāviveka 352–3, 384; Dharmapāla’s contributions to 361–71; early authors on the matter-mental distinction 379; as a framework for understanding Dignāga 285–9; Haribhadra’s integration of 507; as holding there is nothing outside consciousness 286; influence on Saraha 135; Jitāri’s defense of 551, 554; Jñānagarbha’s critique of 453; Jñānagarbha’s use of 459–60; Kamalaśīla’s integration of 484, 486; metaphysical discourse of 386–8; ontological versus epistemological interpretation of 276; in contrast to Madhyamaka 385, 390n22; Ratnākaraśānti’s interpretation of 587–91; Śāntarakṣita’s provisional acceptance of 465, 470–2; Śāntarakṣita’s synthesis with Madhyamaka 472–5; Sthiramati’s formation of 376–88; Vasubandhu’s contributions to 274–81; Vasubandhu’s place in 267–8; see also ālayavijñāna; three natures Yogācāra-Madhyamaka 208, 240, 460, 464, 472, 474, 491 Yogācāra-Svātrantrika-Madhyamaka 476n4, 491, 615 yogic perception (yogipratyakṣa) 286, 299n11, 315, 444, 445, 483, 623, 624

678