Reasons and Empty Persons: Mind, Metaphysics, and Morality: Essays in Honor of Mark Siderits 3031139941, 9783031139949

Citation preview

Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36

Christian Coseru   Editor

Reasons and Empty Persons: Mind, Metaphysics, and Morality Essays in Honor of Mark Siderits

Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures Volume 36

Editor-in-Chief Purushottama Bilimoria, The University of Melbourne, Australia University of California, Berkeley, CA, USA

Series Editor Christian Coseru, College of Charleston, Charleston, SC, USA

Associate Editor Jay Garfield, The University of Melbourne, Melbourne, Australia

Assistant Editors Sherah Bloor, Harvard University, Cambridge, MA, USA Amy Rayner, The University of Melbourne, Melbourne, Australia Peter Yih Jiun Wong, The University of Melbourne, Melbourne, Australia

Editorial Board Balbinder Bhogal, Hofstra University, Hempstead, USA Christopher Chapple, Loyola Marymount University, Los Angeles, USA Vrinda Dalmiya, University of Hawaii at Manoa, Honolulu, USA Gavin Flood, Oxford University, Oxford, UK Jessica Frazier, University of Kent, Canterbury, UK Kathleen Higgins, University of Texas at Austin, Austin, USA Patrick Hutchings, Deakin University, The University of Melbourne, Parkville, Australia Morny Joy, University of Calgary, Calgary, Canada Carool Kersten, King's College London, London, UK Richard King, University of Kent, Canterbury, UK Arvind-Pal Maindair, University of Michigan, Ann Arbor, USA Rekha Nath, University of Alabama, Tuscaloosa, Tuscaloosa, USA Parimal Patil, Harvard University, Cambridge, USA Laurie Patton, Duke University, Durham, USA Stephen Phillips, The University of Texas at Austin, Austin, USA Joseph Prabhu, California State University, Los Angeles, USA Annupama Rao, Columbia University, New York, USA Anand J. Vaidya, San Jose State University, San Jose, USA

The Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures focuses on the broader aspects of philosophy and traditional intellectual patterns of religion and cultures. The series encompasses global traditions, and critical treatments that draw from cognate disciplines, inclusive of feminist, postmodern, and postcolonial approaches. By global traditions we mean religions and cultures that go from Asia to the Middle East to Africa and the Americas, including indigenous traditions in places such as Oceania. Of course this does not leave out good and suitable work in Western traditions where the analytical or conceptual treatment engages Continental (European) or Cross-cultural traditions in addition to the Judeo-Christian tradition. The book series invites innovative scholarship that takes up newer challenges and makes original contributions to the field of knowledge in areas that have hitherto not received such dedicated treatment. For example, rather than rehearsing the same old Ontological Argument in the conventional way, the series would be interested in innovative ways of conceiving the erstwhile concerns while also bringing new sets of questions and responses, methodologically also from more imaginative and critical sources of thinking. Work going on in the forefront of the frontiers of science and religion beaconing a well-nuanced philosophical response that may even extend its boundaries beyond the confines of this debate in the West  – e.g. from the perspective of the ‘Third World’ and the impact of this interface (or clash) on other cultures, their economy, sociality, and ecological challenges facing them – will be highly valued by readers of this series. All books to be published in this Series will be fully peer-reviewed before final acceptance.

Christian Coseru Editor

Reasons and Empty Persons: Mind, Metaphysics, and Morality Essays in Honor of Mark Siderits

Editor Christian Coseru Department of Philosophy College of Charleston Charleston, SC, USA Editor-in-Chief Purushottama Bilimoria

ISSN 2211-1107     ISSN 2211-1115 (electronic) Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures ISBN 978-3-031-13994-9    ISBN 978-3-031-13995-6 (eBook) https://doi.org/10.1007/978-3-031-13995-6 © Springer Nature Switzerland AG 2023 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland

Mark Siderits, Borobudur Temple, Central Java, 2012. Photograph by Esther Viros.

v

Acknowledgments

My sincere gratitude goes first to the outstanding group of scholars for their contributions, and for lending their support and expertise at different stages of this longin-the-making project. The call for a volume to celebrate the work of Mark Siderits came at a time when few if any of us conceived that a crisis of global proportions would alter our personal and professional lives in such dramatic and seemingly irrevocable ways. The immediate and enthusiastic response to that call provided unstinting support and encouragement as the project began to take shape in the wake of the coronavirus pandemic. In addition to thanking all the contributors for tailoring their essays in consideration of Siderits’ influential work, I would also like to thank the three anonymous referees who read the manuscript for Springer, and provided detailed and helpful comments. Grateful acknowledgement also to Routledge for permission to reproduce a slightly altered version of my chapter “Reasons and Conscious Persons”, originally published in A.  Sauchelli (2020), Derek Parfit’s Reasons and Persons: An Introduction and Critical Inquiry, London: Routledge. My heartfelt thanks go to Purushottama Bilimoria, Editor-in-Chief and architect of the Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures Series, for inviting me to be a part of the editorial team, and for considering the volume for this Series. I am also grateful to Christopher Coughlin, senior editor for philosophy at Springer Nature, for shepherding it through the final stages of production, and to the entire production team for their wonderful job in preparing the volume for publication. Lastly, I wish to thank my wife and colleague, Sheridan Hough, for her unwavering patience, enthusiastic support, and above all for helping this editorial endeavor achieve a measure of stylistic and thematic consistency. The original reason for this book is Mark Siderits himself, the person whose innovative work set in motion the many causes and conditions that led to its existence. Whatever one may think about the reality of persons, we ardently wish for the continued flourishing of his diligent scholarship, sharp analytic insights, and unyielding generosity.

vii

Contents

Introduction������������������������������������������������������������������������������������������������������    1 Christian Coseru Part I Mind  Aren’t We Conscious? A Buddhist Reflection on the Hard Problem���������   19 But Georges Dreyfus  Reasons and Conscious Persons���������������������������������������������������������������������   35 Christian Coseru  Descartes and the Buddha—a rapprochement?��������������������������������������������   63 Galen Strawson  The Imperfect Reality of Persons ������������������������������������������������������������������   87 Jonardon Ganeri  Eliminating Selves, Reducing Persons ����������������������������������������������������������  105 Monima Chadha and Shaun Nichols The Curious Case of the Conscious Corpse: A Medieval Buddhist Thought Experiment����������������������������������������������������������������������������������������  121 Robert H. Sharf  How Can Buddhists Account for the Continuity of Mind After Death?�������� 141 Jan Westerhoff What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy and Cognitive Science ��������������������������������������������  165 Evan Thompson On Necessary Connection in Mental Causation––Nāgārjuna’s Master Argument Against the Sautrāntika-Vasubandhu: A Mādhyamika Response to Mark Siderits������������������������������������������������������������������������������  211 Sonam Thakchoe ix

x

Contents

Part II Metaphysics A Post-Reductionist Buddhism? ��������������������������������������������������������������������  231 Matthew MacKenzie  Madhyamaka, Ultimate Reality, and Ineffability ����������������������������������������  247 Jay L. Garfield and Graham Priest The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits and I Really Disagree��������������������������������������������������������������  259 Dan Arnold Anti-Realism and Realism About the Past: A Present for Mark Siderits����������������������������������������������������������������������������������������������  283 Arindam Chakrabarti Buddhist Shipping Containers������������������������������������������������������������������������  295 Koji Tanaka  Finite and Infinite: On Not Making ‘Them’ Different Enough������������������  307 Rupert Read and Christian Greiffenhagen Part III Epistemology Mark Siderits on Anumāna ����������������������������������������������������������������������������  327 Shoryu Katsura  Buddhist Reductionism, Fictionalism, and Expressibility ��������������������������  345 Laura P. Guerrero  The Presupposition Strategy and the Two Truths����������������������������������������  363 John Spackman Part IV Morality  Ideas and Ethical Formation: Confessions of a Buddhist-Platonist������������  387 Amber Carpenter Self-Cultivation Philosophy as Fusion Philosophy: An Interpretation of Buddhist Moral Thought����������������������������������������������������������������������������  417 Christopher W. Gowans Can We Know Whether Śāntideva Was a Consequentialist?����������������������  437 Charles Goodman Selfhood and Resentment��������������������������������������������������������������������������������  459 Rick Repetti Publications of Mark Siderits ������������������������������������������������������������������������  477 Index������������������������������������������������������������������������������������������������������������������  483

Contributors

Dan Arnold  is Associate Professor of philosophy of religions at the University of Chicago Divinity School. He is the author of  Buddhists, Brahmins, and Belief: Epistemology in South Asian Philosophy of Religions  (2005), which won the American Academy of Religion Award for Excellence in the Study of Religion, and of  Brains, Buddhas, and Believing: The Problem of Intentionality in Classical Buddhist and Cognitive-Scientific Philosophy of Mind  (2012), which won the Numata Book Prize in Buddhism, both with Columbia University Press, and editor of Reasons and Lives in Buddhist Traditions, Studies in Honor of Matthew Kapstein (with Cécile Ducher and Pierre-Julien Harter, Wisdom, 2019). He is currently working on an anthology of original translations from India’s Madhyamaka tradition of Buddhist philosophy. Amber Carpenter  is Associate Professor of Philosophy at Yale-NUS College. Her research interests span the ancient Greek and Indian philosophical tradition, with a topical focus on the metaphysics, epistemology, and moral psychology underpinning Plato’s ethics and Indian Buddhist ethics. She is the author of Indian Buddhist Philosophy (Routledge, 2014) and of Portraits of Integrity (with Charlotte Alston, Rachael Wiseman, Bloomsbury, 2020). Her articles have appeared in Proceedings of the Aristotelian Society, Ratio, Phronesis, Ancient Philosophy, The Classical Review, History of Philosophy & Logical Analysis, Australasian Journal of Philosophy, and numerous edited volumes. Monima Chadha  is Associate Professor of Philosophy at Monash University. Her research interests are in the self, consciousness, and mind, which she studies from a cross-cultural and interdisciplinary perspective. Her papers have appeared in Journal of the American Philosophical Association, Australasian Journal of Philosophy, Mind & Language, Philosophy East and West, Philosophers' Imprint, Review of Philosophy and Psychology, Phenomenology and the Cognitive Sciences, Asian Philosophy, Sophia, European Journal of Philosophy, Consciousness and Cognition, and several edited collections.

xi

xii

Contributors

Arindam Chakrabarti  is Professor of Philosophy at the University of Hawai’i, Manoa. He earned his DPhil from Oxford University, working under Peter Strawson and Michael Dummett. He has taught at Calcutta University, University College London, University of Washington, Seattle, SUNY Stony Brook, and Delhi University. After being trained as an analytic philosopher of language, he spent several years receiving traditional training in Indian logic (Navya Nyāya). Chakrabarti has edited or authored ten books in English, Sanskrit, and Bengali, including Denying Existence (Springer 1997), Knowing from Words (edited with B.K.  Matilal, Springer 1994)), Universals, Concepts and Qualities (edited with Peter Strawson, Routledge 2006), Comparative Philosophy Without Borders (with Ralph Weber, Bloomsbury 2016), Realisms Interlinked (Bloomsbury 2019), and Thinking About Food and Clothing: Essays in Quotidian Philosophy (in Bengali) and has published more than one hundred papers and reviews. Christian Coseru  is Lightsey Humanities Chair and Professor of Philosophy at the College of Charleston. He works in the fields of philosophy of mind, phenomenology, and cross-cultural philosophy, especially Indian and Buddhist philosophy in dialogue with Western (classical and contemporary) philosophy. Author of  Perceiving Reality: Consciousness, Intentionality, and Cognition in Buddhist Philosophy (OUP, 2012), he is currently working on an introduction to Buddhist philosophy of mind, titled Moments of Consciousness (under contract with OUP). His papers have appeared in Analysis, Philosophical Studies, Phenomenology and the Cognitive Sciences, Philosophy East and West, Asian Philosophy, Sophia, Zygon, Journal of the Asiatic Society, Voprosi Filosofii, and numerous edited volumes, including the Oxford Handbook of Indian Philosophy (2017), the Routledge Handbook of Consciousness (2018), the Routledge Handbook on the Philosophy of Meditation (2022), the Blackwell Companion to Buddhist Philosophy (2013), Buddhist Philosophy of Consciousness (Brill 2021), and Cross-cultural Approaches to Consciousness (Blackwell 2022). His research has been supported by grants from the National Endowment for the Humanities and the Australian Research Council. Georges  Dreyfus  is Professor of Religion in the Department of Religion at Williams College. He works on issues in philosophy of mind, Buddhist philosophy, and theory and method in the study of religion and modernity. His publications include  Recognizing Reality: Dharmakīrti and his Tibetan Interpreters  (Albany: SUNY Press, 1997), The Svatantrika-Prasaṅgika Distinction (co-edited with Sara McClintock, Boston: Wisdom, 2003), and The Sound of Two Hands Clapping: the Education of a Tibetan Buddhist Monk (Berkeley: University of California Press, 2003), as well as many articles on various aspects of Buddhist philosophy and Tibetan culture. He was the first Westerner to receive the title of Geshe after spending fifteen years studying in Tibetan Buddhist monasteries. Jonardon Ganeri  is Bimal K. Matilal Distinguished Professor of Philosophy at the University of Toronto. His work draws on a variety of philosophical traditions to construct new positions in the philosophy of mind, metaphysics, and epistemology. He advocates an expanded role for cross-cultural methodologies in philosophical

Contributors

xiii

research, together with enhanced cultural diversity in the philosophical curriculum. His research interests are in consciousness, self, attention, the epistemology of inquiry, the idea of philosophy as a practice, and its relationship with literature. He works too on the history of ideas in early modern South Asia, intellectual affinities between India and Greece, and Buddhist philosophy of mind. He is the author of Virtual Subjects, Fugitive Selves: Fernando Pessoa and His Philosophy (OUP 2020), Inwardness: An Outsiders’ Guide (Columbia University Press 2021), Classical Indian Philosophy: A History of Philosophy Without Any Gaps, volume 5 (co-authored with Peter Adamson, OUP 2020), Attention, Not Self (OUP, 2017), Identity as Reasoned Choice (Bloomsbury 2014), The Self: Naturalism, Consciousness, & the First Person Stance (OUP 2012), The Lost Age of Reason: Philosophy in Early Modern India, 1450–1700 (OUP 2011), The Concealed Art of the Soul (OUP 2007), Artha: Testimony and the Theory of Meaning in Indian Philosophical Analysis (Delhi: OUP 2006), Philosophy in Classical India: The Proper Work of Reason (Routledge 2001), and Semantic Powers: Testimony and the Theory of Meaning (Oxford: Clarendon 1999) and editor of Oxford Handbook of Indian Philosophy (2017). Jay  L.  Garfield  is Professor and Chair of the Philosophy Department at Smith College, and Director of the Tibetan Studies in India program. He is also Visiting Professor of Buddhist Philosophy at Harvard Divinity School, Professor of Philosophy at Melbourne University, and Adjunct Professor of Philosophy at the Central Institute of Higher Tibetan Studies. His research addresses topics in the foundations of cognitive science and the philosophy of mind; metaphysics; the history of modern Indian philosophy; topics in ethics, epistemology and the philosophy of logic; the philosophy of the Scottish enlightenment methodology in cross-cultural interpretation; and topics in Buddhist philosophy, particularly IndoTibetan Madhyamaka and Yogācāra. He is the author or editor of over 30 books and nearly 200 articles, chapters, and reviews. Garfield’s most recent books are Losing Ourselves: Learning to Live Without a Self (2022), Knowing Illusion: Bringing a Tibetan Debate into Contemporary Discourse (with the Yakherds, 2021), Buddhist Ethics: A Philosophical Exploration (2021), ̛What Can’t Be Said: Paradox and Contradiction in East Asian Thought (with Yasuo Deguchi, Graham Priest, and Robert Sharf, 2021), The Concealed Influence of Custom: Hume’s Treatise from the Inside Out (2019), Minds Without Fear: Philosophy in the Indian Renaissance (with Nalini Bhushan, 2017), Dignāga’s Investigation of the Percept: A Philosophical Legacy in India and Tibet (with Douglas Duckworth, David Eckel, John Powers, Yeshes Thabkhas, and Sonam Thakchöe, 2016) Engaging Buddhism: Why It Matters to Philosophy (2015), Moonpaths: Ethics and Emptiness (with the Cowherds, 2015) and (edited, with Jan Westerhoff), Madhyamaka and Yogācāra: Allies or Rivals? (2015), all published by Oxford University Press. Charles  Goodman  is Professor of Philosophy and Director of the Program in Philosophy, Politics and Law at SUNY Binghamton, where he is also affiliate faculty in the Department of Asian and Asian-American Studies. His work emphasizes aspects of Buddhist thought that can offer valuable insights for the philosophy of

xiv

Contributors

today, particularly in applied ethics and political philosophy.  He is the author of Consequences of Compassion: An Interpretation and Defense of Buddhist Ethics (OUP 2009), Moonpaths: Ethics and Emptiness (co-author with the Cowherds collaboration, OUP 2016), and of the first complete translation of  Śikṣāsamuccaya in over ninety years, The Training Anthology of Śāntideva (OUP 2016). Christopher W. Gowans  is Professor of Philosophy at Fordham University. His primary research interest are on philosophical implications of moral diversity and conflicts within moral viewpoints and across cultures and traditions, contemporary normative moral philosophies, especially those rooted in the Aristotelian and Kantian traditions, and ancient (Hellenistic, Chinese, and Buddhist) philosophical traditions that understand and promote the attainment of well-being. He is the author of Self-Cultivation Philosophies in Ancient India, Greece, and China (OUP 2021), Buddhist Moral Philosophy: An Introduction (Routledge, 2014), Philosophy of the Buddha (Routledge, 2003), Moral Disagreements: Classic and Contemporary Readings (Routledge 2000), Innocence Lost: An Examination of Inescapable Moral Wrongdoing (OUP, 1994), and Moral Dilemmas (OUP 1987). Christian Greiffenhagen  is Associate Professor in the Department of Applied of Social Sciences at The Hong Kong Polytechnic University where he leads the Video Analysis, Science & Technology (VAST) Research Group, which develops videobased methodologies to study the impact of scientific and technological developments on people’s lives. His publications have appeared in The British Journal of Sociology, International Journal of Computer-Supported Collaborative Learning, New Media & Society, Methodological Innovations Online, Educational Studies in Mathematics, Visual Communication, Education, Communication & Information, Topics in Cognitive Science, Science & Education, Theory, Culture & Society, European Journal of Operational Research, History of the Human Science, International Journal of Social Research Methodology, Journal for the Theory of Social Behaviour, Configurations, Qualitative Research, Semiotica, Journal of Pragmatics, Journal of Classical Sociology, Language & Communication, The International Journal of Medical Robotics and Computer Assisted Surgery, Symbolic Interaction, and numerous conference proceedings. Laura P. Guerrero  is Assistant Professor of Philosophy at The College of William and Mary. Her research is primarily in the areas of Buddhist metaphysics and epistemology. She explores the distinctive approaches various Buddhist thinkers take to questions concerning what is real, what is true, and how we come to know. Her papers have appeared in Comparative Philosophy,  Ratio, and the  Journal of Consciousness Studies, and in several edited volumes. She is currently working on a book about Buddhist metaphysics for which she received a Robert H.N. Ho Family Foundation Program in Buddhist Studies Research Fellowship. Shoryu Katsura  is Professor Emeritus of Hiroshima and Ryukoku Universities. He studied Buddhist Philosophy at Kyoto University where he received his B.A. and M.A. He then entered the Ph.D. program of the University of Toronto and obtained Ph.D. Later he was granted the degree of D.Litt. by Kyoto University by his study

Contributors

xv

of the concept of pervasion (vyāpti) in Indian philosophy. He is the author of Indian Logic (Indojin no Ronrigaku) (Tokyo: Chūōkōronsha, 1998; Reprint: Kyoto: Hōzōkan, 2021) and edited Dharmakirti’s Thought and Its Impact on Indian and Tibetan Philosophy (Wien: Austrian Academy of Sciences, 1999) and The Role of the Example (dṛṣṭānta) in Classical Indian Logic (co-edited with Ernst Steinkellner, Wien: WSTB 2004). His translations include Nāgārjuna’s Middle Way (with Mark Siderits, Boston: Wisdom 2013), which received the Khyentse Foundation 2014 Prize for Outstanding Translation, and the complete Japanese translation of the Gaṇḍavyūha-sūtra (with Yūichi Kajiyama and others, Tokyo: Chūōkōronsha, 1994; Reprint: Tokyo: Iwanami Shoten, 2021). He is the recipient of Japanese Association of Indian and Buddhist Studies Award (1977) and Nakamura Hajime Eastern Academic Award (2010). Matthew  MacKenzie  is Professor and Chair of the Philosophy Department at Colorado State University. He specializes in Buddhist philosophy, classical Indian philosophy, and the philosophy of mind, with a special focus on consciousness, selfconsciousness, and embodiment. His articles have appeared in Philosophy East and West, Phenomenology and the Cognitive Sciences, Asian Philosophy, and Journal of Consciousness Studies, among others. He has published chapters in many edited volumes, including The Oxford Handbook of Virtue, The Routledge Handbook of Skill and Expertise, Routledge Handbook of the Philosophy of Meditation, and Feminist Philosophy of Mind.  He is the author of  Buddhist Philosophy and the Embodied Mind (Rowman & Littlefield 2022). Shaun Nichols  is Professor of Philosophy and Director of Cognitive Sciences at Cornell University. He works in philosophy of cognitive science, and his research concerns the psychological foundations of philosophical thought, in particular the acquisition of philosophically significant concepts and distinctions, especially in the domain of morality. He is the author of Rational Rules: Towards a Theory of Moral Learning (2021), Bound: Essays on Free Will and Responsibility (2015), Sentimental Rules: On the Natural Foundations of Moral Judgment (2004), and Mindreading: An Integrated Account of Pretense, Self-awareness and Understanding Other Minds (with Steven Stich, 2003), and editor of Oxford Studies in Experimental Philosophy (with Tania Lombrozo and Joshua Knobe, 3 vols. 2015-2020), Experimental Philosophy (2 vol., with Joshua Knobe, 2008, 2013), The Architecture of the Imagination: New Essays on Pretense, Possibility, and Fiction (2006), all published by Oxford University Press, and of La philosophie experimentale (with Florian Cova, Julien Dutant, Edouard Machery, Joshua Knobe, and Eddie Nahmias, Paris: Vuibert, 2012), and Moral psychology: Historical and Contemporary Readings (with Thomas Nadelhoffer and Eddie Nahmias, Wiley-Blackwell, 2010). Graham Priest  is currently Distinguished Professor of Philosophy at the Graduate Center, City University of New York, Boyce Gibson Professor Emeritus at the University of Melbourne, and International Research Fellow at the Ruhr University of Bochum. He is known for his work on non-classical logic, particularly in connection with dialetheism, on metaphysics, on the history of philosophy, and on Buddhist

xvi

Contributors

philosophy. He has published over 300 papers in nearly every major logic and philosophy journal. His books include: In Contradiction (1987), Beyond the Limits of Thought (1995), Logic: A Very Short Introduction (2000), Introduction to NonClassical Logic (2001), Towards Non-Being (2005), Doubt Truth to Be a Liar (2006), One (2014), The Fifth Corner of Four (2018), and Capitalism—Its Nature and Its Replacement (2021). For further details, see grahampriest.net. Rupert  Read  is Associate Professor of Philosophy in the School of Politics, Philosophy, Language and Communication Studies at the University of East Anglia. He works in philosophy of language (special focus on Wittgenstein), philosophy of the sciences (especially environmental and social sciences), ecological and political philosophy, and philosophy of film. He is the author of Wittgenstein’s Liberatory Philosophy (Routledge 2021), A Film-Philosophy of Ecology and Enlightenment (Routledge 2018), A Wittgensteinian Way with Paradoxes (Lexington 2012), Wittgenstein Among the Sciences (edited by Simon Summers, Ashgate 2012), Applying Wittgenstein (edited by Laura Cook, Continuum 2008), Philosophy for Life (edited by Matt Lavery, Continuum, 2007), There Is No Such Thing as Social Science: In Defence of Peter Winch (co-authored with Phil Hutchinson & Wes Sharrock, Ashgate, 2008), and Kuhn: Philosopher of Scientific Revolution (coauthored with Wes Sharrock; Oxford: Polity, 2002) and editor of Beyond the Tractatus Wars: The ‘New Wittgenstein’ Debate (with Matt Lavery, Routledge 2011), The New Wittgenstein (with Alice Crary, Routledge, 2000), and The New Hume Debate (with Ken Richman, Routledge 2000). Rick  Repetti  is Professor of Philosophy at CUNY/Kingsborough Community College, member of the Board of Directors of the American Philosophical Practitioners Association (APPA), and fellow of the Center for Contemplative Mind in Society. He is editor of the Routledge Handbook on the Philosophy of Meditation (Routledge 2022) and Buddhist Perspectives on Free Will: Agentless Agency? (Routledge 2017), and author of The Philosophy of Meditation: The Spoken Tao (forthcoming), Buddhism, Meditation, and Free Will: A Buddhist Theory of Mental Freedom (Routledge 2019), and The Counterfactual Theory of Free Will: A Genuinely Deterministic Theory of Soft Determinism (Lambert 2010). Robert H. Sharf  is D.H. Chen Distinguished Professor of Buddhist Studies in the Department of East Asian Languages and Cultures at the University of California, Berkeley, as well as Chair of Berkeley’s Center for Buddhist Studies. He works primarily on medieval Chinese Buddhism but has also published in the areas of Japanese Buddhism, Buddhist art and archaeology, Buddhist modernism, Buddhist philosophy, and methodological issues in the study of religion. He is author of  Coming to Terms with Chinese Buddhism: A Reading of the Treasure Store Treatise  (University of Hawaii Press 2002) and What Can Be Said (with Yasuo Deguchi, Jay L. Garfield, Graham Priest, OUP 2021) and co-editor (with his wife Elizabeth) of  Living Images: Japanese Buddhist Icons in Context  (Stanford University Press 2001).

Contributors

xvii

John Spackman  is Associate Professor of Philosophy and Director of the Program in Neuroscience at Middlebury College. He received his Ph.D. in Philosophy from Yale University, and also completed an M.A. at Columbia University in Religion. His work focuses on contemporary philosophy of mind and its intersections with Buddhist philosophy. He has published articles on such topics as debates concerning the nonconceptuality of perception and aesthetic experience, the structure of concepts, Nāgārjuna, and what contemporary work on the mind-body problem can learn from Buddhist philosophy. He is currently working on a book that develops an approach to the mind-body problem inspired by Madhyamaka Buddhist philosophy. He is the editor of Buddhist Philosophy of Consciousness: Tradition and Dialogue (with Mark Siderits and Ching Keng, Brill 2021). Galen  Strawson  holds the President's Chair in Philosophy at the University of Texas at Austin's College of Liberal Arts. He  taught at the University of Oxford from 1979 to 2000, where he was a Fellow of Jesus College. He was Professor of Philosophy at the University of Reading, UK from 2001 to 2012, and Distinguished Professor of Philosophy at the City University of New York Graduate Center from 2004 to 2007. He has held visiting research positions at the Research School of Social Sciences, Australian National University, Canberra (1993, 2012); the University of Copenhagen (2003, 2011); Princeton University, where he was a Council of the Humanities Old Dominion Fellow (2011); and the École des hautes études en sciences sociales in Paris (2012). He has taught as a Visiting Professor at NYU (1997), Rutgers University (2000), MIT (2010), and Princeton (2011). He is the author of  Freedom and Belief  (Oxford 1986, 2nd edition 2010),  The Secret Connexion: Realism, Causation and David Hume  (Oxford 1989),  Mental Reality  (MIT Press 1994, 2nd edition 2009),  Selves: An Essay in Revisionary Metaphysics (Oxford, 2009), The Evident Connexion: Mind, Self and David Hume (Oxford, 2011, revised paperback edition 2013), and Locke on personal identity  (Princeton 2011, revised paperback edition 2014). He is the principal author of  Consciousness and Its Place in Nature, ed. A.  Freeman (Imprint Academic, 2006). A selection of his philosophical papers, Real Materialism and Other Essays, was published in 2008 (Oxford). Koji  Tanaka  is an Australian Research Council Future Fellow in the School of Philosophy, Research School of Social Science, at the Australian National University. His main research areas are logic, philosophy and history of logic, philosophy of language, and Buddhist philosophy. His current project aims to advance a new theory about the nature of logic and modality, which draws on Buddhist philosophical material. His articles have appeared in Noûs, The Australasian Journal of Logic, Philosophy East and West, Synthese, Australasian Journal of Philosophy, Journal of Chinese Philosophy, and several edited volumes. Sonam  Thakchoe  is Senior Philosophy Lecturer at University of Tasmania and Director of the Tasmanian Buddhist Studies in India Exchange Program. His works in Indo-Tibetan Buddhist philosophy, with a particular focus on the Buddhist metaphysics, epistemology, ethics, philosophy of mind, and the phenomenology of

xviii

Contributors

mindfulness. He is the author of The Two Truths Debate: Tsongkhapa and Gorampa on the Middle Way (Wisdom 2007) and co-author of Moonshadows. Conventional Truth in Buddhist Philosophy (with the Cowherds, OUP 2011), Moonpaths: Ethics and Emptiness (with the Cowherds, OUP 2016), and Knowing Illusion: Bringing a Tibetan Debate into Contemporary Discourse (2 vol., with the Yakherds, OUP 2021). Evan Thompson  is Professor of Philosophy at the University of British Columbia, Vancouver, where he is also an Associate Member in the Department of Asian Studies and the Department of Psychology. He works on the nature of the mind, the self, and human experience. His work combines cognitive science, philosophy of mind, phenomenology, and cross-cultural philosophy, especially Asian philosophical traditions. He is the author of Colour Vision (Routledge, 1995), Mind in Life: Biology, Phenomenology and the Science of the Mind (Harvard University Press, 2007), Waking, Dreaming, Being (Columbia University Press, 2015), and Why I Am Not a Buddhist (Yale University Press, 2020) and co-author of The Embodied Mind (with Francisco Varela and Eleanor Rosch, MIT Press 1991, revised edition 2017). Jan Westerhoff  was educated at Cambridge and the School of Oriental and African Studies. He has taught Philosophy at the Universities of Oxford and Durham and is presently Professor of Buddhist Philosophy at the University of Oxford, and a Fellow of Lady Margaret Hall, University of Oxford. His books include Ontological Categories (2005), Nāgārjuna's Madhyamaka (2009), The Dispeller of Disputes (2010), Twelve Examples of Illusion (2010) Reality. A Very Short Introduction (2011), The Golden Age of Indian Buddhist Philosophy (2018), and The NonExistence of the Real World, all published by Oxford University Press.

Introduction Reason’s Myriad Way: In Praise of Confluence Philosophy Christian Coseru

Mark Siderits’ confluence approach to philosophy, first sketched in his landmark monograph, Personal Identity and Buddhist Philosophy (2003), is emblematic of what has arguably become the most influential way of engaging historically and culturally distant Buddhist thinkers and texts systematically and constructively. For nearly half a century, and rather fittingly for someone enthralled by Madhyamaka, Siderits has successfully charted a middle ground between the text-based, exegetical approach to Buddhist philosophy still dominant in many parts of Europe and East Asia and the methods of contemporary Anglophone analytic philosophy. Indebted to both, yet unconstrained by either, the confluence approach represents Siderits’ unique brand of historically-informed systematic reflection, delivered in the characteristically forceful and tightly argued prose that defines his inimitable style. ‘Confluence’––it should be noted––replaces ‘fusion’, his initial word choice, which, as Siderits himself has recently admitted, failed to catch on, “perhaps because it is too reminiscent of the marketing techniques behind terms like ‘fusion cuisine’” (Siderits 2017, 76). More importantly, ‘confluence’ marks a departure from the ‘comparative’ approach and the idea that the best way to do philosophy across cultural boundaries is by tagging arguments and theories in one tradition on the basis of their resemblance to, or difference from, those found in others. The problem with such comparisons is that they often stop short of critical engagement, something that even its champions acknowledge (although often less as a shortcoming, and more as pretext for claiming that historically and culturally distant figures and texts ought to be understood first in their own terms if one is to avoid the perils of one-­ sided comparisons, the sort that take one of the compared traditions to function as a

C. Coseru (*) College of Charleston, Department of Philosophy, Charleston, SC, USA e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_1

1

2

C. Coseru

tertium comparationis, a third, generic frame of reference for evaluating the other) (Chakrabarti and Weber 2016, 6). As Siderits’ exemplary work reminds us, the philosopher’s task is first and foremost to show that and how we can make progress toward solving philosophical problems by pursuing certain lines of inquiry. For those working at the intersection of multiple spaces of meaning that means considering “what traditions different from our own have had to say about the issues with which we are concerned” (Siderits 2003, 1). But doing so requires that we adopt a dialogical stance and regard philosophical inquiry as an open-ended process of asking questions and pursuing knowledge, rather than as an exploration of fully developed doctrinal structures. The point here is not whether the Buddhist tradition offers any real solutions to our own problems (or vice versa), but whether adopting a broader frame of reference (premised on the binocular advantage principle that two eyes are better than one) can bring a greater measure of clarity about some philosophical problems no matter their origin.

1 Observations of a Life in Buddhist Letters Like many others in his generation, Siderits encountered Buddhism during his high school years while exploring religious traditions beyond the familiar one in which he was raised. Drawn at first to Zen Buddhism through the works of Alan Watts and especially D.T.  Suzuki, whose influential lectures at Columbia University in the 1950s ignited the American Zen boom and its celebration by the Beat movement, Siderits eventually realized that even a modest grasp of Buddhism’s central tenets required proper philosophical training. Although D.T. Suzuki’s work emphasized the compatibility of Buddhism with modern science, the popular image of Zen Buddhism in the 1960s was that of an anti-intellectual tradition extolling the value of non-conceptual intuition and of meditative practice as the sole path to achieving it. It was at Cornell, where Siderits began to study for an undergraduate degree in physics at first, that his attention turned to philosophy. While taking a class with Sydney Shoemaker, which had Wittgenstein’s Blue Book among its assigned readings, Siderits first figured there might be a better way to make sense of the cardinal no-self doctrine than Zen’s paradoxical answers, which had appealed to so many Westerners disillusioned with their institutions and traditional mores in the post World War II period. But whereas Cornell instilled in the young Siderits an appreciation for the rigors of analytic philosophy, the absence of offerings in non-Western philosophy left him hungry for just that. Thus came the decision to pursue an undergraduate degree at a place where he could study both Asian and Western philosophy. Siderits’ early journey into academic philosophy is emblematic of the rather tortuous path an enthusiastic student of non-Western philosophy had to follow half a century ago in order to succeed: leaving Cornell for the University of Hawai’i, Siderits would also take up the study of Sanskrit and Indian philosophy, eventually returning to the East Coast to complete his doctoral dissertation at Yale in 1976 on

Introduction

3

the relation between compassion (karuṇa) and emptiness (śūnyatā) in the works of Nāgārjuna and Nishida Kitaro. Of all his graduate school mentors, Siderits credits in particular his thesis advisor, John Edwin Smith, whose expertise in German idealism helped him understand Nidhida’s own forays into that tradition, and the Platonist Robert S. Brumbaugh, in whose effort to make classical Greek texts relevant to modern philosophical analysis he found inspiration for his own methodological problems. If scholarship is concealed autobiography, one can well speculate that conversations with Kenneth Inada and Karsten Harries, among others whom Siderits acknowledges as influences during his formative years at Yale, might have had something to do with his decision eventually to abandon the East Asian angle of his studies. Inada’s 1970 translation of the Mūlamadhyamakakārikā, emphasized the need for understanding Nāgārjuna’s philosophy in its classical Indian context and with regard mainly to the teachings of the historical Buddha, rather than as the work of a dialectician or metaphysician (in the Western sense) committed to some version of absolute monism or nihilism (Inada 1970, 21). As a consequence, Inada may not have been particularly encouraging of a project dedicated to establishing whether Nāgārjuna’s śūnyatā provides good ground for determining the scope of Nishida’s conception of absolute nothingness. For his part, Karsten Harries is well known as among the first to reflect critically on Wittgenstein’s transcendental stance on language in light of Heidegger’s view about the inadequacy of everyday language as a source of meaning (Harries 1968). Hence, Harries (who took seriously Heidegger’s appeal to poetry as more suitable for articulating our existential condition) likewise may not have found particularly promising a project dedicated to an unabashedly transcendental treatment of language, least of all one inspired by Nishida’s claim that being and value are ultimately grounded in nothingness. The final pivot away from East Asia, however, happened a few years after Siderits completed his Ph.D. and moved to California to take up a visiting position at Sonoma State University. In 1979 he attended an eight-week NEH summer seminar on Nyāya and its theory of epistemic instruments (pramāṇa) directed by J.  N. Mohanty. This was a watershed intellectual moment, following which he decided to focus exclusively on the study of Buddhist philosophy in its original Indian context. A year later he accepted an offer from Illinois State University where he would teach for twenty-eight years until his retirement in 2008. In a felicitous and rather surprising, though not altogether unexpected, turn of events East Asia beckoned once more, this time with a professorship in philosophy at Seoul National University. Siderits welcomed the opportunities of this late career move, and the chance to strengthen connections with colleagues from across the region. Although his tenure at Seoul National University, where he taught until 2012, and occasional lectures at Kyoto University, where he remains a regular visitor, have not rekindled his early interests in Zen and Nishida, they have nevertheless led to many fruitful collaborations, most notably a landmark translation of Nāgārjuna’s Mūlamadhyamakakārikā with Shōryū Katsura, which won the 2014 Khyenste Foundation Translation Prize.

4

C. Coseru

In an earlier attempt at sketching Siderits’ intellectual biography, Jan Westerhoff (2016, 2) noted three conceptual strands that seem to anchor much of his work: first, a recognition that Buddhist philosophy in India is not an exclusively intramural affair but rather develops in close dialogue with its non-Buddhist critics, in particular the Naiyāyikas, who serve as their main interlocutors and whose descriptive metaphysical project provides sharp contrast for the characteristically revisionist metaphysics the Buddhists put forward; second, a preoccupation with some aspects of Wittgenstein’s views concerning the limits of language, specifically about whether it is the form of propositions that determines the shape of reality rather than other way around, and what that might mean for the question of the independent existence of objects, states of affairs, and facts––an issue that is integral to Siderits’ attempt to spell out anti-realism as a rejection of the very notion of independent existence (and its consequences for the Madhyamaka critique of the corresponding notion of svabhāva); and lastly, the adoption of a naturalist perspective that is in keeping with Siderits’ view (also Mohanty’s) that, unlike the internalism that has dominated Western epistemological inquiry in the West since Descartes, epistemology in India, particularly with regard to examining the nature of veridical cognition, developed on an externalist and reliabilist framework. To these three conceptual strands we may add a fourth, centered on a set of interrelated questions about the compatibility of Buddhism’ anti-essentialist metaphysics with its ethical teachings. Is the Buddhist no-self view simply a theoretical construct derived from metaphysical considerations about agency and causality or does it actually reflect our actual existential condition? Siderits has invariably claimed that reductionism about selves is compatible with the pursuit of a moral path even though he acknowledges that the conventional practice of morality (to which the Buddha offers precepts, inspiring tales, and rules of conduct) and Buddhist metaphysical doctrine are actually in conflict (Siderits 2008). Dubbed ‘Buddhist Paleocompatibilism’, this fourth conceptual strand is particularly noteworthy as an illustration of, as he puts it, “the sort of thing that can happen when we bring two distinct philosophical traditions into conversation with one another” (Siderits 2016, 249). What makes it noteworthy is that this is not a view any Indian Buddhist philosopher (as far as we know) actually held. Nor is it a view found in the Western tradition. Rather, it is one that could have been held had Buddhists confronted the (Western) problem of determinism and moral responsibility (that is, the problem whether the sort of freedom necessary for moral responsibility is compatible with determinism). As it happens, they did not. Instead, they faced the somewhat related problem of reconciling karmic justice with the no-self doctrine. ‘Buddhist paleocompatibilism’ functions as a strategy aimed at showing how holding persons morally responsible for their actions can be reconciled with a view of persons as reducible to their parts if: (a) ‘moral responsibility’ is taken to be a property of persons and (b) ‘determinism’ is taken to concern their parts. Siderits agrees with Locke and Rousseau that moral responsibility cannot operate without a conception of freedom as something that pertains to the person as a whole rather than their parts (e.g., it is persons rather than their hands, feet, hearts, lungs, and brains that form the objects of praise and blame). But unlike the classical

Introduction

5

compatibilist who sees freedom as a property of agents rather than their constitutive parts (e.g., the heart’s blood-pumping function is determined by the laws of nature), the paleocompatibilist thinks the idea of a free agent cannot be rescued if agent causation and event causation serve as competing explanations for the occurrence of an event. Why? Because––as Siderits claims––event causation is always poised to win given its ability to explain responsibility for an action as the effect not of the agent’s will, but of a prior event pertaining to some aspect of the agent (her character, nature, upbringing circumstances, etc.). In keeping with the Buddhist two truths doctrine, Siderits thus locates agent causation at the conventional level and event causation at the level of ultimate truth. Semantic insulation between the two levels of discourse, he maintains, allows us to talk of persons as moral agents responsible for their actions. Ultimately, however, there are no such things as persons but only event causation consisting “in the relation of universal concomitance and ordered succession between elementary event-types” (Siderits 2016, 257). As Siderits rather candidly acknowledges in a postscript to his essay on Buddhist  paleocompatibilism, he is not actually sure whether the theory is true. Besides (and as with many of his other attempts to forge new concepts and theories by drawing on Buddhist materials), that is not his main concern. Rather, the point is whether the theory has some degree of plausibility, which, if it does, “should count as evidence that philosophical progress can sometimes occur when distinct traditions enter into conversation” (Siderits 2016, 263).

2 The Confluence Approach: Genealogy, Problem-solving, and the Challenge of Naturalism What, then, are some of the distinctive virtues of the confluence approach that sets it apart from other attempts to do philosophy across cultural boundaries? And what lessons does Siderits’ work and legacy hold for those weighing the merits of this approach over other ways of engaging Buddhist philosophy? First, unlike comparing and contrasting, the confluence approach remains faithful to the dominant conception of philosophy as an intellectual enterprise centered on dialogue and argumentation, in which philosophers pursue unresolved problems by building on the achievements of their acknowledged forbears. The key term here is acknowledged, for philosophy, like other disciplines with a long and venerable ancestry, has its own canonical figures and texts. So, while the confluence approach recognizes philosophy’s complex heritage, it also opens up the possibility that critical engagement with ideas from one tradition in light of another can have both illuminating and destabilizing effects (in much the same way that flow dynamics can take on both concordant and discordant forms where two  stream join one another). Siderits’ innovative and skillful deployment of the methods and conceptual tools of analytic philosophy in pursuing certain interpretive and reconstructive angles has had precisely such an effect: showing, for instance, how a seemingly intractable scholastic debate about the scope of Madhyamaka metaphysics can

6

C. Coseru

benefit from a semantic interpretation, or how compatibilism can offer a solution to reconciling the reductionist view at the heart of Abhidharma Buddhism with the conventional practices of morality. But taking  this approach also means that one possible consequence of adopting a semantic interpretation––for the doctrine of emptiness––is the rather startling notion that the ultimate truth is that there is no ultimate truth. More challenges emerge: for a tradition of thought committed to the view that ultimately there are no (self-like) bearers of moral responsibility, Siderits’ paleocompatibilist solution cannot rescue the Buddhist Reductionist from mereological nihilism. This sort of outcome can be unsettling for those who cannot conceive of Buddhism as anything other than a source of benign insights or who regard its truth claims as valid because ultimately, they are said to be grounded in the transformative experience of the Buddha’s enlightenment. Indeed, as Siderits himself has noted on more than one occasion, the confluence approach is not without its critics, in particular those who think that “Asian and Western traditions have their respective places in two distinct cultures, and that comparison and contrast are consequently the best we can hope for” (Siderits 2017, 76). Of course, it is perfectly understandable why some might worry that confluence can have this diluting and deflating effect if Buddhist ideas, arguments, and views become nothing more than prized items in the analytic philosopher’s toolbox  or, conversely, if  science and scientific findings relevant to philosophy are treated (by Buddhist apologists) as somehow ‘Western’ or specific to the European tradition. For such interventions can end up treating the target tradition as a standing reserve of intellectual skill to be mined as one sees fit, typically in the service of one’s scholarly interests (Coseru 2018, 9–12). Language, of course, is another factor. The adoption of English as lingua franca (so to speak) for Buddhist philosophy can be (some might say, has been) alienating, no less and in some ways perhaps even more so than that of other non-Indic languages Buddhism has adopted as it spread beyond the Indian subcontinent. For such adoption often means operating with a new conceptual vocabulary, in the case of English one shaped by the Greek culture of first millennium BCE, the scholasticism of the Latin Middle Ages, and the predominantly French and German intellectual movements of early modern Europe. In short, one cannot write Asian philosophy in English without importing some of the tacit assumptions of Western conceptual schemas and their metaphysical underpinnings. That raises the complicated question whether the tenets of one tradition can be sustained in another, and at what cost to their original intent. It is only natural to wonder, then, how a conception of philosophy as constrained by its own genealogy could possibly be open to, let alone handle, perspectives and argumentative strategies that are not part of its received canon or articulated in its own recognizable idiom. After all, the charge of parochialism, often levelled against Western philosophy more broadly, reflects this failure to acknowledge the degree to which the genus philosophia actually spans the globe. But the genealogical orientation at the heart of the confluence approach is not about retracing philosophy’s diverse progression across the world, however worthwhile such a project might be.

Introduction

7

Rather, in counterpoising different philosophical traditions the goal is to address (in the hope of solving) perennial problems of philosophy regardless of their mixed ancestry. As with step relatives, our philosophical forbears need not all carry the same genetic blueprint in order to serve as familial sources of wisdom, argumentative skill, and insight. Confluence philosophy thus offers an alternative path to the comparative approach. As such, it recognizes that no matter its geographical location, historical period, or language philosophical inquiry lays claim to universality, and that all its problems and solutions are therefore fair game for critical scrutiny and analysis. In short, when it comes to the confluence approach, there are no holy cows. Of course, critical scrutiny entails  self-scrutiny and willingness to forego, on pain of irrationality, presuppositions that are known to lack any evidence, reason, or justification. That such a boldly systematic approach has shaped the contemporary conversation on Buddhist philosophy in profound and significant ways comes as no surprise. One emblematic aspect of this type of work, clearly evident in his  collected essays (Siderits 2016), is Siderits’ resolve in brief but illuminating postscripts to set the record straight by separating those ideas, theories, and interpretations that have stood the test of time from those that have not––a true example of intellectual humility occasionally punctuated by brief moments of gleeful vindication. Second, like blues rock and Tex-Mex, confluence philosophy implements a syncretic and creative approach to doing philosophy by drawing on, in this case, Indian and Buddhist sources while using the tools and conceptual resources of analytic philosophy. It is an approach meant to add flavor to an intellectual practice that, despite its rigor (or perhaps because of it), is often tedious, inexpressive, and for the most part bereft of existential concerns. If this approach is yet to garner widespread support in the academy that is because—and perhaps Siderits would agree—mainstream Anglophone analytic philosophy continues to source its ingredients from works produced under a strict set of constraints. They must (i) embrace a sort of Moorean ‘common sense’, (ii) use logical analysis as their main methodology, and (iii) consist of nothing more than––as Peter Unger puts it in Empty Ideas: A Critique of Analytic Philosophy 2014––“questions about whether or not certain conceptual connections hold, or certain semantic relations obtain, between certain words, or certain concepts, and certain other words, or concepts” (Unger 2014, 3). Siderits is mindful of these constraints, in particular the third, given his semantic interpretation of the Madhyamaka idea of emptiness. Yet, as his systematic and syncretic analyses demonstrate, doing analytic philosophy with Buddhist ideas and arguments (as opposed to merely writing about them) can sometimes cast perennial problems in a new light and lead to valuable insights. Perhaps owing to him having encountered academic philosophy for the first time at Cornell––a place where proponents of scientific realism such as Sydney Shoemaker and Nicholas Sturgeon began to challenge the instrumentalist anti-realism prevalent during the positivist era––it is also important to recognize the extent to which Siderits’ attempt to spell out both Buddhist Reductionism and the doctrine of emptiness is informed by a naturalist stance. Although his naturalist sensibilities, it seems, have always pulled in the direction of reductive physicalism, Siderits has

8

C. Coseru

found anti-realist arguments against the possibility of capturing the structure of the world as it is independently of our conceptual activity, specifically those based on semantic considerations, to be more compelling. When I first met him––in Calcutta (now Kolkata) in the winter of 1995––he was still hard at work disentangling the Madhyamaka and Advaita Vedānta systems, all the while pondering whether T.R.V. Murti––the author of an influential mid twentieth century book on Madhyamaka––had been right in his assessment that “the Mādhyamika does not deny the real; he only denies doctrines about the real” (Murti, 1955, 218). Incidentally, Murti’s extraordinary claim––that Nāgārjuna had profoundly revolutionized Indian thought in much the same way that Copernicus and Kant transformed astronomy and metaphysics––had gotten an entire generation of young American scholars in the 1960s and 70s off LSD and into the study of Sanskrit, the true gateway to antirealism. In an essay from that period (“Madhyamaka and Naturalized Epistemology”), Siderits welcomes the naturalist turn in contemporary epistemology while contending that its critique of foundationalism and internalism does not go far enough. Against such influential interpreters of Indian philosophy as Mohanty and Matilal, who thought that something akin to a Cartesian internalist epistemological project is also part and parcel of the Indian epistemological tradition (as observed in certain strands of Vedānta, Mīmāṃsa, and Yogācāra thought), Siderits offers a starkly different picture: “The very ideal of epistemology as pramāṇavāda––determining the number and nature of the pramāṇas or reliable means of belief formation––suggests a nonfoundationalist and externalist project. Non-foundationalist because the project presupposes the existence of knowledge, instead of seeking to prove its very possibility. And externalist because it seeks to distinguish between veridical and non-veridical states of the subject in terms of causal factors, and not in terms of states that are necessarily accessible to the subject” (Siderits 2016, 238). While Siderits’ assessment of the scope of epistemological reflection in India rings true for most scholars of that tradition, his claim that the veridical/non-­ veridical distinction is not premised on mental states that are accessible to the subject has been less warmly received, in particular by those who take the Dignāga and Dharmakīrti tradition to offer a more plausible explanation of our epistemic predicament. On this account, while it is true that, say, we can only visually comprehend what we are capable of attaining by means of sight, what we see (or, rather, the event of seeing) cannot be properly dissociated from the act itself and its sensorimotor contingencies. Hence, the Buddhist epistemologist’s contention that we sense not only the property particulars with which we come into contact but also apprehend ourselves (or––in keeping with the no-self view––our episodic cognitive states) as the locus of qualitative experience. Siderits worries that even if it were true that cognitive states possess this unique reflexive property of making themselves (and not merely their contents) know as occurring within this or that mental stream, that can only tell us why we are justified in believing something to be the case, not how the justification for this belief was formed. In other words, veridical access to our own mental states is a function of cognition being a product of reliable causes, not of it standing in the right sort of relation to its mode of presentation (though, if the latter, merely as a byproduct of

Introduction

9

the same impersonal chain of causation). As he puts it, “When I see water I may be said to perceive H2O even if I do not know the chemical composition of water and am thus unable to say that what I see is H2O, for being H2O is intrinsic to water” (Siderits 2016, 241). The claim here is that without understanding the atomic composition of things (or, in Abhidharma terms, their existence as mere trope occurrences), in other words without recourse to science, we are in no position to ascertain whether our justification for taking perception to be a reliable source of knowledge is warranted. Given that many, if not most, of our mundane experiences lack this sort of justification, can we claim to have any knowledge at all in these circumstances? One cannot but wonder whether this view of knowledge is not somehow overly restrictive and perhaps a little impractical considering just how much of the skillful ways in which we navigate the world it rules out (e.g, while an understanding of atmospheric science is a bonus, it seems most Sherpas can successfully summit Everest on inherent adaptations and mountaineering skill alone). But Siderits does not think we should shy away from exploring the implications of this physicalist stance, even if a Buddhist reduction of consciousness might confront us with the spectrum of a Robo-Buddha––something that we must acquiesce to if we are persuaded that neither an experiencing subject nor an inner subjective realm can be part of the Buddhist’s ultimate ontology. Third and last, and in keeping with the title of this volume in honor of his work, is an issue the confluence approach lays bare when the traditions brought into dialogue are on a diverging path on certain matters. Contemporary philosophy, at least in the Anglophone world and its satellites, is an avowedly secular enterprise, deferring to the sciences for its account of what can be said to exist at the most fundamental level and for its understanding of the brain-based processes that realize cognition. Given the preeminent role of perception in grounding basic beliefs (in keeping with the role that causal theory plays in Buddhist accounts of inferential reasoning), it would seem that Buddhism is actually friendly to naturalism, at least in a prescientific sense that reflects commitment to empiricism. But when it comes to its principles of reason (e.g., material implication, contraposition) it is not at all clear that they anticipate something like an analytical system of deductive principles and propositional laws. Certainly, in term of both structure and strategy, these principles codify specific rules of debate for the purpose of settling intramural disputes on a variety of issues (e.g., whether one can know the minds of others, whether sensations follow one another continuously). But they work primarily in support of Buddhism’s soteriological aims, not of some openended method of inquiry. Consider the pattern of argumentation in such canonical works as Points of Controversy (Kathāvatthu) in which the burden of proof switches from one party to the next, and neither offers any positive thesis. This sort of argumentum ad ignorantiam, which is presumptive rather than demonstrative, simply says “I am right because not proven wrong”. And it finds its most obvious illustration in Nāgārjuna’s radical thesis––as embodied by his doctrine of emptiness––that it is not just questions about causation, action, the self, time, arising, dissolution, etc., that lack definitive answers, but rather that reality itself in some sense is beyond the reach of

10

C. Coseru

thought. The radical thesis is that nothing can be revealed about how things are if how things are is to be understood in terms of ultimately real natures (Siderits 2015, 207). This is one issue that Siderits has thought hard about: how can we make sense of the soteriological significance of emptiness? That is, if Buddhism at its core is a soteriological project concerned with overcoming suffering, how can a doctrine that treats persons as mere useful fictions provide an effective basis for cultivating virtues such as generosity, compassion, selflessness, and ultimately serve as conduit for achieving enlightenment? Furthermore, just what it means to say (with the Mādhyamika) that we can only make sense of enlightenment if we affirm the emptiness of all dharmas (i.e., the constitutive elements of existence and/or experience)? In his postscript to “On the Soteriological Significance of Emptiness,” Siderits acknowledges the challenge his semantic interpretation faces: if philosophical rationality must be used to construct and defend a theoretical framework which affirms the truth of emptiness, then one must engage in the very sort of activity, namely, rational self-assertion, that hinders enlightenment. Those of us who go back with Siderits a long way recall that it is his stance on physicalism, and its implications for theories of personal identity, that has anchored much of his philosophical agenda (even as Siderits has continued to ponder just what it is that one can coherently say on behalf of the Mādhyamika). My memories of our early conversations in Calcutta in the winter of 1995 are quite sketchy, but I do remember us debating the issue of personal identity quite a bit. Derek Parfit’s Reasons and Persons (Parfit 1984)––with its robust defense of the idea that the unity of a person’s life could be accounted for in impersonal terms––had been in print for about a decade then and, like many other students of Buddhism in his generation, Siderits wondered whether the controversy its publication had stirred in Western philosophical circles was a good omen for the prospect of getting Buddhism a seat at the table. Since Parfit explicitly acknowledged that his view in some ways resembled what the Buddha said about selves, it seemed obvious that those laboring in the trenches to make Buddhist philosophy accessible to the uninitiated would suddenly find themselves thrust in the middle of that debate. But nearly another decade passed before Siderits brought to fruition his major work, Personal Identity and Buddhist Philosophy: Empty Persons, and made a compelling case for why it is that we should pay attention to the Indian debates if we want to advance the contemporary conversation on personal identity. By this time, and in response to developments in phenomenology and the mind sciences, the discussion had expanded beyond purely metaphysical considerations, to thinking, on the one hand, about the brain-based mechanisms that realize consciousness and cognition, and on the other, about how selves emerge out of the deep continuity between mind and life. In short, to many of those engaged in exploring the many facets of our embodied, social, and environmentally situated lives, the sort of substantive self-view targeted by Buddhists (and Humeans) seemed rather like a philosophical fiction. To take a cue from Metzinger’s Being No One (2003)––a book that landed on library shelves the same year as Siderits’ Empty Persons––nobody ever was or had the kind of full-blown immutable and enduring self that critics thought they were refuting. Certainly, a conception of the ‘soul’, ‘self’ (Gk. psuchê, Skt.

Introduction

11

ātman) or ‘ego’ has informed philosophical speculation on personal identity from ancient times to the present. But, if the Socratic and Upaniṣadic traditions of self-­ scrutiny are any indication, these terms serve largely as placeholders for the innermost principles of subjective individuation. Whether such principles should belong in our ultimate ontology depends largely on whether their function is ineliminable, and if that function is intelligibility it is hard to see how even an anti-essentialist metaphysics could rule it out.  In short, the no-self theorists might be targeting a straw man of their own making. What about the phenomenal or experiential self, that pre-reflective self-­ consciousness that is inescapably perspectival? Although not immune to reductionist and fictionalist interpretations, the dominant question in this case is not whether the experiential self is a process rather than a thing, perhaps something that emerges under certain conditions of psychological development. Rather, the question is which model of the self is best suited to capture the rich repertoire of embodied, psychological, social, and ecological aspects of our conscious lives. Being a self in this experiential sense, as phenomenologists would put it, “is an achievement rather than a given, and therefore also something that one can fail at” (Zahavi 2009, 552). That failure to achieve selfhood could be a problem is, at least in principle, incomprehensible to the Buddhist, for whom selfhood (and any conception thereof) is something to upend rather than achieve. But it shouldn’t be. For the Buddhist too has  a conception of consciousness as something that illuminates, discerns, and makes present. And while the Buddhists themselves debated whether taking conscious states to be ultimately impersonal means that one may fail to articulate their phenomenal and subjective character (or regard such articulation as a deceptive construct), they did not dispute that ‘conscious’ or ‘aware’ is something that one can fail to be. Undeterred, and largely in response to these new developments, Siderits has spent a large part of the last decade and a half spearheading several collaborative projects aimed at addressing the range of problems to do with the nature of self and consciousness by drawing on expertise from across the Buddhist, analytic, and phenomenological traditions. In collaboration with Evan Thompson and Dan Zahavi, he brought together a group of philosophers and Buddhist scholars to consider not merely the range of classical and contemporary arguments for and against the existence of a self, but also debates about its nature and possible identification with certain dimension of consciousness or self-consciousness (Siderits, Thompson, and Zahavi, 2011). More recently, he has championed efforts to understand whether the Buddhist doctrine of no-self could in effect be joined with a plausible theory of consciousness and cognition to address a range of perennial questions about personal identity (e.g., persistence, diachronic and synchronic unity), and the many possible connections these Buddhist debates have to recent controversies in the mind sciences. This is most evident in the recent collection of essays, Buddhist Philosophy of Consciousness: Tradition and Dialogue, co-edited with Ching Keng and John Spackman (Siderits, Keng, and Spackman 2021). Equally noteworthy, and reflective of Siderits’ related and life-long preoccupation with what the limits of language mean for philosophical practice, is his

12

C. Coseru

collaborative venture with Arindam Chakrabarti and Tom Tillemans in bringing together a leading group of scholars to examine whether the notorious apoha theory can explain our ability to use general terms without presupposing the existence of universals and other similar abstract entities (Apoha: Nominalism and Human Cognition, 2016). Similar concerns find articulation in another collaborative venture, this time with The Cowherds (2011), a group of eleven philosophers and scholars with a penchant for Madhyamaka attempting to make sense of the two-truths doctrine elaborated by philosophers in first-millennium India. Last, but not least, Siderits’ sustained commitment to challenging prevailing scepticism that Buddhism contains much by way of serious metaphysics and epistemology has resulted in two of the most lucid, accessible, and engaging volumelength introductions to Buddhist philosophy in the English language to date: the well-­received and widely cited Buddhism as Philosophy (2005, second edition 2021, abridged version 2022) and the just released introduction to Buddhist metaphysics, How Things Are (2022). The latter offers ample evidence and compelling reasons to believe that Buddhist thinkers not only engaged “in serious and sustained efforts to work out what most fundamentally exists”, but that they did so “out of concern to determine how things are anyway, independently of our interests and cognitive limitations” (Siderits 2022, 2). The idea is that unable to appeal to canonical texts an opponent would not regard as authoritative, philosophers with different doctrinal commitments eventually recognized the need for developing a common set of rules for debate. As a result, Buddhists and non-Buddhists alike began “to work out and defend answers to questions that the founders of their systems may not have addressed” (Siderits 2022, 4), In due course, these answers would map the conceptual terrain of a philosophical culture of similar ancestry and equal breadth and depth to those of China, Greece, and the Latin West.

3 A Brief Outline of the Central Questions This volume brings together twenty-two essays written by Siderits’ friends, colleagues, and students as a token of admiration and acknowledgment of his influential work. They cover the core areas of philosophy to which Siderits has made substantive contributions: metaphysics, philosophy of language, philosophy of mind, epistemology, and ethics. Most engage Siderits’ work directly, building on his pathbreaking ideas and interpretations. Many deal with issues that have become a common staple of debates about the scope of philosophical engagement with traditions outside the West. Some focus on key conceptual strategies and interpretive positions: Did Dignāga really change the course of Indian epistemology by introducing the idea of pervasion (vyāpti) of the reason (hetu) by the target property (sādhya-dharma) or was this idea already implicit in the early strata of Indian thought? (Katsura) Can Buddhist accounts of mental causation be understood in terms of necessary causal connections between consciousness moments or are they

Introduction

13

more akin to Humean accounts of causation as the constant conjunction of intrinsic natures? (Thakchoe) Others address broader issues that lie at the heart of the confluence approach, showing how the project of rational reconstruction of positions in one tradition in light of another can lead to new and valuable insights. Are persons mere conceptual fictions or are they in some, however imperfect, sense real? (Arnold, Coseru, Ganeri, Sharf) Does Buddhist reductionism about selves also eliminate the person? (Chadha & Nichols). If not, and if Buddhist Personalism is at least partly correct in claiming that persons, although neither identical not different from the aggregates, are nonetheless real, what are its implications for other issues in Buddhist philosophy of mind? (Coseru, Sharf, Strawson) What about the Madhyamaka critique of Buddhist reductionism: does that threaten not just the person but the entire framework of Abhidharma metaphysics? (Mackenzie, Arnold) Can we use the empirical findings and conceptual resources of cognitive science to unpack the Buddhist analysis of consciousness and cognition without distorting the latter or importing the prejudices and theoretical assumptions of the former? If so, what are some of the strategies and insights of this multidisciplinary research program? (Coseru,  Dreyfus, Sharf, Thompson) Might this research program, for instance, also help to make sense of the Buddhist conception of rebirth? Specifically, might the sort of arguments in defense of constructivist accounts of mental continuity such as we find, for instance, in Nelson Goodman (1978), bolster Dharmakīrti’s classical argument for mental continuity? (Westerhoff) A number of essays focus on a cluster of much debated and still unresolved interpretive problems: Does the Madhyamaka claim that no phenomena ultimately come into being as ordinarily conceived mean that ultimate truth/reality (satya) is ineffable or that there is no ultimate reality and a fortiori no ineffable ultimate reality? (Garfield & Priest). Is an interpretation of Nāgārjuna’s two truths doctrine as a form of dialetheism tenable in the Indian philosophical context or possible only in light of the many transformations that Buddhism underwent in China? (Tanaka) Can fictionalist formulations of Buddhist Reductionism explain how a conventional understanding of the world might be a fictional projection and yet still remain grounded in a more fundamental reality? (Guerrero) Should the svabhāvapresupposition––namely that there are entities that have svabhāva, an independent existence or intrinsic nature––be understood in pragmatic rather than semantic terms and, if so, what implications are there for Siderits’ anti-realist interpretation of the Madhyamaka two truths dialectic? (Spackman) More generally, what are some of the ways in which Abhidharma Reductionism and the Madhyamaka two truths dialectic can help us advance the realism/anti-realism debate? (Arnold, Chakrabarti) Still others ask whether a distinct moral philosophy can be discerned in the ethical teachings of Buddhism, and how best to articulate its principles. For instance, must moral thought be rooted in a conception of the inextricable relation between agency and moral responsibility, or does such a relation obscure the radical picture of the Buddhist ethical project? (Repetti) Is the path of moral and mental cultivation laid out by Śāntideva, for instance, best understood on a virtue ethical or

14

C. Coseru

consequentialist theoretical model? (Goodman) Do Buddhist ethical practices share a common ground with the self-cultivation exercises of the Confucians and, if so, do they underpin a robust moral philosophical program or just a basic account of the good life? (Gowans) Finally, some contributors put to the test received assumptions about whether there could be any common ground between Buddhism and traditions such as Platonism, Cartesian metaphysics, and even mathematical reasoning on such issues as conceptions of infinity. Carpenter asks whether in foregrounding the formation and re-formation of character around similarly ambitious ideals (e.g., the bodhisattva and the philosophos), Buddhism and Platonism might have more in common than meets the eye, their radically opposed metaphysical views notwithstanding. Likewise, Strawson wonders whether there may be less daylight between the Cartesian and Buddhist metaphysics of mind, given affinities between Descartes’s view of the mind as wholly constituted of consciousness and Buddhist accounts of the persistence of awareness captured by such notions as citta-santāna (‘consciousness-­stream’) or bhavaṅga-citta (‘life-continuum mind’)? Lastly, Read and Greiffenhagen ponder the extent to which Madhyamaka dialectics might help articulate the radical difference between the finite and the infinite by resisting the latter’s reification. The variety and breadth of these essays bear testimony to the legacy of Siderits’ impact in shaping the contemporary conversation in Buddhist philosophy and its reverberations in mainstream philosophy, giving readers a clear sense of the remarkable scope of his work. They are offered here in celebration and heartfelt appreciation of the immense debt we owe to his outstanding scholarship, philosophical acumen, and indelible friendship.

References Chakrabarti, A., and R. Weber. 2016. Introduction. In Comparative Philosophy Without Borders, edited by A. Chakrabarti and R. Weber, 1–34. London: Bloomsbury. Coseru, C. 2018. Interpretations or Interventions: Indian Philosophy in the Global Cosmopolis. In Routledge History of Indian Philosophy, edited by Purushottama Bilimoria with Amy Reyner, 3–14. London: Routledge. Goodman, N. 1978. Ways of Worldmaking. Indianapolis: Hackett. Harries, K. (1968). Wittgenstein and Heidegger: The relationship of the philosopher to language, The Journal of Value Inquiry 2 (4): 281–291. Inada, K. 1970. Nāgārjuna. A Translation of his Mūlamadhyamakakārikā with an Introductory Essay. Tokyo: Hokuseido Press. Metzinger, T. 2003. Being No One: The Self-Model Theory of Subjectivity. Cambridge, MA: MIT Press. Murti, T.R.V. 1955. The Central Philosophy of Buddhism. Routledge. Parfit, D. 1984. Reasons and Persons. Oxford: Clarendon Press. Siderits, M., E. Thompson, and D. Zahavi, eds. 2011. Self, No Self?: Perspectives From Analytical, Phenomenological, and Indian Traditions. Oxford University Press. Siderits, M. 2003. Personal Identity and Buddhist Philosophy: Empty Persons. 2nd ed. Ashgate. 2015.

Introduction

15

———. 2008. Paleo-compatibilism and Buddhist Reductionism. Sophia 47: 29–42. ———. 2016. Studies in Buddhist Philosophy. Edited with an Introduction by J. Westerhoff (and the assistance of Christopher V. Jones). Oxford University Press. ———. 2017. Comparison or Confluence Philosophy? In Oxford Handbook of Indian Philosophy, ed. J. Ganeri. Oxford University Press. Siderits, M., C.  Keng, and J.  Spackman, eds. 2021. Buddhist Philosophy of Consciousness: Tradition and Dialogue. Leiden: Brill. ———. 2022. How Things Are: An Introduction to Buddhist Metaphysics. Oxford University Press. Unger, P. 2014. Empty Ideas: A Critique of Analytic Philosophy. Oxford University Press. Westerhoff, J. 2016. Introduction. In Siderits, M., Studies in Buddhist Philosophy, edited with an Introduction by J.  Westerhoff (and the assistance of Christopher V.  Jones). Oxford University Press. Zahavi, D. 2009. Is the Self a Social Construct? Inquiry 52 (6): 551–573.

Part I

Mind

But Aren’t We Conscious? A Buddhist Reflection on the Hard Problem Georges Dreyfus

One of the persistent debates that has animated thinkers in ancient India and in the modern West concerns the nature of consciousness. Is it merely an epiphenomenon that can be reduced to matter, as materialists argue, or does it belong to a different ontological domain, as anti-materialists and idealists hold? In India, the Carvakas held the materialist view, arguing that the mental is merely a manifestation of the body’s capacities, much like inebriating power is a property of fermented grains, and hence has no ontological autonomy. Most Indian thinkers, however, rejected this view and argued that consciousness is a sui generis phenomenon that cannot be reduced to the material domain. In the modern West, the same debate has been front and center since the dawn of the scientific revolution when Descartes sharply distinguished the mental, res cogitans, from the domain of material extended entities. This question is often described as the hard problem of how material brain states can give rise to the mental states that we seem to experience? This is different from the “easy problem,” the much more tractable question of how to explain various mental functions such as the ability to integrate information, categorize objects and focus attention (Chalmers 1996). The main traditional Buddhist take on this topic is best described as an event dualism according to which there are fundamentally two types of phenomena in the universe, material and mental. Although this position does not need to amount to a classical mind-body dualism, it is often understood to entail such a position. Dharmakīrti’s vigorous rejection of materialism is a clear example of such a mind-­ body dualism, which is understood to support the traditional cosmology of rebirth and karma. But this kind of dualism, which corresponded to the zeitgeist of his time, is more difficult to assert in an era when the terms through which the question is understood are connected to scientific inquiry. The modern consideration of G. Dreyfus (*) Williams College, Department of Religion, Williamstown, MA, USA e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_2

19

20

G. Dreyfus

consciousness is no longer purely philosophical but is seen as based on the findings of the natural sciences. Hence, materialism, which was in India a minority position difficult to entertain, has become the dominant view borrowing its authority from the limited but very real successes of the new mind sciences. This predominance of materialism cannot but concern the Buddhists who wish to be relevant to the modern context. What kind of position should they be taking on this question? Should they go along with the dominant materialism and possibly jeopardize some of the central tenets of their tradition, or should they stick to the views expressed by classical Buddhist thinkers and risk irrelevance? One thinker who has addressed this debate within modern Buddhist philosophy is Mark Siderits (Siderits 2011), who has offered an account of consciousness inspired by contemporary cognitive scientific accounts, particularly that of Daniel Dennett’s deflationism (more on this shortly). For Siderits, we should extend Abhidharma reductionism, which argues that reality can be reduced to its ultimate components (partless particles and moments of awareness), to the mental itself and understand consciousness not as some kind of inner state or process but merely as a convenient way to account for the global availability of deliveries of the cognitive systems. Consciousness is merely what needs to be posited to account for how we conceptualize our cognitive processes so as to bring about appropriately coordinated behavior and provide the possibility for corrections. In this essay,1 I first critique Siderits’ deflationism from a modern Buddhist perspective, arguing that his account is both philosophically problematic and antithetical to the Buddhist project as it fails to respect the phenomenological integrity of many of the meditative practices central to the tradition. Hence, it has to be considered as a false friend of modern Buddhists interested in developing accounts compatible with modern scientific views. In the second part, I suggest to modern Buddhists that they may want to avoid the dualism that undergirds most traditional Buddhist accounts and instead adopt a phenomenological understanding of the Buddhist views about the mind. I further argue that this phenomenological approach, which takes seriously how things appear to us while refusing to ontologize these experiences, is in agreement with the Madhyamaka Buddhist philosophy. I further argue that this Madhyamaka refusal to ontologize the mind frees modern Buddhists from having to adopt dualism, which is deeply problematic within a scientific context, while allowing them to introduce into the modern conversation valuable insights found in traditional texts such as the view of consciousness as reflexive. I conclude by showing how this view of consciousness is germane to Buddhist practice.

 I would like to thank all the people who have helped me in sorting these difficult ideas. I cannot mention all of them, but particular thanks are due to Joseph Cruz, Evan Thompson, Mark Siderits, Tom Tillemans, Natasha Judson and many others for their useful comments and feedback. 1

But Aren’t We Conscious? A Buddhist Reflection on the Hard Problem

21

1 What is Deflationism? To start with, let us define the nature of the deflationist project, which argues that the conundrum of consciousness is not a deep question but simply the result of philosophical confusion. This deflationist project is a close relative to another popular philosophical position, eliminativism. There are, however, differences in that the latter (eliminativism) argues for the elimination of intentional terms (terms referring to the mental such as “consciousness,” “desire,” etc.) and their replacement by a scientifically more accurate vocabulary (Churchland and Sejnowski 1992), whereas the former (deflationism) does not share this reformist agenda, arguing that there are good reasons (biological, cultural or even practical) for people to adopt an intentional stance toward the behavior of humans and other animals. Hence, there is no reason to eliminate intentional terms, though they do not reflect accurately what is taking place in reality. In this essay, I will argue against deflationism, which I take to be a more credible view than eliminationism within the context of modern Buddhist thought. In discussing the former I will also, however, address some of the basic intuitions that underlie the latter, since, as Dennett himself recognizes (1988, 44) the two positions share a great deal. The most prominent exponent of deflationism may be Daniel Dennett, who has argued forcefully that there is no hard problem (Dennett 1988, 1991). For him, there are a variety of cognitive processes such as vision that provide us with information about the world and our place therein, but there is no central place, no Cartesian theater in which the results of these processes are displayed. Each cognitive process runs its course and delivers its results through multiple interactions with other processes leading to particular behavioral results. Hence, seeing, for example, is not some kind of interior experience but the result of a series of information processing devices leading to actions in relation to the world of objects delivered to our sense of vision. The sense that there is some kind of witness for whom this cognitive event is happening as well as the sense of unification that we feel as subjects for whom these experiences are taking place are not reflections of what is happening at the level of experience but simply the results of the judgments we pass on the multiple drafts that circulate in our cognitive system. Hence, for Dennett, there is nobody home and no way to make sense of what it is for a subject to experience things over and above how information is cognitively treated. Consciousness is not some mysterious domain yet to be explored, but merely an illusion, the effect of the revisionist unification operated by our conceptual system. The reality is that we are robots, or, rather beings made by robots, lots of them, which have the self-inflating tendency to think that they are more than robots. It this tendency that leads Chalmers (1996) and others to posit a hard problem but this is just a philosophical mistake. In reality we all are zombies who simply fail to recognize our true zombie nature. Within contemporary Buddhist philosophy, this deflationary perspective has been advocated by Mark Siderits, who has argued that we should extend the Abhidharma critique of the self (the idea that the self is not real but merely a useful but mistaken way to conceptualize the body-mind complex) to the mental. For him,

22

G. Dreyfus

consciousness is not some kind of inner state or process but merely a conventional way to account for the global availability of cognitive processes.2 Tackling the classical philosophical (both Indian and Western) question of how we know we have experiences, Siderits recommends that we take seriously the Bhaṭṭa Mīmāṃsā answer: we know that we experience by inferring that we know from the cognizedness of the object that we know. Hence, there is no need to posit a reflexive cognition, which would violate the anti-reflexivity principle (the idea that no agent can act on itself), nor is there the need for a higher order cognition that would somehow render conscious the otherwise unconscious cognitive processes. All that is needed for realizing that we know that we experience is an inference that the object has been cognized. The main difficulty that this Mīmāṃsā position faces is that it is not obvious what cognizedness is over and above the features of the object itself. Siderits suggests that we should just bite the eliminativist bullet and do away with intentional descriptions of the situation, realizing that this cognizedness is nothing over and above the set of information about the object made available to the various cognitive systems by our physical senses. For example, we know that we see a pot when we infer the cognizedness of the pot, which is just the fact that the pot has been made available by our sense of vision for appropriate action. Hence, consciousness is merely a convenient way to deal with the complexity of our cognitive systems, the need to pass judgments on the various cognitive processes that are taking place at any given moment so as to bring about appropriately coordinated behavior and provide the possibility for corrections. Humans are indeed robots or zombies, but they imagine themselves to be real substantial and conscious selves, and this is the source of their problems. According to Siderits’ view, the goal of Buddhist practice is to bring us in touch with our true zombie nature so that we come to realize that consciousness is just a conventional way to sort out and unify the deliveries of our cognitive systems, not a real entity or a process that warrants further exploration over those devoted to the explanation of neurophysiological systems. Hence, accomplished practitioners are best described as Robo-Buddhas, who are awakened to the reality that they are just like us, that is, merely robots who act on automatic pilot. Their superiority is that they have realized their Robo-Buddha nature and developed a bias toward compassion so that their actions are appropriately selflessly robotic, whereas we remain mired in our inflated self-centered delusion.

2 The Problems of Deflationism To those unfamiliar with analytical philosophy of mind, deflationism is bound to seem a strange position, a sophistic trick that removes the hard problem (how material brain states can give rise to mental states) rather than solves it. Nevertheless, we

 For a study of Abhidharma reductionism, see Siderits (2003).

2

But Aren’t We Conscious? A Buddhist Reflection on the Hard Problem

23

may not want to dismiss it too quickly, for this position is not without its virtues despite its counter-intuitive character. It provides a clear (too clear?) and seemingly decisive answer to the hard problem, an answer rooted in the mainstream of the cognitive sciences with seemingly strong naturalistic credentials. Here there is no need to wonder about how consciousness reveals objects over and above the ways in which cognitive systems process information. Since cognitive science has good accounts of information processing systems, we can see how deflationism can suggest that we are well on the way to solving the hard problem. This position could also be especially attractive to Buddhists, since it offers a trenchant critique of the reification of subjectivity, a critique that is not unlike the Buddhist own deconstruction of ordinary self-centered subjectivity. Hence, the modern Buddhist has some ground for following Siderits in adopting this view. My argument here is that despite its virtues deflationism does not offer a productive stance for modern Buddhists, who should see it more as a false friend than as an ally in the struggle against the reification of the self. I rest my case on two considerations, arguing that deflationism is not satisfactory as a theoretical account of cognition, before showing that Buddhists have particular reasons as Buddhists to be deeply suspicious of this account. First the philosophical case, which argues that deflationism provides a poor account of what it means to perceive an object and hence cannot provide a satisfactory account of cognition. Let’s take the example of vision. The deflationist argues that seeing does not involve any kind of phenomenological content and is merely a reflection of how we think about the non-conscious deliveries of our perceptual apparatus. An interesting case is that of the pathological condition of blind-sight in which stroke victims claim to be completely blind but are able to navigate adequately coarse obstacles. The deflationist points to this pathology as showing that there can be information processing without any subjective sense of there being an experience for somebody. The conclusion is that seeing does not require subjective experience and hence that there is no need to posit it in normal cases. My take is quite different. This pathology suggests, on the contrary, the opposite view, namely, that there is phenomenological content in normal cases, content that is lacking in blind sight. Is this not where the difference lies between normal cases and blind sight? For the unfortunate blind sighted, there is information processing but no conscious experience, whereas in the case of normal vision there is a subjective sense that vision is taking place together with non-conscious information processing. Otherwise, we would have to conclude that the difference between the pathological and the normal cases is merely one of conceptual judgment. The deeper problem with the deflationist account is its overly dichotomous picture of the mind, in which all cognitive operations are either unconscious information processes or conceptualizations. There is nothing in between, no experience of vision that mediates between the two realms. Hence, there is no possibility of distinguishing perception from judgment, experiencing from believing that one experiences, being in pain and judging that one is in pain. All there is to experience is the judgments that we make on otherwise unconscious cognitive processes. It is not just that it might be difficult to find sharp boundaries between qualitative aspects of our

24

G. Dreyfus

experience and the propositional content of our judgments, a point I would readily agree with, but that the very distinction between qualitative aspects of experience and judgments can never be made sense of. All there is are mistaken judgments about having experiences and mechanical cognitive processes. This is not to deny that there are interactions at multiple levels between experiences and judgments and that these interactions profoundly affect our psyche. Assessment of experience is relevant to the experiences we undergo and can influence and modify them to an extent of which we are not often conscious. Our feeling of pain, for example, will certainly be influenced by the attitude we have toward this pain. Hence, hypnosis or meditation may be very helpful in transforming painful sensations. Similarly, priming has profound influence on the ways in which we understand our experiences and shows the degree to which we can be influenced unbeknown to ourselves. Hence, Dennett is correct in claiming that it is not possible to posit the existence of phenomenal qualities intrinsic to our mental states. We do not experience objects in piecemeal fashion but rather as a function of the whole of our psyche. Hence, perception cannot be isolated from assumptions, presuppositions and expectations. But it is an entirely different thing to deduce from this more modest point the much more extreme claim that our perceptions are wholly constituted by our attitudes, that there is nothing to our experiences over and above how we conceptualize them. In fact, even this formulation is not acceptable for the deflationist, who argues that it is not the case that we conceptualize experiences in mistaken ways, but, rather that we think that we have experiences whereas in reality this is merely the product of conceptual confabulation, the just-so stories that we tell ourselves to make sense of our actions and reactions. Hence, in this overly intellectualistic picture of the mind, the entire cognitive work is done by the combination of unconscious cognitive processes and conceptual thinking. I also believe that Buddhists have particular reasons as Buddhists to be concerned about this account. This is so not because this view is reductionist, but because it does not provide sufficient resources for making sense of some of the central practices of the tradition. The deflationist account is materialist, reducing the mental to behavioral dispositions and bodily affects, but this is not its main problem. For if modern Buddhists want to engage the mind sciences, they must be ready to run the risk that reductionism may be correct, that science will show conclusively that although the mind appears to be immaterial, in reality it is not. This is not to say that present day science is close to this point, but I would argue that it is difficult to exclude such an outcome in principle. The problem with deflationism is not that it reduces the mental to the physical but that it eliminates the possibility for making sense of how people engage in certain types of Buddhist practice, which seem predicated on the possibility of making a distinction between experiences and the ways they are conceptualized. Let’s take as an example one of the most basic Buddhist practices, the mindfulness of breathing (anapanasati). In this exercise, the practitioner is instructed to focus on her breathing and notice its characteristics: is the breath short or long, shallow or deep, does it change or remain constant, etc. In doing so, the practitioner is also instructed not to conceptualize but merely to observe and develop a nonjudgmental awareness

But Aren’t We Conscious? A Buddhist Reflection on the Hard Problem

25

of the breath. In this way, the practitioner is said to experience the breath directly and to be able to realize that the breath and the whole mind-body complex do not exist as she understood them, that is, as a permanent self-entity, but, rather, as a continuum of ever changing experiences. I believe that accounting for this type of practice becomes highly problematic if one were to accept the deflationist account sketched above. The reason is that a central practice such as the mindfulness of breathing is predicated on the premise that it makes sense, at least at the phenomenological level, to distinguish judgments about the breath from nonjudgmental awareness of the breath. Now it could turn out that this distinction is problematic at a deeper level analysis, and I believe that the modern Buddhist should be ready to entertain this perspective. But even if this were true, this would not entail that there is no way to make sense at the phenomenological level of the distinction between how the breath appears to the person and how she conceptualizes her experience. This seeming distinction may be just that, a seeming that cannot be justified in reality, but there must be at least a seeming for the practice to make sense. It is the very possibility of such a seeming difference between experience and conceptualization that the deflationist cannot accept, for he has to argue not only that this seeming is not justified epistemologically but that it is not justified even at the phenomenological level. The deflationist cannot claim that it just appears to the practitioner that she has experiences, which are in fact nothing but conceptualizations, because he has to deny that the idea of something appearing makes sense over and above the fact that the person (mistakenly) thinks that something appears. It is not the case that there seems to be a difference between observing and judging the breath over and above the ways in which the practitioner conceptualizes her experiences. For him, the practitioner is merely making up the conceptual distinctions that she creates, distinctions that have no basis in reality since there is no experience over and above the judgments that the subject makes, just unconscious cognitive processes being arranged and rearranged conceptually. Hence, the practice as conceived by the tradition is incoherent, being based on the deluded perspective that it makes sense to try to observe what one experiences. This is not to say that the deflationist will be unable to provide some account of the effectiveness of these practices. He may argue that such a practice has beneficial results such as calming the mind and is effective in leading practitioners toward the total mindlessness that he thinks is the goal of Buddhist practice. For, the only practice that can be consistently vindicated by the deflationist is that of total mindlessness, the practice of a fully realized Robo-Buddha abiding in complete cessation and acting spontaneously in the world like clouds producing rain. This is also not to say that there are no Buddhist traditions that offer such a depiction of enlightenment and recommend a practice of complete extinction of the mental. But anybody familiar with Buddhist traditions as they have existed in the last centuries will have to recognize that such a practice does not correspond to the main practices recommended by the tradition and that there are many other practices that seem to presuppose observation of the body-mind complex rather than complete mindlessness. Moreover, such practices often do involve a distinction between experience and conceptualization. The problem with the deflationist account is that it is unable to

26

G. Dreyfus

preserve the phenomenological integrity of this distinction. Although the deflationist may be able to assign some pragmatic value to such a practice, he is unable to make sense of the descriptions of the practice. For him, this description is just confused, being based on the failure to understand that the only valid goal of meditative practice is mindlessness, the realization of one’s true zombie nature. Thus, the deflationist account is deeply problematic for the modern Buddhist not because it is materialist, but because it fails to preserve the phenomenological integrity of the tradition’s description of most meditative practices. The problem with the deflationist account is that it is revisionist, entailing a normative and a priori view about which descriptions can be taken seriously and which ones need to be rejected as being merely born out of confusion. This is why this account must be treated with suspicion by modern Buddhists. They should be open to the possibility that the analyses of meditative practices might turn out to provide a picture of the mind that differs from tradition accounts but cannot and should not accept an analysis that jeopardizes the phenomenological integrity of some of the central meditative practices of their tradition. Buddhists should be committed to save the phenomenon, the fact that there are definite ways in which practices can be described from a first person perspective, ways that are based on how our experience. Hence, modern Buddhists should resist the deflationist attempt to eliminate subjectivity, which is necessary to preserve the phenomenological integrity of many practices central to their tradition.

3 Madhyamaka and the Hard Problem If the deflationist account is unsatisfactory and if the tradition does not provide the resources to solve the hard problem, we may wonder which stance is open to perplexed modern Buddhists? To answer this question, modern Buddhists may want to start from a position of epistemic modesty, realizing that it is not the case that their tradition has all the answers. In the words of Evan Thompson (2020), they should avoid “Buddhist exceptionalism,” the belief that Buddhists have practices (meditations) that provide an insight that make them inherently superior to other traditions. My purpose in this essay is to find a viable stance that shuns this position because I believe that it is important for Buddhists to remain open to the ideas that come from the cognitive sciences, which provide new resources for understanding the human mind, a project that has traditionally been of great interest to Buddhists. This openness to scientific insights should be tempered by the skeptical awareness of the propensity of humans to claim to know more than they actually do. There is ample ground to be skeptical about many of the more inflated claims made by philosophers and scientists for a premature solution to the hard problem. Hence, Buddhists do not need to rush to accept the views put forth by many scientists and philosophers yet should take them seriously. They should also offer their own views based on the resources found in traditional accounts, particularly those of the

But Aren’t We Conscious? A Buddhist Reflection on the Hard Problem

27

Abhidharma, views that do not provide a final answer but are certainly good enough to make contributions to the modern conversation about consciousness. The traditional view of the mind found in the Abhidharma is based on the twin ideas of non-substantiality and dependent origination. In this view, the phenomena given in experiences are not unitary and stable substances but complex and ephemeral formations of events arising within complex causal nexuses. This is true not just of the person, who is not a substantial self but a construct that comes to be only in dependence on complex configurations of multiple mental and material events (the five aggregates of early Buddhism), but also of mental processes, which unfold not through the control of a cognitive executive but on the basis of the interaction of competing factors whose strength varies according to circumstances. There is no enduring experiencing subject, inner controller or homunculus that manages or observes the cognitive process. Rather, the moments of awareness exercise all the functions necessary to cognition and constitute it. Each moment of awareness comes to be in dependence on various conditions (preceding moments of awareness, object, sensory basis, etc). Having arisen, it performs its function and dissolves, giving rise to the next moment of awareness, thus forming a stream of consciousness or continuum (santāna, rgyud) not unlike James’ stream of thought or Husserl’s mental flux. Hence, for the Abhidharma, the mind is not an entity or a substance. It is not a thing, a kind of mechanism that produces thoughts, memories or perceptions, but a constantly changing process, a continuum of arising and ceasing complex and multi-layered mental events, which are, at least indirectly, phenomenologically available. This Abhidharma analysis of the mind is usually thought to support the kind of mind-body dualism central to traditional Buddhist cosmology. This is particularly true of later traditions, which inspired by Dharmakīrti seem to have moved to a position close to a substance dualism. I would argue that such a view is highly problematic within the context of modern scientific accounts. For if mental states exist as real non-material entities, how can they relate to the brain states that they seem to be tightly connected to? This is the fundamental challenge that dualism finds it extremely difficult to meet and this is why I believe that modern Buddhists should avoid this position. To do so, they need to understand that the central concern of the Abhidharma analysis of the mental should be taken to be phenomenological rather than metaphysical, providing descriptions and analyses of the complexity of the components of mental processes as they appear to us, not an analysis of the ontological basis of mental processes, a basis which is not readily available to the kind of analysis central to its project. Hence, I believe that there is no problem for modern Buddhists to take Abhidharma descriptions as offering first and foremost phenomenological analyses that can be used to engage in a dialogue with other accounts of the mind. To put it otherwise, Buddhists should realize that the Abhidharma is best taken as dealing with subjectivity rather than with consciousness in itself. One may then wonder: what about consciousness in itself? This question lies behind many searches for solutions to the hard problem, the attempt to find what consciousness is in reality over and above the superficial appearances that we

28

G. Dreyfus

encounter. This is where Madhyamaka Buddhist philosophy and its refusal to ontologize any phenomenon may be helpful. Madhyamaka can be understood as a cure for our temptation to reify reality and to seek ultimate explanations for what there is. Instead it re-directs us toward taking seriously appearances, the ways things are given transactionally in our practices. Regarding the mind, Madhyamaka suggests that it is not the case that subjectivity is merely a superficial appearance of a deeper more real set of material entities. Rather it is the case that both brain states (observable from a third person perspective) and subjectivity (given from a first person perspective) are partial aspects of what we conventionally call the mind. Brain states and subjectivity are transactional realities that cannot be eliminated but that should not be reified either. Hence, we should resist the temptation to reduce the mental to the material in a way that makes it a mere epiphenomenon. To paraphrase Wittgenstein (1953), subjectivity is not a real something but it is not a nothing either as it is given to us in a plurality of partial and irreducible perspectives. When we think about the ways in which consciousness is given, we can distinguish three perspectives. There is the first-person perspective through which consciousness is given to the subject. In this perspective, objects are given from a certain vantage through a certain affective tone, and consciousness is this subjective perspective, the how it is for the person to undergo that experience. For example, I see my car in front of the house. The car appears to me from a certain angle. I see its color and shape and this brings in me a certain mood. I am elated to have such a beautiful car or saddened to be stuck with such an old clunker. But I am also aware at least non-thematically that I am having this experience. This is the first-person perspective, the subjective point of view. Another irreducible perspective is the third person perspective, that of the scientist who is able to objectify my experience and treat it at least provisionally as an object that can be delimited and studied objectively. In this third person perspective, my reports of seeing the car are correlated with a certain number of measurements that can be determined by any observer with the right equipment. This is the objective point of view. Finally, there is the inter-­ personal perspective, which we can leave aside for the time being, the second person perspective, the view of the person who looks at me while I am looking at my car and understands through empathy the various mental states that I am going through. Many discussions about consciousness seem to want to go beyond these three perspectives to find what consciousness is in itself. This is particularly true of most versions of physicalism which are keen to reduce consciousness to some kind of neural mechanism available to the third person person perspective and to declare the subjective perspective as being superficial if not superfluous. It is precisely this search for what consciousness really is that I would argue should be questioned from a Madhyamaka perspective. There is no such thing as consciousness in itself over and above the various perspectives given to us in our practices. None of these various aspects are more real than the others. Each of them has a limited transactional reality, and to understand them is what it means to understand consciousness.

But Aren’t We Conscious? A Buddhist Reflection on the Hard Problem

29

In a way, the Madhyamaka view of consciousness suggested here is not unlike what has been described as the double aspect view of consciousness, a view often attributed to Spinoza. According to this view, consciousness is given as having two irreducible aspects, the mind as an extended aspect and the mind as subjective aspect. As far as I understand this view, both aspects are understood as belonging to a single real substance not directly available to us. Mādhyamikas should agree that the objective and subjective aspects of the mind are irreducible but they should reject the idea of a unique hidden real substance. For Madhayamaka, consciousness is not hidden, but is, rather, in plain view. This discussion of a modern Madhyamaka view of the mind directly applies to our discussion of the hard problem, which is often understood in causal terms: how to understand how neuronal networks produce or constitute consciousness qua subjective experience. The solution is then taken to entail an explanation of the causal relation between neurophysiology and particular experiences, and this is at times taken to entail the theoretical possibility of reproducing such experiences under artificial conditions (multiple realizability hypothesis). I think that this way of thinking about the hard problem is what Madhyamaka should be skeptical of, for it presupposes that consciousness can be understood as a causally produced real something rather than as a complex process given in a plurality of irreducible perspectives. The search for a material explanation of consciousness (particularly in its subjective dimension) assumes that it can be captured through a causal scheme in which neuronal networks act as causes of mental states. This stance reflects a more general realist metaphysical perspective according to which causality has a central role in explaining how things are in reality rather than merely consisting of conventional and hence limited ways to make sense of reality. This realist understanding of causality when applied to consciousness creates unsolvable quandaries such as the question of the causal role of consciousness. Is consciousness causally inert and if so, how do we explain that at least some of our actions seem to follow from our intentions? And if not, how can a causally active consciousness (particularly in its subjective dimension) be interacting with material phenomena? From the Madhyamaka perspective advocated here, all these quandaries stem from the twin false presuppositions that consciousness in its subjective aspect needs to be explained in scientifically explainable causal terms and that causal relations exist intrinsically over and above the various contexts and perspectives in which they are given. We then cannot but face the hard problem. But if we give up this view of the metaphysical status of causality as argued by Madhyamaka, we may be able to bypass these quandaries and find a way to think of consciousness as being merely correlated to neuronal states rather than produced by them. In this perspective, asking how my subjective intention gives rise to the firing of my muscles appears to be simply the result of a conceptual confusion. It presupposes that we can talk about causal relations concerning the mind in abstraction from the perspectives through which mental states are given to us. It is perfectly feasible to talk about how my intention gives rise to my action understood from the first person perspective. It

30

G. Dreyfus

is also unproblematic to assume that such intention is most likely correlated to particular neuronal states that are given from a third person perspective and that such neuronal states do act as causes of the firing of my muscles. What cannot be asked is how my subjective intention (first person perspective) causes the firing of my muscles (third person perspective) for this assumes that causal relations can be established across the board in abstraction from the perspectives in which mental states are given. To put it otherwise, causal relations do not exist intrinsically but are conceptual constructs that we use to make sense of our experiences. Hence, it is nonsensical to try to use the causal concepts that belong to the third person perspective to explain the first person perspective or build special bridge laws between the two.3 When we give up this metaphysical stance we realize that the hard problem arises from a conceptual confusion about the role and nature of causality in explaining mental states. We also realize that what needs to be understood is not how neuronal states produce mental states but how such neuronal states can be correlated to mental states. Here again, we should remain humble, not making any assumption about the degree of tightness of such correlations. It is most likely that such correlations are extremely complex, that there is no one to one correlation between mental and neuronal states, but this should not discourage us to engage in the difficult work of finding such correlations. In short, what I am suggesting is that we should adopt a position of prudent openness to the wonderful discoveries of modern brain science while remaining aware of its limitations and possible entanglements with the bad kind of metaphysics that is responsible for the hard problem.

4 Theory and Practice I also believe that this Madhyamaka stance allows modern Buddhists to entertain a view of consciousness that harmonizes with Buddhist practice and provides a fruitful way to engage with the mind sciences. This stance is based on the irreducibility of the three perspectives sketched above. It is phenomenological rather than metaphysical, making no claims about what consciousness is in reality and focusing instead on appearances. It is in this perspective that the ideas found in the Abhidharma and other traditional texts are also best understood, that is, as phenomenological resources that contribute to our understanding of consciousness and support Buddhist practice. This is particularly true of the Yogācāra tradition, which offers some important insights such as Dignāga’s and Dharmakīrti’s view of consciousness as reflexive. According to this view, which provides a basis of many important practices, consciousness is not just aware thematically of its object, it is also aware of itself, albeit in a non-thematic way. This self-awareness is neither introspective nor

 Special thanks to Evan Thompson for helping me to formulate this point.

3

But Aren’t We Conscious? A Buddhist Reflection on the Hard Problem

31

reflective, for it does not take inner mental states as its objects. Rather, it is the ability for the mind to keep track of the prominent mental episodes so that the person automatically becomes aware of how her experiences affect her. This self-awareness is particularly relevant to the domain of sentience, the domain of how we feel about what we experience. It allows organisms to be aware of the feeling tones that accompany the experiences they undergo and hence to react adequately to their environment. It is not a separate cognition but simply the registering of conscious episodes. In this way, it provides the basis for introspection, the paying attention to some of the content of mental states.4 To put it otherwise, consciousness is not just concerned with the mapping of the external world, it is also involved in the mapping of how the cognition of the external world affects the organism engaged in such a cognitive process. Hence, consciousness is reflexive, but this reflexivity is not conceptual. Rather it takes place at the most basic level of our interaction with the environment, the level in which bodily awareness registers the ways in which we are affected by our processing of the objects we encounter in the world. In our interactions with our environment, we do not just become aware of objects in abstraction from how these objects affect our body, as if we were disincarnated computers processing information, but, rather we evaluate how the flow of information that we process affects us. Consciousness is not just the apprehension of an object but also involves the keeping track of how cognition of the object affects us. This way of conceptualizing consciousness as having these two aspects has some relevance to the question that is often raised in scientifically informed discussions about the function of consciousness. Why are we conscious rather than just merely processing information like a computer does? The answer becomes clear when we understand consciousness as doubled sided. It is then easy to speculate that having a mind would provide an evolutionary advantage to an organism, an ability to monitor how its interactions with the environment affect it. Thus, consciousness is not just a way for the organism to adapt its behaviors to its environment but it is also a way for the organism to keep track of and to be able to evaluate how its interactions with the environment affect it. The view of consciousness as pre-reflectively self-aware also allows us to gain a better perspective on Dennett’s critique of the idea of consciousness as a Cartesian theater in which perceptions and ideas are displayed for the subject. Dennett (1988) is quite right to critique this idea, which presupposes that consciousness apprehends clearly perceptions and ideas as internal objects, the infamous “qualia” which would then be well-delineated objects existing within the field of consciousness. Positing qualia as internal objects is problematic. This is so not because the qualitative aspects of consciousness don’t exist but because they are simply not objects. To understand them, we need to move away from the idea that consciousness can be understood as the apprehension of the right kind of objects. Dennett (and Siderits)  For a discussion of the idea of consciousness as having a double aspect, see Dreyfus (1997), chaps. 19 & 25. For obvious parallels with the phenomenological view consciousness as pre-­ reflective self-awareness see Zahavi (2005). 4

32

G. Dreyfus

is right to critique of the reification of experience but goes too far in eliminating the possibility of considering the phenomenal qualities of awareness. The idea of a pre-reflective self-awareness is a way of highlighting the qualitative aspect of consciousness without having to explain consciousness as the apprehension of an object by an internal sense. In this view, consciousness does not consist of entities transparently available to the subject through introspection, but of a monitoring of perceptions, sensations and ideas. Most of our information about our mind does not come from introspection, which is retrospective, but from the pre-­ reflective awareness that we have of our mental states, awareness that often delivers impressions rather than clear pictures (Schwitzgebel 2010). My claim is that this phenomenological view of consciousness is in congruence with the Madhyamaka view of the mind as empty. This idea aims at de-reifying the mind, not at negating its transactional reality. In particular, it is not meant to eliminate the subjective aspect of the mind, the ways in which the world discloses itself to us. This appearance should not, however, be treated as a thing but merely as an appearance, that is as a phenomenon empty of substantial reality. These two views that the mind is empty and aware of the appearances form the basis of some of the later Buddhist traditions practices such as Mahāmudrā, which focus on the nature of consciousness as being empty and clear (having appearances). The seventeenth century Tibetan master Karma Chagme (1998) explains: When it [awareness] stares at itself, with this observation there is a vividness in which nothing is seen. This awareness is direct, naked, vivid, unestablished, empty, limpid luminosity, unique, non-dual clarity and emptiness. It is not permanent, but unestablished. It is not nihilistic but radiantly vivid. It is not one, but manifoldly aware and clear. It is not manifold but indivisibly of one taste. It is none other than this very self-awareness.

These meditative instructions offer the glimpse of an experience of awareness as being clear and without content while being self-aware. In this state, awareness experiences itself without the distortions imposed by the dualistic structure through which we usually deal with reality. Consciousness appears as empty since it is non-­ substantial but it is not nothing either since it is luminous, having a basic self-­ presencing or reflexivity that is irreducible to the usual subject-object structures of our ordinary conception of experience. This view of consciousness as being empty and knowing brings together phenomenology and Madhyamaka. As such it is relevant to Buddhist practice and to the contemporary discussions concerning the nature of consciousness. Buddhist traditions offer rich resources for understanding the mind, not just through the meditative practices that scientists can take as the object of their studies, but also through the descriptions and analyses of the mind suggested by the tradition. But what Buddhist doctrines cannot claim is to provide the ultimate authority. They can provide phenomenological descriptions of consciousness that can orient research and constrain the models used by cognitive scientists. They can also provide helpful methods of self-transformation. But it should remain clear that claiming that Buddhism offers important insight about the mind is quite different from the more inflated claims that Buddhism “had figured it out all along” or that we need a new contemplative science to realize the insights of discovered by the Buddha 2500 years ago.

But Aren’t We Conscious? A Buddhist Reflection on the Hard Problem

33

The contribution of Buddhism is not, however, limited to providing valuable phenomenological insights into consciousness. By highlighting the limits of our knowledge in general and critiquing some of the metaphysical presuppositions of the discussion about consciousness, Buddhist philosophy through its Madhyamaka interpretation orients our inquiry in a more fruitful direction. Rather than attempting to find per impossible how consciousness qua subjectivity is produced by neuronal states, it orients us to the more realistic project of neuro-phenomenology, combining the first- and third-person perspectives in order to gain a better understanding of the different aspects of the mind. The Madhyamaka philosophy of consciousness sketched here cannot pre-judge where this project will take us. What it can do, however, is to suggest that rather than looking for a solution to the hard problem, we should pursue the scientific enterprise of finding the complex ways in which mental and neuronal states are correlated. In this way, a Madhyamaka philosophy of the mind can provide a solid philosophical foundation for such scientific enterprise as well as provide a consistent philosophy of the mind that can support Buddhist practice and enter in a fruitful dialogue with the modern sciences of the mind. Hopefully, what I have provided here a step toward achieving this goal.

References Chalmers, D. 1996. The Conscious Mind. Oxford: Oxford University Press. ———. 2002. Philosophy of Mind: Classical and Contemporary Readings. Oxford: Oxford University Press. Churchland, P., and T. Sejnowski. 1992. The Computanional Brain. Cambridge: MIT. Dennet, D. 1988. Quining Qualia. In Consciousness in Contemporary Science, ed. Marceel and Bisiach, 42–77. Oxford: Clarendon. ———. 1991. Consciousness Explained. Boston: Little, Brown. ———. 2003. Freedom Evolves. New York: Viking. Dreyfus, G. 1997. Recognizing Reality: Dharmakīrti’s Philosophy and its Tibetan Interpretations. Albany: Suny. ———. 2003. The Sound of Two Hands Clapping: The Education of a Tibetan Buddhist Monk. Berkeley: University of California. Hanna, R., and M. Maiese. 2009. Embodied Minds in Action. Oxford: Oxford University Press. Huntington, C.W. 1989. The Emptiness of Emptiness. Honolulu: University of Hawaii. Jackson, F. 2002. Epiphenomenal Qualia. In Chalmers, 273–281. Jackson, R. 1993. Is Enlightenment Possible? Ithaca, NY: Snow Lion. Karma, C. 1998. A Spacious Path to Freedom: Practical Instructions on the Union of Mahamudra and Atiyoga, translated by B. Alan Wallace. Ithaca: Snow Lion Publications. Lewis, D. 2002. What Experience Teaches. In Chalmers, 281–295. McGuinn, C. 2002. Can We Solve the Mind-Body Problem. In Philosophy of Mind, ed. D. Chalmers, 394–403. Oxford: Oxford University Press. Nagel, T. 2002. What Is It Like to Be a Bat? In Chalmers, 219–226. Schwitzgebel, E. 2010. Introspection. In The Stanford Encyclopedia of Philosophy. http://plato. stanford.edu/entries/introspection/. Sharf, R. 1998. Experience. In Critical Terms for Religious Studies, ed. M.  Taylor, 94–116. Chicago: University of Chicago. Shoemaker, S. 1981. Absent Qualia are Impossible. In The Philosophical Review XC, 581–599.

34

G. Dreyfus

———. 1982. The Inverted Spectrum. Journal of Philosophy 79: 357–381. Siderits, M. 2003. Personal Identity and Buddhist Philosophy. Burlington: Ashgate. ———. 2011. Buddhas as Zombies. In Self vs. Noself, ed. M. Siderits and D. Zahavi. Oxford: Oxford University Press. Thompson, E. 2015. Dreamless Sleep, the Embodied Mind, and Consciousness. In Metzinger & Windts, vol. 37, 1–19. Frankfurt: Open Mind. ———. 2014. Waking, Dreaming, Being: Self, Consciousness in Neuroscience. In Meditation and Philosophy. New York: Columbia. ———. 2020. Why I am not a Buddhist. New Haven: Yale University Press. Tillemans, T. 2011. Is There a Credible Buddhist Defense of the Mind or Could There be One? In Asian Philosophy and Analytic Thought, ed. G. Priest and K. Tanaka. (in preparation). Waldron, W. 2003. The Buddhist Unconscious. London: Routledge Curzon. Wallace, A. 2007. Contemplative Science. New York: Columbia Unversity Press. Wilson, M. 1984. Rebirth and the Western Buddhist. London: Wisdom. Wittgenstein, L. 1953. Philosophical Investigations. Oxford: Blackwell. Zahavi, D. 2005. Subjectivity and Selfhood. Cambridge: MIT.

Reasons and Conscious Persons Christian Coseru

1 Introduction What justifies holding the person that we are today morally responsible for something we did a year ago? And why are we justified in showing prudential concern for the future welfare of the person we will be a year from now? Whatever our answer to these questions, it seems that we cannot systematically pursue them without in one way or another referring to persons and their identity over time. But while there is widespread agreement that considerations about personal identity must be front and center in any such inquiries, such agreement falls short when it comes to specifying the criteria for personal identity, that is, what this identity necessarily involves, or consists in. Part of the difficulty is that an investigation into the nature of personal identity brings us to metaphysical questions about persons, their ontological status, identity conditions, and persistence over time. The challenge, then, is to pursue these additional questions without losing sight of the practical concerns that prompted them in the first place. In confronting this challenge, I will draw on two seminal works that have shaped the conversation on personal identity in contemporary analytic and Buddhist philosophy: Parfit’s Reasons and Persons and Siderits’ Personal Identity and Buddhist Philosophy. In engaging these works, my aim is to reassess the Reductionist View of personal identity in light of Buddhist Reductionism, a philosophical project grounded on the idea that persons reduce to a set of bodily, sensory, perceptual, dispositional, and conscious elements, which alone are real. Parfit is not only familiar with this Buddhist conception of personal identity, but thinks that the reductionist, no ownership position he defends, which takes persons both to exist and to reduce to their components, is true, and that, as he famously puts it, “Buddha would C. Coseru (*) College of Charleston, Department of Philosophy, Charleston, SC, USA e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_3

35

36

C. Coseru

have agreed” (1984: 273). My goal here is threefold: first, to review Parfit’s Reductionism position and evaluate its main arguments; second, to assess the extent to which Buddhist Reductionism, specifically as Siderits articulates it, supports Parfit’s psychological criterion for personal identity; and finally, to suggest some ways in which Buddhist conceptions of mind and consciousness can help to advance the contemporary debate on personal identity.

2 Reductionism and Personal Identity What is reductionism and how does it relate to the question of personal identity? The literal sense of the term, derived from the Latin ‘reducere’ (lit. ‘bring back’, ‘lead back’), captures rather well the philosophical notion of reduction: to reduce is ‘to lead back’ to something more fundamental than the thing in question. To say that the mental reduces to the physical, that motion reduces to kinetic energy, and that psychology reduces to biology, is to capture what it means for one thing (mind, motion, or psychology) to be brought back to the other (physical, kinetic energy, or biology). Reduction, then, stands for the view that, if entity x reduces to entity y, then y is prior. to, more basic than, or constitutive of x.1 As a philosophical term of art, ‘reduction’ is central both to metaphysical questions of personal identity and to questions in philosophy of mind about consciousness, agency, and the mind-body problem. One is a reductionist if one takes a particular theory or phenomenon to be conceivable in terms of, or reducible to, another theory or phenomenon.2 Parfit is one such reductionist, and Reasons and Persons contains his vastly influential theory of personal identity, which argues against the commonsensical, non-­ reductionist view of persons. According to the non-reductionist view, persons are distinct and discrete entities that exist over and above their bodies and psychological states. Their identity, then, is an irreducible, brute fact of existence, and cannot be explained or described in more basic terms. Whatever persons are, an account of their identity would have to employ person-level descriptive categories of embodied

 Within philosophy of mind, reductionism is concerned mainly with solving the mind-body problem, rather than, as was the case with Carnap (1934) and Neurath (1932/1983), with the unity of scientific theories. Hence, the main motivation for reductionism with regard to questions of personal identity is ontological parsimony. 2  In the case of intertheoretic reduction, the question is whether the atomic and continuum scale models can be bridged. While some reductionists think that it is in principle possible to derive continuum scale entities from atomic scale details (Batterman 1995), others (in particular those who favor top-down modeling) take the view that what we can hope for at most is identifying a set of intertheoretic relations (Berry 2001) rather than a hierarchy of theories operating at different scales. The upshot of this debate is that philosophical theories of reduction, which are typically modeled on scientific theories, must come to terms with the fact that reduction is problematic even in the case of the natural sciences. 1

Reasons and Conscious Persons

37

experience.3 One paradigmatic example for person in this non-reductive sense is the Cartesian Ego. For Parfit the view that there are such entities as Cartesian Egos or souls is representative of a particular intuition about personal identity, according to which we assume that questions of the sort ‘Will I survive the death of my body?’ or ‘Will I be the same person if I were to be teleported elsewhere?’ must have definitive answers. Regardless of whether or not we have answers to these questions at present, given their implications for personal identity, answers must in principle be available. There must be a way to settle these questions one way or another, perhaps on the basis of our very conception of what personal identity entails. What drives this intuition, argues Parfit, is the assumption that our identity must in some sense be determinate (1984: 214). If we reject this intuition and allow for the possibility that Cartesian Egos do not exist, then, argues Parfit, we are in a sense compelled to accept the view of Reductionism. One of the advantages of Reductionism is that it offers new possibilities for reconceiving the problem of personal identity on both metaphysical and empirical grounds. Since the body is the seat of our physical, affective, and mental lives, we may conceive of persons as (i) bodies or as (ii) entities that have bodies, thoughts, and emotions. The first conception can also be understood as an endorsement of one version of the identity view (persons just are bodies), while the second makes the case for the ownership view (persons are the sort of entities that have bodies, thoughts, and other kinds of experiences). As Parfit makes clear, whereas the ownership or constitutive view of personal identity can be easily entertained, and may even fit classical conceptions of persons as property-possessors, the identity view of reductionism opens the door for something more radical: Eliminative Reductionism, the best example of which is the mind/brain identity theory. For the eliminative reductionist, persons, much like nations, are ultimately reducible to their constitutive parts. Just as the concept of a nation cannot be the concept of an entity that is distinct from its people and its territory, so also the concept of a person cannot be that of an entity that exists over and above its body, perceptions, thoughts, and feelings. We may thus conclude, argues Parfit, “that, in that case, there are really no such things as nations. There are only groups of people, living together in certain ways” (2002: 656). What does it mean to extend this analogy and to claim that there are really no such things as persons? On Parfit’s general characterization of reductionism, the existence of persons, then, “just involves the existence of a brain and body, the  One way to understand the difference between the non-reductionist and the reductionist views of personal identity is along the simple/complex divide: the non-reductionist favors the simple, soul or Cartesian Ego, view, whereas the reductionist prefers the complex view that entails relations among physical and psychological states. Holding a soul view, of course, does not necessarily amount to holding a brute fact view, although in the absence of non-circular criteria for personal identity (of the sort required by the complex view) it is hard to tell them apart. What motivates recent defenders of the simple view (e.g., Baker 2013; Lowe 2009, 2013; Nida-Rümelin 2013; Swinburne 2013) is not commitment to a Cartesian Ego, but rather the notion that a specific, perhaps non-conceptual and pre-reflective, type of self-awareness seems indispensable to framing any account of personal identity. I address this point below, in §4. 3

38

C. Coseru

doing of certain deeds, the thinking of certain thoughts, the occurrence of certain experiences” (1984: 2011). The force of this reductionist move is captured by the ‘just consists in’ locution: all of the facts that pertain to a person’s existence––their particular type of body, their psychological profile, the kinds of relations that obtain between their mental and brain states, all of these admit of an impersonal description that neither requires nor presupposes this particular person exist. As a façon de parler, however, we do say that, because we can ascribe thoughts to them, thinkers exist. But just because we can infer from the content of their experience that there are thinkers, it does not follow that their existence is somehow separate from their thoughts. Parfit endorses the impersonal description thesis (1984: 225) even as he acknowledges, with Peter Strawson and Bernard Williams, that any talk of experiences, and of the various relations that obtain between them, is impossible without reference to the persons whose experiences they are. Although he would subsequently disavow it (Parfit 1999), on the grounds that a thoroughly impersonal conceptual schema would be ineffective for the kind of creature we are, the impersonal description thesis plays an important explanatory function for the view that personal identity over time just consists in physical and/or psychological continuity. To motivate both reductionism and the impersonal description thesis, Parfit provides the well-known example of the simple teletransportation (STT) machine. Suppose that in a future century when science is far advanced I am able to travel to a distant place, say Mars, simply by having a machine scan every cell of my body and broadcast the information in real time to another machine on Mars. There is one catch, however: my current body will be destroyed in the process of being scanned. On the up side, a qualitatively identical but numerically different body will be created on Mars. The initiation of the whole process is marked by a loss of consciousness, which is immediately gained once the complete body has been recreated on Mars. In principle, argues Parfit, we should be able to provide a description of the person’s states at t1, before entering the Earth STT, at t2, while being teletransported, and at t3, after exiting the Mars STT, without referring to the persons whose states they are. But while imaginary scenarios such as this do appear to support the impersonal description thesis, they raise a different set of questions, specifically about whether personal identity could be determinate. Upon waking up on Mars, am I the same or a different person? Assuming that the two persons are similar enough psychologically, and that the latter’s personality, consciousness, and memories are caused by the former’s, we would have to conclude that they really are the same person, but at different stages. In this case, causal dependence would indeed serve as a necessary and sufficient criterion for personal identity. But as Parfit concedes, such scenarios are not entirely unproblematic for the psychological theory of personal identity. Consider the second scenario, where an advanced teletransportation (ATT) machine has the same scanning capabilities as the STT, plus the added feature of leaving the scanned body intact. In this case, having undergone teletransportation, and on the assumption that the person I am on Earth is psychologically very similar to the one I am now on Mars, one would find oneself in two places at once. One advantage of this second scenario is that it provides a compelling reason for adopting some version of the ‘body’ theory of

Reasons and Conscious Persons

39

personal identity. But neither scenario addresses the many objections to the impersonal description thesis such as, for instance, that reductionism cannot explain in a non-question-begging way how exactly we get persons from impersonal psychophysical elements. The teleportation examples aim to drive home the point that personal identity is indeterminate, and that reductionists must contend with a situation in which questions about the mode of existence of persons may yield no definitive answers.4 Notwithstanding these problems, Parfit thinks such examples are compelling enough to warrant that we take reductionism seriously, and presents us with two distinct conceptions of reductionism about persons: (1) a person just is a particular brain and body and a series of interrelated physical and mental events; and (2) a person is an entity that is distinct from a brain and body and such a series of events; (1984: 211) Although he thinks that (1), which articulates the eliminativist version of reductionism, is justified in some instances (it is right to claim that “there were really no witches, only persecuted women” (2002: 657)), his conception of persons as rational and moral agents clearly compels him to accept (2). Furthermore, Parfit insists on the importance of thinking about persons not in terms of their mental states, which would entail that they must be the states of some entity, but in terms of the occurrence of certain experiences, deeds, and thoughts. Thus understood, all reductionists would accept the view that: (3) a person’s existence just consists in the existence of a brain and body, and the occurrence of a series of interrelated physical and mental events (1984: 211). Now contrast this view with that of the non-reductionist: (4) a person is a separate existing entity, distinct from their brain and body, and their experiences (1984: 210) But Parfit clearly distinguishes between two non-reductionist views. The common one, already noted, is that of a person as a purely mental entity, a Cartesian Pure Ego, soul, or spiritual substance. On this view, “personal identity over time does not just consist in physical and/or psychological continuity” (1984: 210). Rather, the Cartesian Ego is an altogether different metaphysical entity. The second non-reductionist view denies that persons are separately existing entities, distinct from their brains, bodies, and experiences:

 Parfit supports his claim regarding indeterminacy with the case of the Combined Spectrum, which, as the name suggests, involves all the possible variations of physical and psychological connectedness between such substantially altered person stages that the person I am now would in no way be connected with the one at the further end of the spectrum. 4

40

C. Coseru

(5) though we are not separately existing entities, personal identity is a further fact, which does not just consist in physical and/or psychological continuity (1984: 210) Parfit calls this the Further Fact View. The typical non-reductionist view articulates a familiar conception of personal identity, whose basic premise is that human beings (and perhaps other forms of life) must be conceived in terms of something essential, without which we could not ascribe to them the properties of persons. Unlike the typical non-reductionist view, the further fact view poses a conundrum: what we have here is both the notion that persons are not metaphysically distinct entities and the notion that their existence cannot be accounted for in purely impersonal terms. On the further fact view, then, the concept of person is simple and unanalysable. Parfit’s account of personal identity, which had initially drawn on the reductionist no-self view of Buddhist Abhidharma,5 may be more profitably associated with one Buddhist school of thought, Personalism (Pudgalavāda), that did indeed defend a conception of person as irreducible (more about the arguments adduced in its support in the next section). Before we turn to the Buddhist discussions of personal identity, a brief overview of Parfit’s psychological criterion of personal identity is in order.6 Parfit takes the ‘criterion’ for identity to stand for “what this identity necessarily involves, or consists in” (1984: 202). In the case of physical objects such as the pyramids or the Moon, that simply refers to the concrete spatio-temporal continuity of the object in question. So far as our physical body is concerned, the criterion of personal identity would simply consist in the persistence of this body and brain over time. Parfit offers what he assumes is a better version of the ‘physical criterion’: what is necessary for personal identity is the persistence of enough of the body and the brain over time, such that even if we were to lose some of the body and brain (through amputation or hemispherectomy) the same person would continue to exist. Nonetheless, on the physical criterion of personal identity, the case of STT would not count as persistence over time and neither would that of rebirth, which presuppose that something of oneself continues beyond the destruction of the body. This is largely the reason why Parfit thinks the physical criterion does not suffice, and why instead he puts forward a psychological criterion of personal identity, which he understands as involving two relations:

 The first to draw attention to their similarity was Collins (1985), though in a subsequent article (Collins 1997) he softened his position, regarding the two views as similar rather than identical. Similar attempts to draw together the two views and to adopt a Parfitian conceptual schema in pursuing various projects in Buddhist philosophy are found in Bastow (1986), Stone (1988, 2005), Giles (1993), Duerlinger (1993), Basu (1997), Siderits (1997, 2015), Perrett (2002), Ganeri (2007), and Sauchelli (2016). 6  Parfit first articulated his views on personal identity as early as his 1971 paper “Personal Identity”, so that by the time Reasons and Persons was published, his views were widely known, and had already generated a great deal of discussion. 5

Reasons and Conscious Persons

41

Psychological connectedness is the holding of particular direct psychological connections. Psychological continuity is the holding of overlapping chains of strong connectedness (1984: 2006) Although Parfit takes connectedness, which is a matter of degree, to be a more important relation, he does not think it serves as an adequate criterion for personal identity. Strong connectedness is a transitive relation, since the person I am today is strongly connected to the one I was yesterday, and to the one I was 3 days ago. But it cannot be true that the person I am today is strongly connected to the one I was 20 years ago, and so, despite its importance, the strong relation of psychological connectedness is too problematic for personal identity: “Because identity is a transitive relation, the criterion of identity must also be a transitive relation. Since strong connectedness is not transitive, it cannot be the criterion of identity” (1984: 206). What about psychological continuity? Parfit uses the continuity relation to introduce his ‘psychological criterion’ of personal identity: There is psychological continuity if and only if there are overlapping chains of strong connectedness. X today is one and the same person as Y at some past time if and only if (2) X is psychologically continuous with Y, (3) this continuity has the right kind of cause, and (4) it has not taken a ‘branching’ form. (5) Personal identity over time just consists in the holding of facts like (2) to (4). (1984: 206)

The key elements here are (2) psychological continuity and (3) right causal determination, and Parfit uses right causal determination to further distinguish between narrow, wide, and widest versions of the psychological criterion depending on whether the cause is identifiably the correct one, any reliable cause, or any cause whatsoever. These additional criteria provide a basis for introducing the memory element, since being able to remember past experiences as one’s own is essential to the psychological criterion of personal identity. Memory thus stands as an example of strong connectedness between, for instance, an experience once had and later remembered or an intention once formed and later acted upon. These sorts of direct causal connections that memory, intention, and action provide are typically of a short duration. I seldom act on intentions formed 20 years ago, and distant childhood memories have already been embedded in a complex self-narrative that had undergone revision and embellishment over time (even though Parfit takes it to be a logical truth that we can only remember our own experiences).7 So, while there may be few direct connections between, say, the 50-year-old adult and the 10-year-old child, they are still indirectly connected through multiple overlapping chains of direct connectedness. The overlapping chains model thus serves to illustrate Parfit’s

 Parfit gets around the problematic issue of the relation between episodic and semantic memory by appealing to q-memory (quasi memory––a new category of memory first introduced by Shoemaker (1970)). According to Parfit, “I am q-remembering an experience if (1) I have a belief about a past experience which seems in itself like a memory belief, (2) someone did have such an experience, and (3) my belief is dependent upon this experience in the same way in which a memory of an experience is dependent upon it” (1971: 15). 7

42

C. Coseru

understanding of the criterion for psychological continuity, whose functionality is such that we can talk of the child I was once and the adult I am today in one breath. While the narrow account of causal determination serves as a plausible way to understand persistence over time, it cannot accommodate the simple teletransportation case. For this reason, Parfit thinks we need the wide and even widest views of causality. The wide view allows any reliable causal chain to serve as a criterion for personal identity. The widest view is even more permissible, allowing for any causal chain to play this role (a requirement for the persistence of identity in the ATT case). Since Parfit claims that there is no good reason to prefer the narrow view over the others, we are confronted with a problematic notion of persistence: just what it means to say the person I am now on Mars is psychological continuous with the one I was on Earth an hour ago is not exactly clear. And it is even more puzzling to entertain what it would be like to talk of personal identity between the two versions of myself in the ATT case. What is clear is Parfit’s motivation for articulating these two views, both of which allow for a conception of persistence over time that reduces the person to certain chains of causal connectedness between mental and physical states. The issue is not whether the same individual persists overtime, including in the ATT case. Rather, the problem is how best to guarantee there is enough resemblance between the different person stages. Even so, in articulating his psychological criterion, Parfit does not aim to secure a strong foundation for personal identity; rather, he wants to demonstrate that the persistence of identity is not an important factor when it comes to matters of moral and practical concern. Now that we have identified the two primary elements of Parfit’s theory of personal identity, let us consider to what extent Buddhist Reductionism supports Parfit’s psychological criterion for personal identity.

3 Buddhist Reductionism, Personalism, and the No-Self View As is well known, in an effort to demonstrate that his views are not the product of a particular culture and epoch, but rather apply to all people at all time, Parfit turns to, inter alia, Buddhism, where he finds an early reductionist stance not unlike his own. As he notes: I claim that, when we ask what persons are, and how they continue to exist, the fundamental question is a choice between two views. On one view, we are separately existing entities, distinct from our brain and bodies and our experiences…The other view is the Reductionist View. And I claim that, of these, the second view is true… Buddha would have agreed (1984: 273)

Whether or not Buddha would have agreed depends in large measure on whether Buddhist Reductionism articulates something close to a psychological criterion of personal identity. I will argue that it does, although in a way that emphasizes the structural dynamics of conscious experience over accounts of psychophysical causal relations (of the sort that Parfit’s reductionist view of persons endorses). One of the

Reasons and Conscious Persons

43

clearest articulations of Buddhist Reductionism about persons is found in The Questions of Milinda (Milindapañha), an eloquent para-canonical work in the style of a Platonic dialogue detailing an exchange between Nāgasena (a Buddhist monk) and Milinda (better known as the Bactrian King Menander I) aimed at articulating a nominalist conception of personal identity. In a much-quoted passage, Nāgasena explains that, although he responds to a certain appellation––his name––there is no corresponding referent: Sir, I am known as ‘Nāgasena’; my fellows in the religious life address me as ‘Nāgasena’. Although my parents gave (me) the name ‘Nāgasena’…it is just an appellation, a form of speech, a description, a conventional usage; ‘Nāgasena’ is only a name, for no person is found here (Horner 1963: 34)

Parfit is familiar with The Questions of Milinda, and appeals to this and to several other excerpts from Buddhist texts8 to make the case that “the Reductionist View is not merely part of one cultural tradition. It may be, as I have claimed, the true view about all people at all times” (1984: 273). That Parfit’s view and that of the Buddhists bear a strong resemblance is both obvious and well documented.9 What is less obvious is the extent to which Buddhist metaphysics supports the Reductionist View. To answer this question, I will turn to one of the most influential attempts to show how Buddhist resources could help Parfit address some of the difficulties of his reductionist account while working out the implications of a Parfitian-style conceptual framework for Buddhist philosophy in general, and for Buddhist Reductionism in particular: Siderits’ Personal Identity and Buddhist Philosophy (1997/2015).10 Following Parfit, Siderits too acknowledges that reductionism comes in a variety of forms, and that when situated in the Buddhist context, one must not overlook the (historical) fact that not all Buddhists are reductionists. Nonetheless, within the context of early Buddhist scholasticism (viz., Abhidharma), all Buddhist philosophers who, in addition to being reductionists about composite entities such as chariots, forests, and armies, also take a reductionist stance with regard to persons, may be identified as Buddhist Reductionists. As Siderits explains, a distinctive feature of Buddhist Reductionism is the two truths framework, which takes sentences to be either conventionally or ultimately true depending on whether they lead to successful practice or to what is taken to be ultimately real (2015: 16). For instance, Buddhists take partite entities such as chariots to be conceptually constructed and thus not ultimately real. Thus, chariots reduce to ‘axle’, ‘felly’ and ‘linchpin’, which  This text, along with Vasubandhu’s Treasury of Metaphysics (Abhidharmakośabhāṣya) and Buddhaghosa’s Path to Purity (Visuddhimagga), serve as his main source of Buddhist ideas about personal identity. Parfit includes several passages from these texts, excerpted from Stcherbatsky’s (1919) seminal work on the Buddhist no-self doctrine. He also cites Collins’ (1982) landmark study of the early Buddhist doctrine of anatta. See also Collins (1997) for a response that considers the extent to which support for a reductionist account of personal identity on the Parfitian model can indeed be found in the Buddhist tradition. 9  For an overview of the impact that Parfit’s categorical framework and theoretical perspectives (on metaphysics and ethics) have had on Buddhist philosophy, see Hanner (2018). 10  All references are to the second edition, Siderits (2015). 8

44

C. Coseru

in turn reduce to their elemental parts, and so on. On this account, only impartite unanalysable entities are ultimately real. Siderits reworks the two truths framework into an account of semantic insulation, which he uses to justify the Buddhist Reductionist view that partite entities cannot be ultimately real. As he himself admits, this view, which articulates the position of mereological nihilism, “will strike many as giving an implausibly austere ontology” (2015: 14). Siderits finds some of the arguments adduced in its support compelling, though, as I will argue, for reasons that are not entirely clear. What, then, motivates Buddhist Reductionists’ mereological nihilism? The answer lies in the widely-shared but, according to the reductionist, obviously mistaken non-reductionist stance that most people adopt: When the Buddhist Reductionist claims that strictly speaking chariots do not exist, that ‘chariot’ is a mere conventional designator for a collection of part arranged chariot-wise, this is meant to serve as an example of widely held belief that turns out not to be strictly speaking true, but also of a belief whose acceptance is perfectly understandable given the demands of everyday life. The point is not to get us to stop believing there are chariots. The point is to help us see how we could all be mistaken about ‘I’. For the mechanism that generates belief in a real ‘I’ is the same; hypostatization. What we have in both cases is a many masquerading as a one (Siderits 2015: 98)

Does use of the first-personal pronoun and handling of the middle size dry goods that populate our mundane existence have such dramatic effects on belief generating processes and embodied practices? And if they do, how exactly can the dualist semantics of the two-truths framework undo our ingrained hypostatization proclivities? The mereological nihilist singles out the effects on belief that configuration processes––by which various parts in the world come to be arranged chariot-wise and person-wise––can have when the demands of everyday life constrain one’s ontological outlook. Speedy and effective long-distance travel may demand that one avails oneself of various modes of transportation. Should we admit cars, trains, and airplanes in our final ontology? Sure, the car in the commercial is more than a wheeled and motorized mode of transportation; it is also a symbol of comfort, elegance, efficiency, and status. It has its own event horizon, and it promises to augment human experience in profound ways. Nonetheless, such demands should be recognized for what they are: alluring invitations to reify that which is ultimately only a collection of parts. The belief’s acceptance may well reflect conventional linguistic practices that, once fixed, make possible assertions in which it is actually true that there are chariots and cars. The problem, as Siderits carefully explains, is that with such customary linguistic practices in place it becomes possible to assert both that there are such things as chariots and that there are certain parts that can be arranged chariot-­wise. Semantic insulation between the two types of discourse may indeed give rise to anomalies, especially if one were to assert that both types of entities are real. But Siderits, following Thomasson (2007), thinks such anomalies indicate that the conditions under which something comes to be called a ‘chariot’ are not always fixed by rules that apply to the discourse about parts, and so the existence of chariots, and of ordinary objects in general, can be vindicated.

Reasons and Conscious Persons

45

However, it is an open question whether the dualist semantics of the two truths framework brings Buddhist Reductionism in line with the particular type that Parfit favors. Siderits thinks that it does, and considers a version of the teleportation scenario that aligns the two accounts, but in a way that does not eliminate the problem of indetermination. The question whether parts arranged a certain way are identical with or distinct from the chariot does not admit of a determinate answer, and neither does the question whether the swapping of parts from one to another chariot-wise arrangement results in an identical or different arrangement. Siderits’ proposed solution involves metalinguistic analysis: What we can do is ascent to a meta-language and discuss those relations between our uses of chariot on the one hand and ‘axle’ and the like on the other that explain why, when the parts are assembled in a certain way, we say there is a chariot, and why there may be cases where we do not know whether to say there is a chariot, or to say that there is diachronic identity between some present chariot and an earlier chariot. We can talk about the whole, or we can talk about the parts, but we cannot talk about both in the same breath. (2015: 110)

The metalinguistic analysis solution may seem compelling enough in the case of chariots and other inanimate objects, but it confronts a paradox: since the parts taken together in some degree of combination do not add up to a chariot, there are no such things as chariots. Parfit confronts the paradox while answering an objection from Bernard Williams against the view that we can get personal identity from a psychological spectrum. The objection concerns the sorites paradox. Run in the reverse, the argument considers a scenario in which by means of small enough surgical interventions a surgeon could cause a person to cease to be psychologically continuous with previous stages of herself. If a surgeon could alter in small increments the neurochemistry of the brain in such a way as to profoundly transform the psychological features that make up who you are, does that mean you have ceased to exist? Parfit’s response considers the semantics of vague terms, and his solution to the sorites paradox comes in the form of an appeal to stipulation: we simply decide how many grains of sand to call a heap. But whereas stipulation may work for natural kinds such as forests it does not work for artifacts such as chariots. A chariot is not simply an arbitrary number of parts, but rather a definitive set of parts arranged such as to function as a ‘chariot’. These may vary depending on whether one considers extra features and embellishments, but one could agree on a minimum number of parts without which the concept chariot could not be meaningfully applied to parts arranged chariot-wise. But even if we adopted such a provisory stance, does that mean the argument would work in the case of persons? If Parfit is right, the physical criterion is not enough for personal identity, which is why we need to consider the full spectrum of psychological states. Since persons are defined primarily in terms of the subjective and phenomenal character of their conscious mental states, how does Buddhist Reductionism solve the personal identity problem? It would seem that any account of persons requires person-level descriptive categories, a point that, as we noted above, Parfit himself concedes.

46

C. Coseru Even Reductionists do not deny that people exist. And, on our concept of a person, people are not thoughts and acts. They are thinkers and agents. I am not a series of experiences, but the person who has these experiences. A Reductionist can admit that, in this sense, a person is what has experiences or the subject of experiences. This is true because of the way we talk. What a Reductionist denies is that the subject of experiences is a separately existing entity distinct from a brain and body, and a series of physical and mental events. (1984: 223)

But in retaining a meaningful conception of ‘person’ as thinkers and doers, Parfit’s view departs from mainstream Buddhist Reductionism only to join the ranks of the Buddhist Personalists. As expounders of the ‘personalist view’ (pudgalavāda), Buddhist Personalists do agree with the general tenet of Buddhist Reductionism that there are no such things as enduring substantive selves. They also retain the bundle view and take ‘persons’ to be conceptualized in dependence upon the psychophysical aggregates. Where they differ, however, is in admitting that persons, although neither identical nor different from the aggregates, are nonetheless real. It is unclear whether for the Buddhist Personalists persons enjoy the same ontological status as the aggregates, in part because they exhibit novel properties that neither of the aggregates do. What is clear, however, is that although the Personalists admit that the notion of ‘person’ is conceived in dependence upon the aggregates, they do not regard it as a conceptual fiction. Sustained efforts to reject Buddhist Personalism as an orthodox position form an integral part of Buddhist Reductionism. Historically, Buddhist Personalists were understood to have forgone commitment to the no-self doctrine. However, it is now possible to accommodate their views on an emergentist or supervenience account of phenomenal properties.11 It is instructive that what makes Buddhist Personalism the target of criticism is a certain insistence on taking what we would now call the first-person perspective seriously, particularly as it finds articulation in discussions of the relation between consciousness and content. The key issue under dispute is the concept of person itself: if it stands simply for the collection of aggregates, then the Buddhist Reductionist is right to insists on its fictional status. However, if the concept of person is meant to capture something determinate, such as the view that persons are self-conscious subjects, then it is hard to see how the reductionist can make the case that the concept of ‘person’ lacks a proper referent. Consider, for instance, Vasubandhu’s refutation of personalism, and the argument that, since an individual

 Buddhist Personalists (Pudgalavāda) include the Vātsīputriyas and the Sāṃmitīyas, though their exact views on persons are hard to reconstruct since they are mostly preserved in unsympathetic treatments by their critics, mainly in the Points of Controversy (Kathāvatthu), Vasubandhu’s Treasury of Metaphysics (Abhidharmakośabhāṣya), Harivarman’s True Attainment Treatise (Satyasiddhiśāstra), Devasarman’s Body of Consciousness (Vijńānakāya), and Śāntarakṣita’s Compendium of Principles (Tattvasaṃgraha). The only source that presents their views unbiasedly is the Sāṃmitiya Nikāya Śāstra, now preserved only in Chinese (an English translation in Venkataramanan (1953)). The prevalent interpretation, gleaned from the Points of Controversy, is that Pudgalavādins take persons to be ultimately real. According to most of the other sources, Pudgalavādins think persons occupy an ontological position somewhere between conventional and ultimate reality. See Priestly (1999) for detailed studies of Buddhist Personalism, and also Châu (1999) for a general survey and reconstruction of their views. 11

Reasons and Conscious Persons

47

can only possess ‘person-properties’ on account of possessing some sort of essence (svabhāva), and since Buddhist metaphysics is anti-essentialist, there is no ontological basis for admitting entities defined in terms of their possession of person-­ properties.12 On the argument put forth here, person-properties are those very things that allow us to use the first-person pronoun meaningfully in sentences such as ‘I am walking to my office’ or ‘I am thinking about the conversation we had yesterday’. The motivation for resisting such language use, according to Vasubandhu, reflects its unintended consequences: conceiving of the aggregates together as forming a person is a slippery slope to thinking there are such things as selves. To be clear, Vasubandhu does not reject the view that we do appear to ourselves to be more than just a body, or a bundle of feelings and thoughts. Rather, he rejects the notion that feelings, thoughts, and the body are something that we possess. Such a view would entail an existence apart from the aggregates. But since we are nothing but thoughts feelings, perceptions, memories, and bodies of a particular shape, size, and gender, no such residual existence is anywhere to be found. In claiming that persons are conceived in dependence upon or in reliance upon the aggregates, though neither identical nor different from them, the Personalists invoke the conceivability principle. The dispute, then, hinges on whether or not the object of conception must be identical with its causal basis. If no such identity relation were to obtain, then conceiving would lack reference, and would not point to anything in the world. But even on a strongly non-referential view, conceiving must be regarded as an effective epistemic practice that entails, inter alia, reference to nonconceptual entities (regardless of their ontological status). On a referential view of conception, thus, to conceive of persons is to conceive ourselves as persons. In conceiving of becoming a father I conceived of myself as a father. In conceiving of unicorns, I conceive of them as possible or impossible. On the personalist view, the aggregates may serve as a causal basis for our conception of person without being identical with it. Conception is indeed reliant on the presence of some manifest phenomena, but conceiving of ‘ourselves’ in dependence upon the aggregates does not make ‘us’ identical with them. We conceive of fire in relation to fuel and of friendship in relation to common interest, but fire is not identical with fuel, nor is friendship identical with common interest. The Buddhist Personalist, pace Vasubandhu, wants to make the case that the object of conception is not always identical with its causal basis, which is why we can both retain a meaningful conception of person and be reductionists about selves. It would seem that this particular version of Buddhist Reductionism, which retains the concept of person (though not that of self) would support Parfit’s conception of personal identity in the STT case, though not in the ATT case. Since the person is conceived in dependence upon the causal series, though not identical with it, I am sufficiently connected to the previous instance of myself prior to entering the Earth STT.  The conceivability criterion of personalism makes the ATT case  The refutation, known as Treatise on the Negation of the Person, is appended as Chapter 9 of Vasubandhu’s Treasury of Metaphysics (Abhidharmakośabhāṣya). See Duerlinger (2003) for a detailed treatment of this section of Vasubandhu’s treatise. 12

48

C. Coseru

problematic, given that there is no way to differentiate between conceiving of myself in dependence upon the causal series now present on Mars as opposed to those still on Earth. Might the person so conceived by Buddhist Personalists be a further fact, and could that be a mitigating factor against mereological nihilism? Siderits considers this objection by engaging some recent work in ontology, specifically Merricks’ (2001) defense of eliminativism. Merricks’ view is that strictly speaking there are no such things as statues, only “atoms arranged statuewise”. In denying the existence of macroscopic objects, Merricks is not simply arguing against conventional uses of language, of the sort that take statement such as “there exists a statue” to mean “there exist atoms arranged statuewise” (2001: 3). Nor is he denying the existence of statues because statues are nothing over and above atoms arranged statuewise, and thus not mereologically distinct from their parts. Rather, he puts forward a complex argument from causal overdetermination: admitting that there exist such things as statues causally overdetermines their effects (because whatever effect can be attributed to a statue is already an effect of its microphysical parts). Since such overdetermination is not observed, we must admit that there are no such things as statues.13 However, Merricks think the causal overdetermination argument does not apply in the case of human beings, because humans have causal powers that exceed those of their constitutive psychophysical parts. Most notably, humans instantiate consciousness, a property that does not supervene on the body’s microphysical properties. Merricks offers an interesting argument in support of this view operating on the premise that conscious and subjective mental properties bear the mark of being intrinsic. ‘Intrinsic properties’ are those properties “that an object can exemplify even if that object and its parts (if any) are the only objects that exist” (2001: 92). What is distinctive about humans is precisely that there is no metaphysical necessity that their intrinsic causal powers be implied by the existence of intrinsic properties of, and spatiotemporal and causal interaction among, their constituent elements. The argument goes as follow: suppose P, a conscious being, accidentally slices off her left index finger and thus shrinks. Suppose that at the very same moment of amputation, the atoms that compose her remain just as they were before amputation. The implication is that those atoms that are constitutive of P after amputation, are also constitutive of it immediately before amputation. Now, consider the pre-amputation finger-complement that is not identical with P (on the grounds that P has a part, the finger-complement, which the finger-complement itself simply lacks). If that is the case, and if we assume a relational view of consciousness (as a composite of atoms grouped by spatiotemporal and causal interrelations), then we must concede that before amputation there are two consciousnesses present, one for P and the other for the finger-compliment. But since we only ever have one such entity, P, the thesis that there are two consciousnesses must be false. Merricks thinks that to

13

 For a response to the argument from overdetermination, see Sider (2003).

Reasons and Conscious Persons

49

postulate the existence of a conscious pre-amputation finger-complement is to engage in an unacceptable multiplication of persons. If it is true that the schema of the overdetermination argument does not apply to the effects a human causes by virtue of being conscious, then humans cannot be treated as epiphenomenal entities. As composite entities capable of causing effects that our parts cannot redundantly cause by virtue of being conscious, we humans can resist the eliminative sweep of the reductionist. Now, Siderits is willing to grant that, were the argument from overdetermination to succeed, and the effects of conscious states are not systematically overdetermined, then we would have a forceful case for there being persons with causal effects that are irreducible to those of their constitutive psychophysical elements (2015: 112). This argument may not rehabilitate composite entities such as chariots, baseballs, or trees. But it would establish Parfit’s further fact version of Non-Reductionism: persons are not just suitably arranged psychophysical aggregates that exhibit novel causal powers. Rather, persons are what these specific aggregates are constituted as, which is why persons do not reduce to the aggregates in the same way that the chariot reduces to its component parts, and the forest to the collection of trees. The difference between Buddhist Reductionism and Buddhist Personalism should be clearer by now: whereas the reductivist claims that mental states strongly supervene on physical states and thus are reducible to the latter, the personalist takes them to possess novel causal powers that cannot be ontologically analyzed in terms of the properties of the physical states upon which they supervene. It is possible to interpret Buddhist Personalism as endorsing an emergentist view of persons and not simply a non-reductive physicalist view, but that would dependent on how we interpret the ‘in reliance upon’ clause. Siderits’ response to the arguments against causal overdetermination takes up the problem of compatibility between four different positions Buddhist Reductionism must consider when debating the ontological status of persons: (1) dualism, (2) reductive physicalism, (3) non-reductive physicalism, and (4) emergentism. When considered in the context of debates about wholes and parts, these positions correspond respectively to (1) mereological realism, (2) reductive mereological nihilism, (3) non-reductive mereological nihilism, and (4) emergentism (2015: 114). Drawing on contemporary debates on these issues,14 Siderits admits that the non-reductive physicalist position (viz., mental states supervene on physical states but are not reducible to them) ought to, at least in principle, be compatible with emergentism (viz., mental states do supervene on physical states, but since they exhibit novel causal properties, physicalism must be false). The question, however, is whether conscious mental states, and by extension the person whose states they are, are fundamental despite being dependent upon the intrinsic properties of, and the spatiotemporal and causal interaction among, their constituent elements. Siderits thinks attempts to answer this question that appeal to the notion of downward causation face an explanatory gap, and so efforts to close that gap inevitably “push the non-reductivist in the direction of reductionism and

14

 In particular, Crane (2010), Noordhof (2010), and Kistler (2010).

50

C. Coseru

identity” (2015: 116). So, the position is best classified as a version of the “non-­ reductive supervenience view” (2015: 117). But he does not agree that Buddhist Personalism offers a legitimate way to preserve the irreducibility of person-­ properties necessary for, say, negotiating moral desert, and finds functionalist accounts of mental content compelling enough the resist the non-reductionist move. It is debatable whether the further-fact view of Buddhist Personalism is as inherently unstable as Siderits claims. In the next section, I will argue that it is not, and that considerations about self-consciousness threaten certain aspects of the Reductionist View, and thus render the conception of person (as conceived in dependence upon the aggregates, though neither identical nor different from them) essential for psychological accounts of personal identity.

4 Reflexivity, Agency, and the Unity of Conscious Experience The much-quoted passage from The Questions of Milinda (Milindapañha) discussed above––in which Nāgasena argues against the ultimate reality of persons–– stands in sharp contrast to an equally important, but less celebrated, critical statement from his interlocutor, King Milinda, concerning the moral and pragmatic consequences of the no-self doctrine: If, most revered Nāgasena, no person is apprehended, then who gives you the offerings that you receive as a Buddhist monk? Who enjoys those things? Who practices disciplined conduct? Who enters into contemplation? Who attains to the goal of nirvāṇa, which is the outcome of the path of cultivation? If, revered Nāgasena, one was to slay you, his would not be the crime or murder. (Horner 1963: 100)

As this passage makes clear, the problem of personal identity is not exclusively metaphysical, but engages central issues in ethics and moral psychology. Do the moral and mental forms of cultivation at the heart of Buddhist practice demand a robust notion of agency? Buddhism is indeed unique in articulating a theory of action that, it seems, dispenses altogether with the notion of agent causation. Even though practices of moral and mental cultivation form an integral part of the Buddhist path, there is no stable self or agent who bears the accumulated responsibility for initiating those pursuits, and seemingly no normative framework against which some dispositions, thoughts, and actions are deemed felicitous, and thus worthy of cultivation, while others are not so deemed. The agent-neutral metaphysical picture of Buddhist scholasticism thus challenges traditional conceptions of moral agency. One way to address this challenge is to claim that the discourse of ‘persons’ and the discourse of ‘causes’ are compatible in so far as they belong in two distinct and incommensurable domains.15 Compatibilists who adopt this position typically cite evidence from social and cognitive psychology to show that any robust notion

 Proponents of this move include, inter alia, Siderits (1987, 1997, 2008), Flanagan (2002), and Meyers (2014). 15

Reasons and Conscious Persons

51

of agent causation must be incoherent.16 But, as I have argued elsewhere, such solutions compromise traditional conceptions of moral responsibility and render ethical conduct indistinguishable from merely pragmatic acts (Coseru 2016). Indeed, despite the dominance of an ultraminimalist account of agency, there are good and compelling reasons to give Buddhist Personalists credit for insisting that in so far as the aggregates operate in a person-constituting way, persons are ineliminable from our discourse about agency and moral responsibility.17 Does Buddhist Personalism provide a closer analogue for Parfit’s theory of personal identity (and its psychological reductionism) than Buddhist Reductionism, with its stated mereological nihilist view that there are no such things as composite entities? I think that it does, for conceivability reasons: the psychological criterion is plausible only to the extent that I can conceive of myself in dependence upon a sufficiently similar series of aggregates. But the conceivability principle concerns the epistemological, rather than the ontological, dimension of personal identity, which brings up an altogether different set of considerations, specifically about the relations that obtain between self-referential mental states (those that presuppose the notion of oneself as an agent) and self-consciousness. The epistemological dimension is framed by a different set of questions that pertain not to what awareness supervenes on but to its structure and specific properties, specifically: What in particular accounts for a mental state becoming an instance of self-consciousness? Does self-consciousness require that its self-referential relation is present to itself as an object? If we can answer these questions, we can make progress in understanding the relation between self-referential mental states and self-consciousness. And if we can get clarity about the nature and character of self-consciousness, we are in a better position to understand what it is that makes us persons.18 Recall the schematic analysis of the five aggregates that informs the Buddhist Reductionist account of personal identity. In this analysis, only the body is a physical aggregate stricto sensu. Feelings, perceptions, dispositions, and consciousness can acquire an objective aspect, but are not properly speaking empirically tractable phenomena. Nor are they things, that is, abstract entities with well-defined properties and functional characteristics. What this reductive analysis of persons in terms  See Caruso (2012), Smart (2006), and Wegner (2002) for various attempts to prove the illusory nature of experiences of mental causation. While not conclusive, Nahmias et  al. (2004) review experimental data that seems to favor compatibilist over incompatibilist accounts of free will. 17  See Carpenter (2015) for a sympathetic treatment of Buddhist Personalism that proposes a new, constitutive or developmental model for understanding the Pudgalavāda, aimed at extending the categorical framework of Buddhist metaphysics. On this model, the aggregates (of perception, consciousness, etc.) relate to each other in a person-constituting way, which she regards as “the most minimal explanation possible” that must be accepted if we are “to do justice to the phenomena without adopting a substantialist theory of self” (2015: 27). 18  A similar strategy is at work in Garrett (1998), though the focus there is on connections between ‘I’-judgments and self-consciousness. Garrett finds Anscombe’s view that the first-person singular pronoun is not a referring term unsupported, and builds a strong case for language use as an intentional activity: although not all language use exhibits intentional activity, self-consciousness must be understood as consisting “in the presence of an intention to self-refer” (Garrett 1998: 96). 16

52

C. Coseru

of their constitutive features is meant to capture is not what person are made of, but rather what human experience is constituted as: specifically, as series of intentional and self-referential mental events. Consider the paradigmatic example of pain: as a sensation, pain is not reducible to the physical substrate, say a finger, in which it is instantiated (nor presumably to a mere physiological response). Rather, pain is constituted as a distinctly qualitative phenomenon whose intentional content cannot be dissociated from its subjective aspect. There is no such thing as generic or impersonal pain (understood strictly in terms of, say, the activation of Aδ- and C- fibers following an intense stimulation of nociceptors) apart from phenomenologically foregrounded sensations of some kind: of burning, stinging, or throbbing. Feelings may define the quality of the impressions that result from contact with an object, with the implication that they perhaps stand in a causal relation with these objects. In the schematic analysis of Buddhist scholasticism, they are categorized as mental states conditioned by habitual tendencies (vāsanā), which, in turn, they condition. Likewise, apperception (saṃjñā), the capacity to make intelligible or cause to be understood, although dependent on a multiplicity of psychological factors, captures the datum of experience only as unified into a single percept (since what Abhidharma psychology understands as a simple apprehension and identification of a sensible trope actually involves a significant amount of unification of sensory input that goes on behind the scene). Volitions too fit the same profile, with one important difference: rather than attending to the object at hand or providing a sort of transcendental unity of apperception, they bring forth future states of existence. As dispositions to act in certain ways, they cleave the mental domain into two classes of conditioned phenomena: those that are internal to consciousness, such as, for instance, obsessive dispositions like greed and delusion, and those that are dissociated from it, usually taken to refer to latent dispositions typically comprising various biological and physical traits.19 This dynamic model of personal identity is not incompatible with the notion that there are phenomenally primitive features of experience, features that any analysis of the structure self-consciousness must ultimately reveal. When the Buddhist Personalists insist that a mere functional account of the aggregates will not suffice to explain why an action counts as murder they draw attention to the specificity and individuality of a given bundle of aggregates, and hence, of its actions and consequences: “If the self were absolutely non-existent, then there could not be killing nor would the killer have killed anything. There would be nothing like theft and robbery…good and bad would yield neither freedom nor bondage; even bondage would have no one bound. There would be neither the doer nor the deed not any result thereof” (Venkataramanan 1953: 177). Indeed, understanding why something is categorized as murder and not simply as the rearranging of aggregates presupposes a conception of intentional action that is unintelligible without reference to persons. If living beings were mere conventional designations, as Nāgasena contends, then there would be no non-arbitrary way to assign guilt in the case of killing

 See, e.g., Treasury of Metaphysics (Abhidharmakośabhāṣya) 2.23–34  in de la Vallée Poussin (1980: 45–62). 19

Reasons and Conscious Persons

53

a sentient being, since killing a living ox would be no different than destroying a clay ox.20 Unlike clumps of clay, living beings are characterized primarily in terms of their capacity for responsive and intentional action: they can both do things and have things done to them in a way unavailable to insentient objects. Persons, unlike mountains or chariots, are not simply generic unities of aggregates, but self-­ disclosing wholes. Persons are what they are by virtue of the fact that their aggregates are perceived to belong together.21 Aggregates grouped together in a person-constituting way are specific to themselves in a way that mountains grouped together or chariot parts assembled together are not. What are the implications of this position for Buddhist Reductionism in particular, and for the Reductionist View of personal identity in general? There are various responses to this question. One may take the reflexive structure of self-­consciousness to point to deeper aspects of consciousness as conscience,22 or delineate several ways in which pronominal and indexical uses of the ‘I’ can engender a false sense of self.23 One may also claim that the perceived unity of consciousness results from certain patterns of attention to, and grasping of, various features of experience.24 Whether one takes self-consciousness to have a distinctive structure and clearly specifiable content, for the Buddhist Reductionist no minimal account of agency can escape the antirealist stance of the no-self doctrine and its implications, namely that all modes of ‘I-making’ (and ‘I-sensing’) characteristic of self-referential mental states must be conceptually constructed. This is precisely the reason why Siderits, for instance, thinks that we should take only causally efficacious particulars as providing metaphysical grounding (Siderits 2015: 115). Buddhist Reductionism does indeed pose a significant challenge for conceptions of phenomenal consciousness that take consciousness to consist in more than just a succession of associated ideas or a construct of post-hoc rationalization. However, by treating its contents as transient episodes arising within a continuum of causally interconnected states, Buddhist Reductionism itself confronts an epistemic explanatory gap: if all there is to the experiential domain is a stream of momentary, objectdirected mental events, just what it means for an object to be known, and by whom? Can Buddhist Reductionism, with its account of primitive atom-like ‘qualitons’ of experience, provide an adequate conceptual basis for mapping out the distinctly self-intimating character of knowledge episodes? To answer this question let me briefly consider an epistemological attempt to negotiate the difference between egological and non-egological accounts of self-consciousness, drawing principally on the work of Dignāga.

 For Harivarnan’s discussion of some of the ethical implications of the no-self view in Satyasiddhiśāstra, see Priestly (1999: 96). 21  Carpenter (2015: 23f) thinks the “belonging together” clause is not just a matter of spatiotemporal proximity, a feature shared by all aggregates that belong together, but something having to do with their functionality and development, even though in the end she admits that functionality alone is not sufficient to distinguish persons from chariots. 22  This is the strategy adopted by Harvey (1995). 23  Collins (1982) offers precisely such an analysis. 24  This position, first articulated by Albahari (2006), is developed at length by Ganeri (2017). 20

54

C. Coseru

Building on the Abhidharma analytic project, Dignāga, not unlike Brentano, advances a dual-aspect theory of mind, which takes the subjective or qualitative aspect (svābhāsa) and the object-oriented or intentional aspect (viṣayābhasa) to be a constitutive features of the structure of cognitive awareness.25 In large measure, this theory rests on three distinct claims: (1) that we are directly aware of each occurrent mental state; (2) that each occurrent mental state has a dual aspect: it has both subjective and objective content; and (3) that each occurrent mental state is also reflexively self-conscious. The first claim goes against the view that occurrent mental states are ultimately impersonal or anonymous. To hold such a view would be akin to claiming that experience lacks a distinctly subjective character or possess such character only by virtue of distortions embedded in the structure of ‘I-making’ and ‘self-grasping’ tendencies.26 The second claim identifies subjectivity and intentionality as distinct structural features rather than discrete elements of the mental domain. The subjective aspect is constitutive of an implicit openness to what is given, while the objective aspect captures what the mental state is about: an object or mental content of some kind. Finally, the third claim is intended to capture the mode of presentation of all conscious cognitive states. Incidentally, the first and third claims can also be read as making the case that effortful self-knowledge––of the sort gained through introspection, intersubjective reports, or reflective attitudes––depends on tacit or non-propositional modes of acquaintance (Cassam 1995). This understanding of the structure of conscious experience is not unlike that put forward by those who argue that our subjectivity is immersive rather than egological. Consider, for instance, this statement from Zahavi, who takes Heidegger’s lead on this issue: “I am acquainted with myself when I am captured and captivated by the world. Self-acquaintance is indeed only to be found in our immersion in the world, that is, self-acquaintance is always the self-acquaintance of a world-immersed self” (Zahavi 2005: 82).27 On this view, self-acquaintance is not something that occurs apart from our immersion in the world and its complex set of intentional and intersubjective relations. Reflecting on the mode of presentation of our mental states, P.  F. Strawson likewise observes, “our desires and preferences are not, in general, something we just note in ourselves as alien presences. To a large extent they are we” (Strawson 1992: 134). To claim, thus, that access to our mental lives is always mediated in some  See, e.g., Compendium of Epistemology (Pramāṇasamuccaya) 1.10, in Hattori (1968: 29). See also Coseru (2020) for a discussion of Dignāga’s theory relative to the problem of self-knowledge. 26  This view is attributed to Asaṅga. As he observes in the Compendium of the Great Vehicle: “How does one know that mind (manas) in the sense of ‘afflicted mind’ (kliṣṭa-manas) exists? Without it, there would be no uncompounded ignorance, that is, a basic ignorance not yet associated with all the faults (doṣa), but serving as their base (āśraya). Indeed, introspective awareness (manovijñāna) must also have a simultaneous basis, as do the five types of empirical consciousness whose support are their material organs. Such a simultaneous support can only be found in the ‘afflicted mind.’” (Mahāyāna-samgraha 1.6–7, in Lamotte 1938: 36). 27  For a paradigmatic account of this world-immersed conception of selfhood, drawing on the work of Kierkegaard, see Sheridan Hough’s (2015) compelling existential-phenomenological investigation of the “knight of faith”, Kierkegaard’s example of a fully realized human being who finds non-thematic meaningfulness in their ongoing engagement with the world. 25

Reasons and Conscious Persons

55

fashion or another, perhaps by participation in a shared domain of language and reflection––is to ignore these essential features of self-consciousness. Dignāga’s account of reflexive self-consciousness (svasaṃvitti) does raise some legitimate concerns, for instance, about the absurd consequence of admitting that all cognitions, by virtue of being self-presenting, are epistemically warranted. Yet the main issue, as I have argued at length elsewhere (Coseru 2012, Chapter 8), is specifying the criteria under which occurrent mental states becomes epistemically salient. A reductive causal account that treats consciousness as emerging from elements that themselves lack phenomenal properties, is necessarily incomplete if it does not explain how those elements acquire the phenomenal properties that they do when grouped together. Of course, not all groupings of microphysical elements need necessarily exhibit a reflexive structure.28 But those that do call for an explanation that considers the whole person and not just its constitutive elements. In putting forward an account of consciousness as self-intimating, Dignāga is thus concerned to explain how a conscious mental episode, which is irreducible, can become the vehicle of self-knowledge. In that sense, he shares a common ground with, inter alia, Aristotle, Descartes, Locke, Brentano, Husserl, Sartre, and many contemporary defenders of the view that the nature of consciousness is such that all conscious mental states are non-representationally or pre-reflectively aware of themselves.29 The picture they present corresponds to what some have termed non-reductive “one-level theories of consciousness”: that is, theories which propose that consciousness is essentially a matter of having or being an awareness of a world that does not require a prior (representational) awareness of our own mental states (Thomasson 2008). Reflexivism or the thesis that self-consciousness consists in mental states being implicitly conscious of their occurrence thus serves as a grounding principle, enabling the intentional and subjective aspects of experience to emerge co-­ constitutively in each instance of cognitive awareness. Can such reflexive self-consciousness provide enough metaphysical grounding for thoughts of the sort ‘I am in pain’? Do such self-ascriptions require independent criteria (or the normative, indexical, or ownership kind) for individuating streams of conscious episodes to manifest as occurring for me? Can conscious mental states occur without any sense of whose states they are? Can there be consciousness without self-consciousness?

 Micropsychism, roughly the view that human consciousness derives from a combination of micro-conscious entities, might be taken as the exception here (see, e.g., Strawson 2006). But micropsychism faces the well-known combination problem (Seager 1995): how exactly do microlevel entities with their own very basic forms of consciousness combine together to form human consciousness? In response to the combination problem, some defenders of panpsychism reject the atomistic metaphysics of micropsychism in favor of cosmopsychism, which holds that individual human consciousness is derived from, or an expression of, a single unified field of consciousness (Jaskolla and Buck 2012; Shani 2015, 2018; Nagasawa and Wager 2016; Goff 2017). Cosmopsychism faces the opposite problem, namely the decombination problem: how do we get from a unified field of consciousness to individual human and animal consciousness? 29  See Strawson (2015) for a defense of the self-intimation thesis that draws together these major figures. 28

56

C. Coseru

If reflexive self-consciousness has a distinctive character, a specific givenness or for-me-ness, then misascriptions (as in the case of thought insertion or various pathologies of the self, such as ego dissolution) would be unintelligible if thought were transparent with regard to its occurrent for-me-ness (Bermudez 1998; Zahavi and Kriegel 2015). Reflexivism, then, is simply a statement about the self-­intimating character of conscious mental states. It is the condition for the possibility of self-­ knowledge, rather than the achievement of self-knowledge. Because agency entails awareness of action and its consequences, persons (as self-conscious agents) are ultimately real in a way that chariots and forests are not. Reflexivism allows for a minimal conception of subjectivity, specifically one that works against the idea that there could be experiences that are generic or impersonal until they are attended to in reflective or introspective thought. As such, it addresses not only the problem of self-consciousness, but also the problem of subjectivity, of why it is that experiences present themselves as being not only about something, but for someone. It is only to the extent that experiences exhibit what has been called the ‘dative of manifestation’––the fact that every experience is necessarily an experience for (or given to) someone (Prufer 1975)––that we are in a position to understand what it is that makes us persons. We may conclude that what makes the further fact view of there being persons plausible is the structure of consciousness itself. This way of understanding persons is not unlike Locke’s view of persons as self-conscious rational subjects. Neither a biological entity, nor an immaterial soul, a person is what it is by virtue of the fact that it, …can consider it self as it self, the same thinking thing, in different times and places; which it does only by that consciousness, which is inseparable from thinking, and, as it seems to me essential to it: It being impossible for any one to perceive, without perceiving, that he does perceive. When we see, hear, smell, taste, feel, meditate, or will any thing, we know that we do so. ... And as far as this consciousness can be extended backwards to any past Action or Thought, so far reaches the identity of that Person (Locke 1689 2.27.9, 335)

The Buddhist Personalist may eschew the requirement that the self-conscious subjectivity in play be manifest only in thought as rational deliberation and action, but agree that without self-consciousness it is hard to understand how self-concern, which is essential for moral cultivation, would be possible. For in looking to reward or punish persons for their deeds we do not look for a particular body, mental state or immaterial substance of some sort, but rather “to connections between the consciousness of the one who did the deed and the one who is being held accountable” (Ainslie and Ware 2014: 248).30 And Buddhist Personalists may also agree with Hume and the Parfitian version of the Reductionist View that our perceptions or sensible intuitions reveal no such thing as a fixed or stable self. Awareness of the inner contents of our mind lacks a necessary relation to a unified self. But while it  As Ainslie and Ware make quite clear, Locke not only defends the view, typically associated with Sartre, that consciousness is reflexive (that is, consciousness entails self-consciousness), but also (and more controversially) that consciousness is not “limited simply to our occurrent mental states” but rather extends its contents from one conscious moment to the next (2014: 249). 30

Reasons and Conscious Persons

57

may be true that we do not need to be self-conscious in a higher order or metacognitive sense in order to perceive and think, the operations of perception and thought do presuppose that conscious experiences are unified in such a way that their contents can be synthesized. And it is to the unity of conscious experience, achieved in the constitutive operations of the aggregates, that Buddhist Personalists point to when they claim that persons are ultimately real.

5 Conclusion Parfit’s motivation for rejecting the Non-Reductionist View concerns the claim that there is more to being a person than what accounts of physical and psychological continuity can satisfactorily provide. The Non-Reductionist makes the ultimately unsupported claim that “the unity of consciousness at any time is explained by the fact that several experiences are being had by a person” (1984: 275). Split brain cases and teleportation thought experiments, however, make it hard to explain unity by insisting on the importance of maintaining personal identity over time across different streams of consciousness. In conceiving of unity in terms of the ascription of different experiences to a particular subject, we are confronted with the unwarranted assumption, namely that “there can be, in a person’s life, subjects of experiences that are not persons” (ibid.). Such cases are better explained, argues Parfit, by appealing to the Reductionist Psychological Criterion, which claims that “at any time, there is one state of awareness of the experiences in one stream of consciousness, and another state of awareness of the experiences in the other stream” (ibid.) In developing the criteria for psychological reductionism, Parfit’s theoretical perspective points in the direction of a formal but variant structure of awareness. This structure ought to, at least in principle, be able to accommodate continuity without the sameness or identity that ascribing such states to something like a Cartesian Ego would entail. It is unlikely that Buddha and most Buddhist Reductionists would have agreed with this part of his theory. Although it is in keeping with the principles of the no-self view, Parfit’s insistence on an external criterion for individuating or indexing awareness to a particular stream of consciousness may be at odds with what I have argued are certain salient and ineliminable features of phenomenal consciousness, specifically its self-reflexivity. If self-reflexive conscious episodes lack an intentional structure, then they cannot provide the minimal sense of internal distance necessary for subjectivity. On the dual aspect model of consciousness sketched above (see also Coseru 2015: 66, 79), mental streams are differentiated by being covariant with intentional behavior, which presupposes that subjectivity and intentionality are structural features of consciousness, rather than constructs or relations among discrete mind moments (as presupposed by mereological nihilists). This essay has considered whether Buddhist Reductionism provides support for Parfit’s psychological criterion of personal identity given considerations about the seemingly irreducible character of phenomenal consciousness. Taking Buddhist Personalism as a point of reference, I have argued that reductionism about selves

58

C. Coseru

can be reconciled with the seemingly irreducible character of self-consciousness. Furthermore, Buddhist philosophers who follow in the footsteps of Dignāga and his successors provide a viable, if not altogether unproblematic, model for how this reconciliation might proceed. Whether or not this epistemological project is constrained by ontological assumptions about the constitutive elements of persons is open to debate. What is less controversial is that in rejecting a permanent locus for experience, the Buddha created an opportunity for the problem of personal identity to be explored not only on metaphysical but on empirical and phenomenological grounds as well. What sets apart non-reductive theories of persons as self-conscious subjects from their alternatives, then, is the notion that personal identity is not really a matter of the persistence or continuity of enough of the constitutive (physical and psychological) elements of persons, as demanded by the psychological criterion. Rather, on the model put forward here, it is a matter of understanding that the principle of individuation is neither external nor transcendent but internal to the structure of consciousness and its operations.

References Ainslie, D.C., and O.  Ware. 2014. Consciousness and Personal Identity. In The Routledge Companion to Eighteenth Century Philosophy, ed. A. Garrett, 245–264. Routledge. Albahari, M. 2006. Analytical Buddhism: The Two-Tiered Illusion of Self. Hampshire: Palgrave Macmillan. Baker, L.R. 2013. Personal Identity: A Not-So-Simple View. In Personal Identity: Complex or Simple, ed. G. Gasser and M. Stefan, 179–191. Cambridge: Cambridge University Press. Bastow, D. 1986. Self-construction in Buddhism. Ratio 28 (2): 97–113. Basu, A. 1997. Reducing Concern with Self: Parfit and the Ancient Buddhist Schools. In Culture and Self: Philosophical and Religious Perspectives, East and West, ed. D.  Allen, 97–109. Boulder: Westview Press. Batterman, R.W. 1995. Theories Between Theories: Asymptotic Limiting Intertheoretic Relations. Synthese 103: 171–201. Bermudez, J.L. 1998. The Paradox of Self-Consciousness. Cambridge, MA: MIT Press. Berry, M.W. 2001. Chaos and the Semiclassical Limit of Quantum Mechanics (Is the Moon There When Somebody Looks?). In Quantum Mechanics: Scientific Perspectives on Divine Action, ed. R.J. Russell, P. Clayton, K. Wegter-McNelly, and J. Polkinghorne, 41–54. Vatican Observatory CTNS Publications. Carnap, R. 1934. The Unity of Science. London: Kegan Paul, Trench, Trubner, and Co. Carpenter, A. 2015. Persons Keeping Their Karma Together: The Reasons for the Pudgalavāda in Early Buddhism. In The Moon Points Back, ed. K.  Tanaka, Y.  Deguchi, J.  Garfield, and G. Priest, 1–44. New York: Oxford University Press. Caruso, G.D. 2012. Free Will and Consciousness: A Determinist Account of the Illusion of Free Will. Lanham: Lexington Books. Cassam, Q. 1995. Introspection and Bodily Self-Ascription. In The Body and the Self, ed. J.L. Bermudez, A. Marcel, and N. Eilan. Cambridge, MA: MIT Press. Châu, Thich Thiên. 1999. The Literature of the Personalists of Early Buddhism. Trans. S. Boin-­ Webb. Delhi: Motilal Banarsidass. Collins, S. 1982. Selfless Persons: Imagery and Thought in Theravāda Buddhism. Cambridge: Cambridge University Press.

Reasons and Conscious Persons

59

———. 1985. Buddhism in recent British Philosophy and Theology. Religious Studies 21: 475–493. ———. 1997. A Buddhist Debate About the Self; and Remarks on Buddhism in the Work of Derek Parfit and Galen Strawson. Journal of Indian Philosophy 25 (5): 467–493. Coseru, C. 2012. Perceiving Reality: Consciousness, Intentionality, and Cognition in Buddhist Philosophy. New York: Oxford University Press. ———. 2015. Perception, Causally Efficacious Particulars, and the Range of Phenomenal Consciousness. Journal of Consciousness Studies 22 (9–10): 55–82. ———. 2016. Freedom from Responsibility: Agent-Neutral Consequentialism and the Bodhisattva Ideal. In Buddhist Perspectives on Free Will: Agentless Agency? ed. R.  Repetti, 92–105. New York: Routledge. ———. 2020. Whose Consciousness? Reflexivity and the Problem of Self-Knowledge. In Buddhist Philosophy of Consciousness: Tradition and Dialogue, eds. M. Siderits, C. Keng, and J. Spackman, Leiden: Brill. Crane, T. 2010. Cosmic Hermeneutics vs Emergence: The Challenge of the Explanatory Gap. In Emergence in Mind, ed. G.  Macdonald and C.  Macdonald, 22–34. Oxford: Oxford University Press. de la Vallee Poussin, L. 1980. L’Abhidharmakośa de Vasubandhu. Vol. 1. Bruxelles: Institut Belge des Hautes Études Chinoises. Duerlinger, J. 1993. Reductionist and Nonreductionist Theories of Persons in Indian Buddhist Philosophy. Journal of Indian Philosophy 21 (1): 79–101. ———. 2003. Indian Buddhist Theories of Persons: Vasuabndhu’s “Refutation of the Theory of a Self”. London: Routledge Curzon. Flanagan, O. 2002. The Problem of the Soul: Two Visions of Mind and How to Reconcile Them. New York: Basic Books. Ganeri, J. 2007. The Concealed Art of the Soul: Theories of Self and Practices of Truth in Indian Ethics and Epistemology. Oxford: Oxford University Press. ———. 2017. Attention, Not Self. Oxford: Oxford University Press. Garrett, B. 1998. Personal Identity and Self-Consciousness. London: Routledge. Giles, J. 1993. The No-self Theory: Hume, Buddhism, and Personal Identity. Philosophy East and West 43 (2): 175–200. Goff. 2017. Consciousness and Fundamental Reality. Oxford: Oxford University Press. Hanner, O. 2018. Buddhism as Reductionism: Personal Identity and Ethics in Parfitian Readings of Buddhist Philosophy; from Steven Collins to the Present. Sophia 57: 211–231. Harvey, P. 1995. The Selfless Mind: Personality, Consciousness and Nirvāṇa in early Buddhism. Richmond: Curzon Press. Hattori, M. 1968. Dignāga, On Perception. Cambridge, MA: Harvard University Press. Horner, I.B. 1963. Milinda’s Questions. Vol. I. London: Luzac & Company. Hough, S. 2015. Kierkegaard’s Dancing Tax Collector: Faith, Finitude, and Silence. Oxford: Oxford University Press. Jaskolla, L.J., and A.J.  Buck. 2012. Does Panexperientialism Solve the Combination Problem. Journal of Consciousness Studies 19 (9–10): 190–199. Kistler, M. 2010. Strong Emergence and Freedom––Comment on A. Stephan. In Emergence in Mind, ed. G. Macdonald and C. Macdonald, 240–251. Oxford: Oxford University Press. Lamotte, E. 1938. La Somme du Grand Véhicule d’Asaṅga (Mahāyānasaṃgraha). Louvain-la-­ Neuve: Université de Louvain. Locke, J. 1689/1975. In An Essay concerning Human Understanding, ed. P.  Nidditch. Oxford: Oxford University Press. Lowe, E.J. 2009. More kinds of being: A further study of individuation, identity and the logic of sortal terms. Malden (MA), Blackwell. ———. 2013. The Probable Simplicity of Personal Identity. In Personal Identity: Complex or Simple, ed. G. Gasser and M. Stefan, 137–155. Cambridge: Cambridge University Press. Merricks, T. 2001. Objects and Persons. Oxford: Clarendon Press.

60

C. Coseru

Meyers, K. 2014. Free Persons, Empty Selves: Freedom and Agency in Light of the Two Truths. In Free Will, Agency, and Selfhood in Indian Philosophy, ed. M. Dasti and E.F. Bryant, 41–67. New York: Oxford University Press. Nagasawa, Y., and K. Wager. 2016. Panpsychism and Priority Cosmopsychism. In Panpsychism, ed. G. Brüntrup and L. Jaskolla, 113–129. New York: Oxford University Press. Nahmias, E., S. Morris, T. Nadelhoffer, and J. Turner. 2004. The Phenomenology of Free Will. Journal of Consciousness Studies 11 (7–8): 162–179. Neurath, O. 1932/1983. Sociology in the Framework of Physicalism. In Neurath, Philosophical Papers 1913–1946, ed. R.S. Cohen and M. Neurath, 58–90. Dordrecht: Reidel. Nida-Rümelin, M. 2013. The Non-descriptive Individual Nature of Conscious Beings. In Personal Identity: Complex or Simple, ed. G. Gasser and M. Stefan, 157–176. Cambridge: Cambridge University Press. Noordhof, P. 2010. Emergent Causation and Property Causation. In Emergence in Mind, ed. G. Macdonald and C. Macdonald, 69–98. Oxford: Oxford University Press. Parfit, D. 1971. Personal Identity. The Philosophical Review 80 (1): 3–27. ———. 1984/87. Reasons and Persons. Oxford: Clarendon Press. ———. 1999. Experiences, Subjects, and Conceptual Schemes. Philosophical Topics 26 (1–2): 217–270. ———. 2002. Reductionism and Personal Identity. In Philosophy of Mind: Contemporary Readings, ed. David J. Chalmers, 655–661. Oxford University Press. Perrett, R. 2002. Personal Identity, Minimalism, and Madhyamaka. Philosophy East and West 52 (3): 373–385. Priestley, L. 1999. Pudgalavāda Buddhism: The Reality of the Indeterminate Self. Toronto: Centre for South Asian Studies, University of Toronto. Prufer, T. 1975. An Outline of Some Husserlian Distinctions and Strategies, Especially in the Crisis. Phänomenologische Forschungen 1: 189–204. Sauchelli, A. 2016. Buddhist Reductionism, Fictionalism About the Self, and Buddhist Fictionalism. Philosophy East and West 66 (4): 1273–1291. Seager, W.E. 1995. Consciousness, information, and panpsychism. Journal of Consciousness Studies 2: 272–288. Shani, I. 2015. Cosmopsychism: A Holistic Approach to the Metaphysics of Experience. Philosophical Papers 44 (3): 389–417. ———. 2018. Beyond Combination: How Cosmic Consciousness Grounds Ordinary Experience. Journal of the American Philosophical Association 4: 390–410. Shoemaker, S. 1970. Persons and Their Pasts. American Philosophical Quarterly 7 (4): 269–285. Sider, T. 2003. What’s so Bad about Overdetermination? Philosophy and Phenomenological Research 67 (3): 719–726. Siderits, M. 1987. Beyond Compatibilism: A Buddhist Approach to Freedom and Determinism. American Philosophical Quarterly 24 (2): 149–159. ———. 1997. Buddhist Reductionism. Philosophy East and West 47 (4): 455–478. ———. 2008. Paleo-compatibilism and Buddhist Reductionism. Sophia 47: 29–42. ———. 2015. Personal Identity and Buddhist Philosophy: Empty Persons. 2nd ed. Aldershot: Ashgate. Smart, J.J.C. 2006. Metaphysical Illusions. Australasian Journal of Philosophy 84 (2): 167–175. Stcherbatsky, T. 1919. The Soul Theory of the Buddhists. Bulletin de l’Academie des Sciences de Russie 13: 823–854; 937–958. Stone, J. 1988. Parfit and the Buddha: Why There Are No People. Philosophy and Phenomenological Research 48: 519–532. ———. 2005. Why There Still Are No People. Philosophy and Phenomenological Research 70: 174–192. Strawson, P.F. 1992. Freedom and Necessity. In Analysis and Metaphysics. Oxford: Oxford University Press.

Reasons and Conscious Persons

61

Strawson, G. 2006. Realistic Monism: Why Physicalism Entails Panpsychism. Journal of Consciousness Studies 13 (10–11): 3–31. ———. 2015. Self-intimation. Phenomenology and the Cognitive Sciences 14: 1–31. Swinburne, R. 2013. How to Determine Which is the True Theory of Personal Identity. In Personal Identity: Complex or Simple, ed. G. Gasser and M. Stefan, 105–122. Cambridge: Cambridge University Press. Thomasson, A.L. 2007. Ordinary Objects. Oxford: Oxford University Press. Thomasson, A. 2008. Phenomenal Consciousness and the Phenomenal World. The Monist 91 (2): 191–214. Venkataramanan, K. 1953. Sāṃmitīyanikāya Śāstra. Visva-Bharati Annals 5: 155–243. Wegner, D. 2002. The Illusion of Conscious Will. Cambridge, MA: MIT Press. Zahavi, D. 2005. Subjectivity and Selfhood: Investigating the Fist-Person Perspective, 2005. Cambridge, MA: MIT Press. Zahavi, D., and U. Kriegel. 2015. For-Me-Ness: What It Is and What It Is Not. In Philosophy of Mind and Phenomenology, ed. D.  Dahlstrom, A.  Elpidorou, and W.  Hopp, 36–53. London: Routledge.

Descartes and the Buddha—a rapprochement? Galen Strawson

1 Introduction: Substance Descartes’s conception of the mind is nothing like what most people suppose. I believe it may have interesting affinities with certain Asian—even Buddhist!—conceptions of the mind. I’m not qualified to comment on the Asian side, so I’m going to describe what I take to be his position and invite others to judge. Perhaps the only thing one can be sure of, when offering intertraditional comparisons of this kind, is that both sides will feel affronted. So it goes. This piece will succeed in its purpose if it reduces even by one the number of routine mistaken invocations of Descartes. I’ve published various versions of this view over the years, and draw on them here.1 As far as I know it has been met with perfect silence, and it is in the spirit of a Zen exercise that I offer another variant. I conclude with a rebuttal of Lichtenberg’s famous but feeble criticism of Descartes. My central point can be made without paying much attention to what is known as Descartes’s ‘substance dualism’—his official view that there are two

 See e.g. Strawson (1994: ch. 5, 2011: §§7.3–7.4; the previous versions are seriously defective because they don’t take sufficient account of the role of the brain in Descartes’s theory). When I cite a work, I give the date of first publication or composition; the page or section reference is to the edition listed in the bibliography. In the case of quotations from languages other than English I cite a standard translation but don’t always give it. I use bold italics to mark an author’s emphasis and italics to mark my own. References of the form ‘AT2.570’ are to the Adam and Tannery edition of Descartes (in this case volume 2, page 570); references of the form ‘C3.137’ are to the Cottingham et al. translations of Descartes (in this case volume 3, page 137). 1

G. Strawson (*) The University of Texas at Austin, Department of Philosophy, The College of Liberal Arts, Austin, TX, USA © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_4

63

64

G. Strawson

fundamentally different kinds of concrete being, mental and material.2 Two remarks, though, before I put it aside. First, I say ‘his official view’ because there’s good reason to think that Descartes knew that he couldn’t actually rule out a materialist monist view of the kind he discussed with Regius.3 Second, the term ‘substance dualism’ is already highly questionable insofar as it comes with a standard picture of Descartes as Mr. Substance. This is hugely misleading: Descartes regularly stressed the point, usually attributed to Hume (or at least Locke), that we have no positive conception of substance whatever. When Mark Siderits writes, in his lovely book Buddhism as Philosophy, that for Abidharma ‘the substances are useless cogs in the machine, they explain nothing’ (2007: 114), he could be glossing Descartes. Desmond Clarke has an excellent discussion of this question in his book Descartes’s Theory of Mind, which I am going to cite frequently. He shows that the notion of substance operates for Descartes as a kind of ‘dummy’ concept, invoked only insofar as Descartes goes along with—without in any way clarifying—the philosophically orthodox idea that all qualities must in order to exist somehow ‘be in’ (inesse) something ‘as in a subject’.4 So much, very briefly, for Descartes’s ‘substance dualism’. Let me add a remark about his religion. I’m with him when he says that ‘mingling religion with philosophy … goes quite against the grain with me’ (AT2.570/C3.137). That said, I obviously won’t bracket anything in his metaphysics that might be thought to flow from his religious commitments, for example his unargued commitment to the existence of an irreducible plurality of individual immaterial souls, which sits so strikingly alongside his view that the material universe is in some fundamental sense a single thing.

2 Duration and God Some brief notes, next, about duration, God, and the meaning of cogitatio. Descartes holds that a lifespan can be divided into innumerably many parts, each completely independent of the others, so that it does not follow from the fact that I existed a little while ago that I must exist now unless there is some cause that as it were creates me all over again at this moment, i.e. which conserves me … the same power and action are needed to conserve anything in

 Here ‘concrete’ simply contrasts with ‘abstract’: some people thing that numbers and concepts are examples of abstract being. 3  See Regius (1647: 294–295). In his 1648 reply to Regius, Descartes reiterates his official—and religiously orthodox—position. See Clarke (2003: ch. 9). 4  See §6 below for examples of Descartes’s direct repudiation of the orthodox terminology. It’s important to remember that he is anxious not to offend philosophical and religious orthodoxy, and to use standard terminology as far as possible (see p. 000 below). For a discussion of the issue as it arises in Descartes’s intense exchange with Hobbes, see Strawson (forthcoming). 2

Descartes and the Buddha—a rapprochement?

65

existence at each individual moment of its duration as would be required to create that thing de novo if it were not yet in existence (AT7.49/C2.33, see also AT7.109/C2.78–9) …. Thus … conservation and creation only differ relative to our way of thinking, and not at all in reality (AT9.39)

This invites comparison with the Buddhist doctrine of momentariness, according to which the existence of the universe consists of a series of distinct flashes of being: as an entity vanishes, it gives rise to a new entity of almost the same nature which originates immediately afterwards. Thus there is an uninterrupted flow of causally connected momentary entities of the same kind, the so-called santāna … the world … is at every moment distinct from the world in the previous or next moment, … linked to the past and future by the law of causality.5

Plainly any such view raises a question about how the flash of being that constitutes the existence of the universe at one moment accords as it does with its predecessor and successor. Descartes has God to do the job, recreating the universe at every instant in a way that seems to take the causality out of the entities themselves. That said, he also writes that ‘I understand nature, generally considered, as nothing other than God himself, or, if you will, the ordered system of created things set up by God’.6 I simply want to note this. When it comes to the question of duration, Descartes doesn’t have some heavy and supposedly paradigmatically ‘Western’ conception of what substance (or a substance) is that puts him at odd with more explicitly processual views. My understanding is that Buddhist momentarists don’t hold that there are literally innumerable moments—an idea which follows, for Descartes, from his assumption of the literally infinite divisibility of temporal duration. Even so the similarity in the matter of the flash theory of being is worth noting. It provides a useful context for Russell’s arch and misguided ‘Humean’ complaint: What [Descartes] was to himself is not best described as a single entity with changing states. The single entity is quite otiose. The changing states suffice. Descartes to himself should have appeared as a series of events, each of which might be called a thought, provided that word is liberally interpreted. It was this series of ‘thoughts’ which constituted Descartes’s ‘mind’.7

It’s also true that Descartes standardly talks as if minds are genuine continuants, things which have, as persisting individuals, a moral destiny. This may be thought to be warranted by the fact that there is on his view no real difference between conservation and creation. For if conservation—persistence—in being is really (‘just’) continual recreation, so too continual recreation is a species of genuine conservation in being—a kind of genuine persistence. We needn’t make much of this, however,

 Rospatt (1995: 1). According to one Buddhist estimate this happens 16,000,000 times a second. Some physicists hold that time itself is quantal, and we may perhaps look in the region of the Planck time—about 10−43 of a second—when looking for a measure. 6  CSM2.56, AT7.80 (per naturam enim, generaliter spectatam, nihil nunc aliud quam vel Deum, ipsum, vel rerum creatarum coordinationem a Deo institutam intelligo). 7  1950: 148. See also the discussion of Lichtenberg in §12 below. 5

66

G. Strawson

when speculating about Asian (in particular Buddhist) parallels in Descartes, because the idea that entities like ourselves have some kind of persistence, quite irrespective of their temporal flashiness, seems to be matched in Buddhist thought by the essentially diachronic idea of karma. From now on, I’ll use conventional language that suggests persistence—words like ‘stream’, ‘process’, ‘mind’—on the understanding that one can convert it into strict Cartesio-Buddhist flash-theory terms as required. Whatever makes it the case that individual entities like ourselves can be supposed to have a diachronic moral destiny finds some echo in the fact that it is appropriate to talk of an individual karma trail across time.

3  Cogitatio = Consciousness It’s well known that Descartes standardly uses the word ‘thought’ or ‘thinking’ (cogitatio, pensée) in an extremely broad way to cover all conscious occurrences whatever, including all sensing or feeling, willing, and imagining: Thinking. I use this term to include everything that exists within us in such a way that we are immediately conscious (conscius) of it. Thus all the operations of the will, the intellect, the imagination, and the sense are thoughts (AT7.160/C2.113) …. Understanding, willing, imagining, having sensory perceptions … all fall under the common concept of thinking or perception or consciousness (AT7.176/C2.124)

For this reason, I’ll regularly use ‘consciousness’ and ‘experience’ interchangeably, instead of ‘thought’ or ‘thinking’, in presenting Descartes’s views. Since ‘consciousness’ and ‘experience’ are themselves philosophically vexed terms, let me say that I’ll use them only in their very broadest sense to mean actual, live, occurrent conscious experience of any sort whatever; actual occurrent experiential ‘what-it-is-­ likeness’, ‘phenomenology’, feeling, sentience, qualial consciousness of any sort, however primitive, however sophisticated (this seems to be what most analytic philosophers mean by ‘consciousness’ today). On the Buddhist side I’ll take it that ‘consciousness’ and ‘experience’ cover all (occurrent) elements or dharmas that are instances of the four ‘mental’ or nāma skandhas. The best translation of ‘cogito ergo sum’, given this use, is ‘I’m conscious therefore I exist’, or ‘I’m experiencing therefore I exist’. It’s certainly not ‘I think therefore I exist’ or ‘I’m thinking therefore I exist’—not if ‘think’ is understood to have any sort of distinctively cognitive implication.8

 Descartes’s broad use of ‘thought’ and ‘thinking’ to mean any kind of conscious experience survives in Western philosophy well into the twentieth century. It’s only relatively recently (perhaps in the last sixty years) that this use has been misunderstood to mean something essentially intellectual or cognitive. (Note, though, that Descartes does on occasion use ‘think’ in a more narrowly cognitive sense closer to our present-day use. See Rozemond (1998); Strawson (forthcoming).) 8

Descartes and the Buddha—a rapprochement?

67

4  C is a Consciousness-Process So far, perhaps, so good. Now for the central question. What is a Cartesian mind— call it C? C, I propose, is quite unlike a Cartesian mind as usually imagined—call it C*. C* is some sort of unit of immaterial stuff that [i] is a vehicle or bearer of states of consciousness [ii] has (therefore) some nature or mode of being other than consciousness [iii] can continue to exist even when there is no consciousness. C, by contrast, is just a stream of consciousness, a consciousness-process. It is not [i] a vehicle or bearer of states of consciousness, it does not [ii] have any nature or mode of being other than consciousness and it cannot [iii] continue to exist even when there is no consciousness, i.e. in a state of unconsciousness: [1] C = thinking = consciousness, consciousness-process. That’s all there is to C, ontologically speaking. The stream of consciousness doesn’t happen or ‘live’ in something else.9 Descartes holds the same view as the fifth-century thinker Mamertus. Hobbes saw this quite clearly (see Strawson forthcoming), and Priestley took it for granted when, discussing Mamertus, he found in some of his expressions the peculiar opinions of Descartes. For, he says, the soul is not different from the thoughts, that the soul is never without thought, because it is all thought [i.e. conscious process] (Priestley 1782: 250)

Many have wondered why Descartes held that the mind is always conscious—that it can’t possibly exist (in dreamless sleep, for example) in a state of unconsciousness. Here is the explanation! The mind can’t possibly exist in a state of unconsciousness because it’s wholly constituted of consciousness. If consciousness stopped it would cease to exist. There are no doubt difficulties in this view. That doesn’t change the fact that it is Descartes’s view. At one point in the Principles he treats the terms ‘mind/soul’ and ‘thinking’ as strictly interchangeable, writing of ‘what we call our soul or our thinking’ (AT9B.10/C1.184, ‘ce que nous appellons notre âme ou notre pensée’). Later in the Principles he writes, unequivocally, that consciousness … must be considered as nothing else than conscious substance itself …, that is, as mind (AT8.30–1/CSM1.215).10

 Who came up with C*? I’m not sure. I think that Spinoza, Leibniz, Berkeley and Hume all reject it. Among the leading Western early modern philosophers only Locke appears to make use of it (and his own deep inclinations are—I suspect—materialist). 10  non aliter … cogitatio concipi debet, quam ipsa substantia cogitans …, hoc est, quam mens (I’ve edited out the strictly parallel claim about extension, and so changed the plural verb ‘debent’ to the singular ‘debet’). 9

68

G. Strawson

It may seem surprising that this fact about Descartes is not more generally known. In one way it is surprising, in another way not. In its more popular versions, the small area of the history of philosophy that I know something about contains some astonishing—and seemingly ineradicable—misconceptions (Locke on personal identity, Hume on causation, Hume on personal identity). Kant says that many historians of philosophy, with all their intended praise, … attribute mere nonsense … to past philosophers. They are incapable of recognizing, beyond what the philosophers actually said, what they really meant to say. (1790: 160)

He underestimates how often historians don’t even take in what past philosophers actually say.

5 Substance and Attribute Objection. ‘Surely there is, on the one ontological hand, the substance, the Cartesian mind or soul, C, and on the other or at least further ontological hand, its attributes, and in particular what Descartes calls its ‘principal’ attribute, i.e. consciousness? Surely the substance is not literally identical with the attribute(s)?’ Like it or not, this is Descartes’s view. The substance, for him as for Spinoza, is literally identical with its attribute(s).11 Descartes is one of a considerable number (including Kant, Nietzsche, and Schlick, in addition to Spinoza, in the Western traditions, and no doubt many in the Eastern traditions) who grasp the deep and difficult metaphysical truth (such I believe it to be) that an object (anything that we pick out as an object) is not something that is in any manner ontologically over and above its total intrinsic qualitiedness.12 The quotations in the last section are unequivocal, although Descartes (anxious not to cause undue upset in a time when unorthodoxy could lead to death) continues for the most part to use the language of substance and attribute in the received manner. When Frans Burman in his famous conversation with Descartes explicitly asks him to confirm his apparent view that there is indeed no ‘real distinction’, no ontological distinction, between a substance and its attributes, Descartes does so: the attributes [of a substance], when considered collectively, are indeed identical with the substance. (AT5.155/1648: 15)

He speaks here of attributes in the plural simply because anything that exists concretely has, in his scheme of things, in addition to its ‘principal’ attribute (either  Spinoza speaks of “substances or, what is the same, their attributes, and their affections” (1677: IP4d). 12  It’s important not to call this view ‘the bundle theory of objects’, because the name brings with it the framework that the view repudiates, and this makes it seem problematic in a way that it isn’t. Briefly: the bundle is taken to be a bundle of qualities, and the notion of a quality that is deployed is understood as standing in essential—constitutive—contrast to a notion of an object that, given the contrast, can’t be just its qualitiedness. See e.g. Strawson (2021). 11

Descartes and the Buddha—a rapprochement?

69

consciousness or extension), the attributes of existence, duration, and number. But the only specific-character-constituting attribute of mind—the principal attribute— is: consciousness (thinking). Conscious process is in other words the whole stuff— the whole concrete being—of the mind. Existence, duration, and number are not themselves concrete, stuffy. When reading Descartes, then, one mustn’t fall back into thinking of conscious experience as somehow just a property of a something else that is ontologically distinct from it. This falling back is almost irresistible for us, but Descartes specifically rules it out—even while he continues for the most part to talk in the accepted way in order to not to get into too much trouble. A principal attribute—a concrete instantiation of a principal attribute13—is itself a res, a thing, according to Descartes. Certainly we naturally speak of the principal attribute of a thing, as if the attribute were in some way distinct from the thing. There is nevertheless a fundamental sense in which we speak imprecisely—albeit in a manner that our natural ways of thinking and talking force on us. For, again, the attributes of a thing (principal or characteristic attribute + attributes of existence, duration, number) are that thing, are identical with that thing. To that extent the phrase ‘attribute of a thing’, though natural and useful in conveying ideas to others, is misleading; although it still serves to distinguish the characteristic nature of the particular thing from its (mere) concrete existence.14

6 Substance and Attribute 2 Contrary quotations can be found. Shortly after confirming to Burman that the attributes of a substance are identical with the substance, Descartes speaks conventionally of the substance ‘underlying’ the attribute (AT5.156/1648: 17). His basic commitment is nevertheless clear. The distinction between the notion of an attribute like thinking/experiencing or extension ‘and the notion of substance itself is, he says, a merely conceptual distinction’ (AT8A.31/C1.215), like the distinction between triangularity and trilaterality. It is in other words a distinction that can be made in thought, a ‘distinction of reason’, not a ‘real’ distinction; where to say that there’s a real distinction between two things is simply to say that each can (or does) exist in reality without the other existing. Clarke is unequivocal: there is for Descartes no real distinction … between a thing and its properties … in the case of material things … it would certainly be a mistake to think of their substance as some kind of reality that is independent of their properties’. (2003: 215, 217)  Adding the attributes of existence, duration (existence in time), and number to the attribute consciousness gives it concrete existence. 14  One might say that the work done by ‘res’ in the phrases ‘res cogitans’ and ‘res extensa’ is simply the assertion of real (concrete) existence, while the work done by ‘cogitans’ or ‘extensa’ is identification of the type of existent. 13

70

G. Strawson

Nadler is no less clear: Descartes’s ‘considered position…is that while there is a conceptual distinction between substance and attribute … there is not a real distinction between them. Substance and attribute are in reality one and the same’ (2006: 57). Rozemond agrees: much of what [Descartes] says suggests that the principal attribute constitutes the entire substance and that there is no bare subject of inherence at all … the substance just consists in a principal attribute. I think that this is indeed [Descartes’s] view. But his commitment to this view is not entirely without problems. (1998: 11)15

Certainly, one can never when reading Descartes infer commitment to traditional views from use of traditional terminology: I pay no attention to the way in which certain words have been used in recent times in the schools … All I do is to notice what particular words mean in Latin, so that, when I lack appropriate words, I shall transfer to my own meaning whatever words seem most suitable. (AT10.369/C.1.14)

His philosophy is in part a covert operation: ‘I hope that readers’, lulled by the use of traditional terminology, ‘will gradually get used to my principles, and recognize their truth, before they notice that they destroy the principles of Aristotle’ (AT3.298/ C3.173). Shortly after his assertion of substance–attribute identity in response to Burman’s unavoidably direct question (AT5.155), Descartes talks again in conventional fashion of a substance underlying (substerno) its attribute (AT5.156). Cottingham (1976: 77–9) has a good discussion of this, but I don’t think he’s right to suggest that we should give priority to Descartes’s use of a word that accords with the conventional position over his earlier unequivocal (and ‘Aristotle-destroying’) assertion of the radical substance–attribute identity position. So Descartes entirely rejects C*, the Lockean picture of the immaterial mind or self that is usually assumed to be his view. For according to this picture, as already remarked, (a) there is some sort of immaterial mind-substance or mind-stuff X (b) X is the ‘ground’ or ‘bearer’ of conscious mental goings-on (c) X can continue to exist in the absence of any conscious mental goings-on from which it follows that (d) X has some nature other than conscious mental process. We can represent the difference between C and C* in a diagram and at the same time introduce C+, a third possible picture of an immaterial mind (Fig. 1). Everyone agrees that Descartes rejects (c), in holding that a mind or subject must always be thinking (= conscious), but his claim in the Principles that ‘each substance has one principal property which constitutes its nature or essence … and thinking (= consciousness) constitutes the nature of thinking (= conscious)

15

 See also Principles 1644: 1.62. I discuss some further objections in Strawson forthcoming.

Descartes and the Buddha—a rapprochement?

71

X continuously existing mind or soul represented by thick continuous line Y gappy process of thinking/consciousness represented by thin gappy line Z continuous stream of thinking/consciousness represented by thin continuous line Standard picture of immaterial soul: Y going on in ontologically distinct X [+a +b +c +d] C*

Possible picture of Descartes’s view: Z going on in (and essential property of) ontologically distinct X [+a +b +c +d] C+

Descartes’s view: Z = X [+a –b –c –d] C

Fig. 1  Three pictures of an immaterial mind

substance’ (AT8A.25/C1.210) is often read as if it allowed, as in C+, (d) that the mind has some other necessary manner of being that is not occurrent thinking = conscious experience.16 This reading is, however, extremely problematic; for to claim that something Y constitutes the nature of something X is to claim that nothing else does. It can also, of course, seem extremely problematic to reject (d). This is because we think of a mind as something that essentially has powers, powers that must (we may suppose) be in some way ontologically over and above the being of the conscious stream. At the very least, it seems that the stream must be somehow thicker than the stream of conscious contents one is experiencing at any given time. This is something we must address. For the moment, though, it seems clear that Descartes rejects all of (b)–(d), and allows (a) only in taking it (as remarked) that there is no real distinction between ‘the thing … we call …a “substance”’17 and its attributes. His picture is C. The non-material mind or self or soul or subject is: a consciousness-process. Descartes is thought of as clunky Mr Substance, as already

16 17

 It’s arguable that Berkeley endorses C+.  AT7.222/C2.156; my emphasis.

72

G. Strawson

remarked, but for him mental substance is just consciousness–process (and the process is flashy). So when Descartes famously states that the mind or soul can’t exist in the absence of consciousness—when he holds that consciousness is an essential property of mind in this sense, a property it can never lack—this isn’t any sort of extra stipulation on his part, a special and seemingly rather odd18 condition added to an already existing conception of the mind. The reason why a mind in which no thinking is going on is as impossible as a physical object without extension is simply, and again, that mind is consciousness; it’s wholly and literally constituted of occurrent conscious process. This immediately explains why Descartes holds that it can’t possibly exist without being conscious.

7 The Powers of the Mind There still seems to be a serious difficulty in this position—the difficulty, already noted, of finding a ground or ‘place of residence’, a manner of real existence, for mental faculties or capacities or powers like will and understanding, for innate ideas, and for what Descartes calls ‘intellectual memory’. Where can they possibly be lodged, given the rejection of (d)—given the idea that C, the thinking subject is, metaphysically, nothing other than conscious process? It’s true that Descartes doesn’t really believe in faculties as entities, holding that ‘the term “faculty” denotes nothing but a potency’,19 but potencies or powers also seem to need a place of residence (a manner of real existence) of a sort that seems hard—impossible—to supply if all one has to hand at any particular time—say t—is an individual human being’s current conscious process with the particular extremely limited (conscious) content that it has at t. The difficulty might be diminished if Descartes’s use of ‘thinking’ could be supposed to extend to non-conscious occurrent contentful mental goings­on. Then we might suppose C to be a great, deep, thick ongoing process—a great sinewy stream of non-conscious mental content as well as conscious content—that could in some manner carry all C’s capacities. But it seems that we can’t do this: according to Descartes all thinking is conscious, consciousness. And even if we could, it would remain obscure how the categorical being of a rich process of non-­ conscious occurrent mental goings-on could serve as the place of residence of C’s capacities. It may be that a similar problem arises about the being of the mental skandhas (see e.g. Siderits 2007: §3.2), given that the current set, which we may suppose to be heavily preoccupied with a cow, cause others which cause others (etc.) which … later include an understanding-French-involving skandha. For it seems that the

 For this opinion, see Locke (1689–1700: 2.1.10). Locke failed to grasp Descartes’s conception of the soul. 19  AT8B.361/C1.305. 18

Descartes and the Buddha—a rapprochement?

73

understanding-French-involving skandha must somehow inherit everything that underlies its possibility of being what it is from its causal predecessors. Can we mitigate the difficulty? The first thing to do, perhaps, is to try to forestall a possible misunderstanding of my use of the term ‘conscious process’ or ‘consciousness-­process’. Whatever consciousness-process C is, it’s not a mere streaming of passive content. It is rather something that must be conceived as inherently live, dynamic, active, powerful. Now anything that concretely exists at all, and that therefore has categorical (actual, existing) properties, ipso facto has powers.20 So to say that the existence of conscious process involves the existence of powers is not to say anything in any way peculiar. We can (and must) think of consciousness-­ process C as having powers, like any concretely existing thing whatever; and plainly the powers of a thing, considered specifically as such, need not be obviously manifest at all times in what one might call the overt being of the thing (the ‘overt being’ of C at any given time is C’s having the total conscious experience it has at that time, the conscious experience that constitutes its existence). There is, then, a completely harmless—unavoidable—sense in which the process of thinking is not fully manifest to itself (unless—I suppose—one is God). Descartes makes the point explicitly when replying to Arnauld: ‘although we are always actually aware of the acts or operations of our minds’, he says, ‘we are not always aware of the mind’s faculties or powers, except potentially’.21 But there still appears to be an enormous difficulty: how can any human consciousness–process, of the sort with which we are all familiar, possibly carry or harbour all the powers—or, more neutrally, properties—we associate with an individual human mind like C—will, understanding, memory, linguistic ability, possession of concepts, innate or not?22 How can the categorical being of C—which is, again, simply consciousness-process, consciousness-process which appears at any particular time t to have only some very particular and extremely limited content— possibly wholly constitute what one might call the power being of C at t? Right now I’m absorbed by a white cow in a field, swishing its Sideritsian tail, and my ­knowledge of French and algebra isn’t manifest in this experience. Somehow Descartes has to find room for a mind or self or subject with sufficient ‘ontic depth’—sufficient capacitational depth, adequate ‘storage’. And this last requirement may seem to sink his position—unless perhaps Leibniz can help, by modifying it in some way.

 This is my view. It is very widely held, but the idea of perfectly ‘inert’ material substance was also widely embraced in the seventeenth century. 21  AT7.246–7/C2.172. Descartes continues: ‘by this [i.e. the word ‘potentially’] I mean that when we concentrate on employing one of our faculties, then immediately, if the faculty in question resides in our mind, we become actually aware of it’ (AT7.246–7/C2.172). 22  It’s precisely this problem that leads Hume to give up his official empiricist account of the nature of mind—‘all my hopes vanish’—in the Appendix to his Treatise (1739–1740: Appendix §20). 20

74

G. Strawson

It may perhaps be said that the problem disappears given a flash theory of being, but even on the terms of a flash theory of being we will need successive flashes that have capacity-carrying content or ‘depth’ over and above the evident conscious content experienced by the human individual in the present moment, if there is to be any real plausibility in the idea that each flash gives rise causally to the next. For although I am currently delighted by a cow, I may in the succeeding flashes hear and understand something in French and reply appropriately in that language. Sautrāntika Abidharma, facing very much the same problem, posits capacity-­ carrying mental dharmas called ‘seeds’ to do the required work (see e.g. Siderits 2007: §§6.3–6.5). The same problem presumably motivates the various accounts of ‘storehouse consciousness’ (Sanskrit ālaya vijñāna) and ‘minimal sentience’ (bhavaṅga citta), although they may vary on the question of whether or not there is really some sort of consciousness at all times (as a strong parallel to Descartes would require). There are clear connections even though the Cartesian position accepts the notion of a persisting mind while the various Buddhist positions reject it. At the very least they face a similar difficulty. What is to be done? The fact that all concrete entities necessarily have powers allows us to put aside the objection that the whole being of the consciousness-­ process that is C at t must be fully manifest in consciousness at t—manifest in the ordinary everyday consciousness of the subject of experience that we take to be identical with C. When it comes to finite minds like ours, the power being of something needn’t be—can’t be—fully manifest in the way it currently appears to us in our conscious experience. This is so even when the something is ourselves, and even when it is wholly constituted of consciousness. Descartes explicitly says of immaterial minds that ‘I understand them as powers or forces of some kind’ (AT5.270/C3.361), and we can register this by adding an explicit reference to power into the existing equation [1] (C = thinking = consciousness), and give it a new number to mark the addition: [2] C = consciousness = thinking/power of thinking. [2] is perhaps not easy to understand. The being of the power of the consciousness-­ process at any given time has to be nothing over and above the being of the actuality of the consciousness-process at that time, and the consciousness actuality may still seem too thin, at any given time, to be (or ‘ground’) the power. But it is, I submit, Descartes’s view, so that any difficulty in it is his difficulty. It’s also wholly in accord with a radical processual view of concrete reality that is I believe correct independently of Descartes.23

 It’s worth comparing the case of matter, or rather physical stuff in general. There is I believe no difficulty in saying that the being of the power of physical stuff is nothing over and above the being of the actuality of physical stuff. I take it that this is in fact correct. 23

Descartes and the Buddha—a rapprochement?

75

8 The Cartesian Brain We can solve almost all of the storage problem straight away. It’s enough to note the central place of the body, and above all the brain, in Descartes’s overall account of our mental lives.24 All our ideas, memories, knowledge are stored there, not in C. The human imagination is in Descartes’s view a wholly neural phenomenon. Objection. ‘The brain is fundamental in Descartes’s theory, but it can’t do all the work of storage. He’s committed to the existence of innate ideas, which must be located in C, along with what he calls “intellectual memory”, i.e. wholly non-­ corporeal memory, “memory that depends only on the soul” (AT3.48/C3.146).’ Reply. Innate ideas don’t constitute a separate problem, for they are in Descartes’s view nothing over and above powers. ‘When we say that some idea is innate in us, we do not think that it is always present to us; in that sense no idea would be innate. We mean only that we have within us a power to produce the idea in question’ (AT7.189/C2.132; see Clarke 2003: 162). Descartes explicitly rejects any conception of innate ideas as intrinsically contentful items ‘always actually depicted in some part of our mind, in the way in which many verses are contained in a book of Virgil’ (AT11.655). So the supposed problem posed by innate ideas reduces to the problem—of how C can be intrinsically powerful. Descartes’s occasional references to ‘intellectual’ or purely soul-based memory can’t be accommodated in this way, but they have a strangely glancing character in his writings, like his reference to ‘internal emotions which are produced in the soul only by the soul itself’ (AT11-1.440/C1.381). They don’t sit well with his overall theory, or his clearly stated view that ‘purely spiritual beings have no genuine recollection at all’.25 For present purposes, though, we may accept that Descartes commits himself to the existence of purely C-based memory, and that the storage problem remains even when we have built powerfulness into C, as in [2] C = consciousness = thinking/power of thinking. I think we find further striking—overwhelming, if indirect—support for the attribution of [2] to Descartes when we compare his conception of mind with his conception of matter. I’ll come to this in §10.

9 ‘Thin’ Subjects I want first to consider another reason to attribute the Mamertian view of the mind (i.e. [1]/[2]) to Descartes. I think it constitutes further powerful support for the attribution, although it’s not clear that more support is needed, given the textual evidence set out in §§4–6. But before that I need to step back a little.  ‘Above all the brain’: Descartes holds that ‘a lute player, for instance, has a part of his memory in his hands’ (AT3.48/C3.146). 25  AT3.425; Clarke (2003: 99–105) has a very good discussion of this point. 24

76

G. Strawson

Consider René, a human being, a subject of consciousness or experience (I’ll use ‘experience’ instead of ‘consciousness’ because of the naturalness of the phrase ‘subject of experience’). What is a subject of experience? There are several views. There is, first, the thick conception of the subject, according to which the subject is the whole organism, e.g. (to take our own case) the whole human being. There is, equally, the traditional inner conception of the subject, according to which the subject is a persisting entity which isn’t the same thing as the whole human being, but is instead a particular persisting complex brain-system, say, or alternatively, perhaps, a Lockean, C*-type immaterial soul (such an immaterial soul may not be thought of as literally inner, but the idea is clear). In both the thick and the traditional inner case it’s standardly supposed that a subject of experience can continue to exist when there is no experience that it is the subject of. Both these conceptions of the subject allow for the possibility that René can continue to exist as subject of experience during dreamless sleep of the kind we believe human beings regularly experience (or rather do not experience). According to the thin inner conception of the subject of experience, by contrast, a subject of experience is something that exists only if experience exists that it is the subject of. I believe this to be the most important philosophical conception of the subject, and it is in any case the one that will concern me here. A thin subject, then, is something that is essentially experientially ‘live’. If there’s no experience/experiencing, there’s no subject of experience: [3] [subject of experience → experience] and since it’s a necessary truth that an experience entails an experiencer [4] [experience → subject of experience] we can express the thin conception of the subject of experience as a biconditional: [5] [experience ↔ subject of experience]: a subject of experience exists if and only if experience exists—experience that it is the subject of. It’s important to be clear that [4] says nothing about the ontological category of the subject. In particular, it carries no suggestion that the existence of the subject of experience is in any way ontologically over and above the existence of the experience, the experiencing. All it does is point out is the fundamental (trivially true) sense in which experience is necessarily experience-for—experience-for-someone-­ or-something. Experience necessarily has this structure even if that all that exist are experience flashes. Pain is essentially pain-for, it’s essentially felt, hence, necessarily, felt by someone-or-something. [4], then, is compatible with any view that admits the existence of pain (and accordingly values compassion), and therefore, I take it, with the vast majority of Buddhist views. Whatever the no-self doctrine denies, it doesn’t deny this. Some

Descartes and the Buddha—a rapprochement?

77

may like to say that there is subjectivity but no subject, but there’s no issue here, for my use of the term ‘subject’ adds nothing whatever to what they allow.26 We may call the subject as represented by the thin conception of the subject the ‘thin subject’. We may then express the present point by saying that this is Descartes’s conception of the subject: [6] C is a thin subject. This is not news, in the context of this paper, because it follows immediately from [1] C is wholly constituted of (conscious) mental goings-on. If [6] is also, as I believe, the most natural Buddhist conception of the subject of experience—remember that the existence of subjects of experience, however momentary, must be allowed as soon as the existence of experience, e.g. pain, is allowed—then we have a further point of connection. It may be said that Cartesian subjects are sempiternal, whereas Buddhist subjects come in flashes. But we have already seen that Cartesian subjects also come in flashes. Objection. ‘This can’t be said to be a Buddhist view. Certainly there is experience; but no experiencer. “The Buddhist does not countenance … an experiencer” (Siderits 2007: 128). Reply. Here Siderits is using ‘experiencer’ in the standard way that builds in commitment to [i] some sort of substance–property distinction and [ii] some idea of the experiencer as something persisting. Both of these commitments are explicitly cancelled by the thin conception of the subject. This leaves us free to assert [4] as an absolute metaphysical truth, as we must—while remaining wholly friendly to Buddhist views.27

10 Mind and Matter Many will agree that §§4–6 already provide a sufficient defence of [1], and therefore [6], but there’s a further consideration which is I think of great importance. I have in mind the deep—all-the-way-down—resemblance between Descartes’s conception of the nature of mind and his conception of the nature of matter. Descartes holds that matter is quite literally nothing other than extension. This being so, it seems hardly surprising that he holds that mind is quite literally nothing other than thinking (i.e. experiencing, consciousness). Descartes’s view about matter is often derided. One common objection is that there can be no difference, for him, between a cubic metre of deep space and a cubic metre of lead or cheese. What such objections show, however, is that Descartes  I discuss this in a number of places. See e.g. Strawson (2010/2017: 167–8).  I once wrote a paper called ‘What is the relation between an experience, the subject of the experience, and the content of the experience?’ Strawson (2003/2008) The answer to the title question, I propose, is ‘Identity’; strict metaphysical identity. 26 27

78

G. Strawson

(obviously!) doesn’t conceive of space or extension in anything remotely like the way we do in everyday thought. It’s arguable that he conceives of it more in the way Steven Weinberg does when he proposes that all physical objects of the sort we take ourselves to have to do with are made of ‘rips in spacetime’, spacetime being itself a physical object, an essentially substantial something that is itself, in some immovable sense, the only thing there is.28 One might add that in conceiving of space— extension—as substantial, Descartes is also in effect conceiving of it as something inherently powerful, in spite of his official conception of matter as passive. This, again, is in accord with orthodoxy in current physics, according to which there is simply no such thing as empty space(time), but rather the ‘quantum vacuum’, a substantial physical thing. On this view, there is strictly speaking only one material thing or substance, the spatially extended universe, one big extended thing ‘with different nubbly gradients of texture’29 at different places that amount to trees, people, railway lines, and so on—an idea which, once again, seems profoundly in accord with the spirit of much present-day physics—relativistic quantum field theory—and cosmology. It also appears to accord with some version of the doctrine known as Indra’s Net. Consider any proper part of the material universe that we are naturally inclined to pick out as a truly independent single thing or substance. There are no such truly or radically independent things in Descartes’s metaphysics—no plurality of irreducibly independent material substances. Whether or not one accepts this thing-monist idea,30 the material universe is in Descartes’s view a ‘plenum’, that is, it contains no literal vacuum (the definitional opposite of plenum), no place that isn’t occupied by matter. How could there be such a place, given that extension is matter—that extension is itself something concrete, substantial? So too, it contains nothing that isn’t interdependent with everything else. Descartes takes this thing-monist view of matter to be compatible with ordinary conventional talk of multiple distinct material objects, and he never explicitly suggests that there is in the same way really only one mind—one res cogitans. He takes it, as remarked, that there are many irreducibly numerically distinct individual minds, fully in line with conventional Christian eschatology. He claims—but without ever offering any account of their identity and individuation conditions—that ‘each of us understands himself to be a thinking being and is capable in thought of excluding from himself every other substance’ (8A.29). It’s worth noting, though, that this form of words doesn’t unequivocally exclude the view that we are all ultimately one thinking substance. In their absorbing book The Rise and Fall of Soul and Self Ray Martin and John Barresi observe that ‘although officially Descartes subscribed to a doctrine of personal immortality, how personal it was is open to

 Weinberg (1997: 20).  Catherine Wilson, in correspondence. 30  A thing monist (e.g. Spinoza) holds that there is only one thing. A stuff monist (e.g. Hobbes) holds that there is only one kind of stuff, although there may be many distinct things. 28 29

Descartes and the Buddha—a rapprochement?

79

question’ (2006: 136). Some have suggested that when Spinoza went on to explicitly endorse the idea that our minds are ultimately aspects (or modes) of one single universal mind, he was simply rendering Descartes’s philosophy fully consistent.31 I’m going to leave this line of thought hanging. Even if one doubts that there is anything resembling the ātman-brahman view in Descartes (he would certainly have known of Averroës’ cognate view), an absolutely fundamental similarity between his view of mind and his view of matter remains: just as he holds that spatial extension is literally all there is to the latter, he holds that thinking (consciousness, experiential process) is literally all there is to the former. Plainly we need to make a truly radical adjustment to our conception of extension if we want to acquire any remotely realistic sense of what Descartes thinks matter is when he equates matter and extension. We need to adjust it to the point where it allows us to accommodate the fact that Descartes is as realist about ordinary massy physical objects as we are, while at the same time holding that the existence of matter is simply the existence of extension. We shouldn’t therefore be in the least surprised, but should rather expect, that we will need to adjust our conception of mind no less radically if we want to acquire any remotely decent sense of what Descartes thinks mind is—and of what he thinks an individual mind is. We will need to adjust it to the point where it can accommodate the fact that Descartes is as realist as we are about the capacities of minds —minds coupled to bodies and in particular brains which solve almost all if not all their ‘storage’ problems—even though he holds that there is nothing more to their existence than conscious process. His conception of the nature of experiential process must be at least as rich—at least as power-­full—as his conception of the nature of concretely existing extension needs to be if it is to accommodate all the seeming massy phenomena of the nonmental physical world in the way he thinks it does. One hasn’t begun to understand Descartes’s conception of mind until one sees that it invites exactly the same kind of initial puzzlement as his conception of body. (‘Body can’t be just extension!’ ‘Mind can’t be just a stream of conscious goings-on!’) Does this help with the problem of ‘ontic depth’ mentioned above? It may still be hard to see how. For whatever the nature of this depth, these riches, the whole categorical being of mind or self or subject is still held to consist of nothing other than mental process—conscious mental process (= [1]). And when we ask what this conscious mental process is supposed to consist in, we may still be inclined to think that it must (by definition) be something whose whole categorical being at a given time t is phenomenologically manifest in consciousness at t—something, in fact, whose whole categorical being just is what is phenomenologically manifest in consciousness at t. And when we then turn to consider C at t, and consider what is phenomenologically manifest to C, who is perhaps absorbed in watching a cow, it seems painfully obvious that there just isn’t enough room, in the total (categorical) being of C’s conscious mental process, for all the powers we want to lodge in C at

 Spinoza offered a systematization of Descartes’s philosophy in his Principles of Cartesian Philosophy (1663). 31

80

G. Strawson

t—even though the body and brain can wholly solve the problem of content storage (one’s knowledge of French, Arabic, chess, and so on). It’s true that that which constitutes C’s fully explicit occurrent content-process at t is necessarily the ground of certain powers, powers that are not themselves consciously displayed at t, simply because (in any sound metaphysics) all categorical being is also and necessarily and at the same time power being. But if—since—it can’t be the ground of all the powers and capacities we want to attribute to C, there must be more conscious content occurring somewhere in C, as part of C; because, again, all we have, when it comes to furnishing categorical grounds for powers and capacities, is occurrent conscious content. It seems, then, that we have to reject the thesis that the whole being of C’s mental process at t is fully manifest (self-manifest) to C in consciousness at t, and allow that the conscious process that constitutes C at t can after all include something of which C is not conscious or only dimly conscious. One question is whether some variant of the notions of ālaya-vijñāna (‘storehouse consciousness’) or bhavaṅga citta (‘minimal sentience’) or indeed citta-santāna (‘mind-stream’) can provide the necessary accommodation. I would be extremely interested to hear of any possible parallels. The move may seem to be immediately ruled out given Descartes’s view of the self-transparency of mind. But the thesis of the self-transparency of mind is not something additional to the thesis that the categorical being of the mind consists wholly of conscious process, it’s part of it; and I think we may now do five helpful things: (i) recall the passage where Descartes says he conceives of minds ‘as powers or forces of some kind’ (ii) note again the respect in which powers qua dispositional neither are nor can be manifest in the essentially occurrent (i.e. non-dispositional) content of the process of consciousness, but may nonetheless reside in the process of consciousness (iii) note again the primordial parallel between Descartes’s conception of matter as extension—something that seems initially far too thin—and his conception of mind as conscious process—something that again seems initially far too thin (iv) signal the important qualification in his statement of the self-transparency thesis in the First Replies, which allows for different—and perhaps very faint— degrees of awareness: ‘there can be nothing in me of which I am in no way [nullo modo] aware’ (AT7.107/C2.77) and finally (v) note a passage in which Descartes seems to state explicitly that the thinking— the conscious process—may at a given time be the place of residence of mental powers that are not grounded merely in the existence of the particular and limited conscious content of the thinking at that time: just as ‘the memory of material things depends on traces [vestiges] that reside in the brain’, he says, so too ‘the memory of intellectual things depends on … traces that reside in thinking [la pensée] itself’ (AT4.114).

Descartes and the Buddha—a rapprochement?

81

11 A Leibnizian Solution? Objection. ‘This can’t work. A genuine acknowledgement of what one might call the ‘power being’ of C amounts—in spite of (i)–(v)—to giving up [1]. It amounts to giving up the view that C is wholly constituted of actual conscious process. It is irredeemably illegitimate to expand [1] into [2] in the way you have done.’ Reply. This may be a problem in Descartes’s position rather in this account of his position. But I don’t think it’s insuperable. It may be—for one thing—that a broadly speaking Leibnizian solution is available. A full exposition of this idea would require another paper, but I’ll conclude with a sketch of how it might go. We need to be careful when we talk of powers, but it’s perfectly acceptable to say that all powers require categorical grounds (it’s just that one shouldn’t in saying this wrongly think that their being—that in which their existence consists—is distinct from or over and above the being of their categorical grounds). By hypothesis, the only categorical grounds available in the case of C are actual conscious goings-on. We must therefore suppose that conscious goings-on ground (constitute) all C’s powers. If we do this, it seems that we can hold on to [1]—the existence of C is wholly a matter of the existence of conscious mental goings-on—even as we think of C as intrinsically mentally powerful. What we can no longer hold on to, it seems, is the thesis of the full self-transparency of mind. We can continue to hold that all mental process is necessarily conscious, live, and it seems that we must do so if we are to present a position that can hope to be genuinely Cartesian. But we must then allow that there is conscious mental process that is part of what constitutes one’s mind at a given time t, and constitutes in particular its needed ontic (capacitational) depth at t, although—how to put it?—the conscious subject that one experiences oneself to be at t, and experiences as having the experiences one has at t, isn’t conscious of it, or (rather) is conscious of it only in a very dim way. Does this make sense? We already have Descartes’s crucial qualification of the self-transparency thesis: ‘there can be nothing in me of which I am in no way [nullo modo] aware’ (AT7.107/C2.77). That may already be enough (compare the fact that ālaya-vijñāna or storehouse consciousness is sometimes said to be not unconscious but ‘subliminal’). Another way to try to make sense of it, I think, is to suppose that one’s mind (or any unenlightened mind), a stream of consciousness, consists of more than one ‘locus’ of awareness (not ultimately, perhaps, but for all practical experiential purposes). Let ‘top subject’ name the putative entity picked out at the end of the last paragraph: the conscious-subject-that-one-normally-experiencesoneself-to-be-and-to-be-having-the-experiences-one-experiences-oneself-as-having. One can then express the present proposal by saying that there may be more to the full metaphysical reality of the subject of experience that one is—that one is correctly said to be—than the ‘top subject’ as just defined. One can allow that experience occurs, experience that is correctly said to be constitutive of the being of the subject of experience that one is considered as a whole, in loci of awareness (within

82

G. Strawson

the overall subject of experience that one is) that are other than the top-subject locus of awareness. The phenomenological content of this experience must be fully consciously manifest somewhere, because its being fully consciously manifest is an essential feature of what its existence consists in, and since it can’t be fully manifest in the experience of the top subject, it must be fully manifest in other loci of awareness— although one can again allow, in line with the nullo modo qualification, that the top subject always has some sort of dim awareness of it. This may seem to sit ill with Descartes’s doctrine of the indivisibility of mind, but I think that it may be part of what Leibniz has in mind when he postulates tiny experiences or conscious mental goings-on (petites perceptions) that are genuinely partly constitutive of one’s mind, although one, i.e. the ‘top’ subject, is not conscious, or is only very dimly conscious, of their content (see e.g. Leibniz c. 1704: 54). Jointly they constitute a mental entity that has (coupled as it is to body and brain) the necessary ‘thickness’ to be the place of residence of all one’s mental ­powers. And this may be so whether they’re perceptions of which we are in some sense aware although they’re individually too small for us to notice (the distant roaring of the sea is the sum of the separately indiscernible sounds of many individual waves), or whether there is really no sense in which we (the top subjects) can be said to be aware of them.32 If this picture makes sense, as I think it does, then it provides a means of greatly expanding the categorical being, and hence (equivalently) the power being, of C. I think, in fact, and contrary to Descartes, that we are in fact wholly material beings, but I also think it extremely likely (empirically very plausible) that there are loci of consciousness in us, so considered, of which we as ‘top’ subjects are unaware (or only very dimly aware) even when we judge ourselves to be fully awake and conscious. Does this idea oblige me—or Descartes—to say that an individual’s existence as a subject of experience is partly constituted by or at least dependent on many subjects in an ontologically substantial way that is fatal to Descartes’s thesis of the indivisibility of the mind? I think not. Not even the most passionate proponent of the indivisibility and ‘substantial simplicity’ of the mind can deny that the mind is highly complex. Many may still doubt that the picture offers a sufficient response to the question how anything that consists of nothing but conscious mental process can have the

 Some think that Leibniz’s notion of petites perceptions or unconscious perceptions extends to cover goings-on that are completely unconscious in a modern sense of ‘unconscious’ according to which they have no phenomenological content at all. I believe that this is wildly anachronistic, and that it is for Leibniz definitionally true that anything that is called a ‘perception’ or a ‘representation’ or a ‘thought’ has phenomenological content. A perception fails to be ‘conscious’ (if it does) simply in not being a distinct—or distinguishable—object of explicit awareness in a creature’s everyday experience. 32

Descartes and the Buddha—a rapprochement?

83

whole mental power being of subjects of experience like ourselves.33 Certainly more needs to be said. I feel no difficulty with the idea, although I’m a ‘stuff monist’ and an all-out materialist, because it seems to me that the least implausible view of the intrinsic nature of the physical stuff which constitutes concrete reality is that it is—or at least essentially involves—conscious experience. I think, in other words that panpsychism in some form is the least implausible position to hold about the nature of concrete physical reality in the present state of our knowledge. And here my use of the word ‘physical’ is, as always, the same as Russell’s: the word ‘physical’, in all preliminary discussions, is to be understood as meaning ‘what is dealt with by physics’. Physics, it is plain, tells us something about some of the constituents of the actual world; what these constituents are may be doubtful, but it is they that are to be called physical, whatever their nature may prove to be. (1914: 150)

Let me finish with a remark about Lichtenberg—and ‘pure consciousness experience’.

12 Lichtenberg Lichtenberg objected that Descartes didn’t have the right to assert cogito, i.e. ‘I am having conscious experience’, only ‘There is conscious experience’, or ‘It’s conscious-­experiencing’—as in ‘It’s raining’ (1765–99: 190, §18). His objection has been endlessly repeated and praised, but I don’t think it’s valid. Descartes uses ‘I’ simply because our thought about ourselves naturally and inevitably occurs for us in terms of ‘I’. He uses ‘I’ successfully as a referential term to refer to something that certainly exists (see §8 above), as he carries through his thought-experiment of doubting everything he can, without claiming to know the metaphysical nature of the something. He states explicitly that he doesn’t know its metaphysical nature: What else am I? I will use my imagination. I am not that structure of limbs which is called a human body. I am not even some thin vapour which permeates the limbs—a wind, fire, air, breath, or whatever I depict in my imagination; for these are things which I have supposed to be nothing. Let this supposition stand; for all that I am still something. And yet may it not perhaps be the case that these very things which I am supposing to be nothing, because they are unknown to me, are in reality identical with the ‘I’ of which I am aware? I do not know, and for the moment I shall not argue the point, since I can make judgements only about things which are known to me. I know that I exist; the question is, what is this ‘I’ that I know? (1641: 18)

 But perhaps this is to take the claim that our mental being is ‘everything that exists within us in such a way that we are immediately aware of it’ (AT7.160/C2.113) too narrowly. Why not understand this immediate awareness (and self-transparency of mind) to extend beyond the contents of present occurrent conscious experience to cover all those dispositional contents (beliefs and other mental states) of which we may be said to be immediately aware inasmuch as they are immediately available to us? (Thanks to Stephen Gaukroger and James Hill for prompting this thought. What they have in mind seems to be what Descartes calls ‘potential’ awareness in the quotation on p. 8.) 33

84

G. Strawson

He doesn’t know. He might for all he knows be something that’s more like a process or a property than a self-subsistent substance as ordinarily conceived. He might be just a ‘thin vapour’ permeating the body, i.e. a wholly physical phenomenon, nothing like a ‘Cartesian subject’ as ordinarily and mistakenly conceived of (a Lockean C*—see p. 000). The word ‘I’ carries no such implication. The force of cogito in Descartes’s use is in fact precisely ‘There is conscious experience going on’, ‘There is conscious process’. It is precisely what Lichtenberg criticizes it for not being. Later, in the Sixth Meditation, Descartes argues that he is indeed a substantial something. But this claim, once again, is not the claim that he is a C*. It’s the claim that he is C, something wholly constituted of consciousness—something whose existence Lichtenberg also and equally admits. Once again: Descartes doesn’t hold that there is some substance distinct from his conscious experience that is doing the experiencing—some substance distinct from his stream of consciousness that is the independently substantial locus of the stream of consciousness. He only holds, with Lichtenberg, that there is conscious process. This processual phenomenon is the ‘substance’ in question, and it is the thing referred to by ‘I’ when one uses ‘I’ to refer only to one’s mental being. The use of ‘I’ is unexceptionable. There remains the thought that this consciousness process is an individual consciousness-­process, one among others. But, first, it is not clear that Lichtenberg rejects this claim when he says that Descartes was only entitled to assert ‘there is thinking (i.e. conscious experience) going on’ (Lichtenberg was not then concerned with scepticism about other minds). Secondly, and as already remarked, it’s not clear that Descartes holds that an individual human mind is an irreducibly separate entity—even though he is concerned not to tread on orthodox toes. Material objects as ordinarily and conventionally understood are not irreducibly separate substances, on Descartes’s view, but are, rather, and ultimately, parts of one single thing— Extension. It’s arguable that the deep trend of Descartes’s thought is, as Spinoza saw, towards the parallel but at that time dangerously unorthodox view about minds: all minds are ultimately parts of one thing—Mind. Objection. ‘All this time you’ve been forgetting Descartes’s claim that “even if all the accidents of the mind change, so that it has different objects of the understanding and different desires and sensations, it does not on that account become a different mind” (AT7.14/CSM2.10). This sufficiently shows that C can’t be simply and wholly constituted of its contents. It must have some principle of identity—and therefore some kind of being—that is independent of, over and above, all its particular contents.’ Reply. Not so. One can supply a principle of diachronic (flash-being-diachronic) identity for a mind without supposing that it has any kind of being over and above its particular contents. Causal linkage is enough—this episode or flash giving rise to this next one. This mind is this particular stream (flash-stream)—that’s it. Here again there seems to be a connection with certain Eastern conceptions of the mind. Objection. ‘Any such appeal to causal linkage amounts in effect to supposing that there is more to C than all the particular conscious contents.’ Reply. Conscious contents are inherently powerful, like all concrete things, and there seems to be no good reason to allow this objection. And here there is perhaps

Descartes and the Buddha—a rapprochement?

85

another resource: a conception of conscious process (well developed in Eastern philosophy) that allows that it can exist and be genuine conscious process while being ‘pure’ in some sense, i.e. without having any particular experiential contents of the sorts with which we are familiar. On this view, a cross section of C at any given time in the waking day will indeed reveal nothing but consciousness, conscious process, ‘thinking’, but while some of it will be ‘coloured’ (contentful in the everyday ways with which we are familiar), some of it will be pure; and a great repository of capacities. If Descartes has something like this in mind, he can allow for the cessation of all ordinary content in C (e.g. in dreamless sleep) while still holding that the mind is always conscious—as it must be since (once again) it is wholly constituted of consciousness. A cross section of C in dreamless sleep will reveal consciousness-process, thinking, but no particular everyday contents.34

References Clarke, D. 2003. Descartes’s Theory of Mind. Cambridge: Cambridge University Press. Cottingham, J. 1976. Commentary. In Conversations with Burman. Trans. with a philosophical introduction and commentary by J. Cottingham. Oxford: Clarendon Press. Descartes, R. 1618-60/1964-76. Oeuvres, ed. C. Adam and P. Tannery. Paris: Vrin. ———. 1619–50/1991. The Philosophical Writings of Descartes, vol. 3. Trans. J.  Cottingham et al. Cambridge: Cambridge University Press. ———. 1641/1985. Meditations Objections and Replies. In The Philosophical Writings of Descartes, vol. 2. Trans. J. Cottingham et al. Cambridge: Cambridge University Press. ———. 1648/1976. Conversations with Burman. Trans. with a philosophical introduction and commentary J. Cottingham. Oxford: Clarendon Press. ———. 1644/1985. The Principles of Philosophy. In The Philosophical Writings of Descartes, vol. 1. Trans. J. Cottingham et al. Cambridge: Cambridge University Press. Hume, D. 1739–40/2000. A Treatise of Human Nature, ed. D.F. Norton and M. Norton. Oxford: Clarendon Press. Kant, I. 1790/1973. On a Discovery. In The Kant-Eberhard Controversy, trans. H. Allison. Baltimore, MD: Johns Hopkins University Press. ———. 1781–7/1933. Critique of Pure Reason. Trans. N. Kemp Smith. London: Macmillan. Leibniz, G.C. 1704/1996. New Essays on Human Understanding. Trans. and ed. J. Bennett and P. Remnant. Cambridge: Cambridge University Press. Lichtenberg, G.C. 1765–99/2000. The Waste Books. Trans. R.J. Hollingdale. New York: New York Review Books. Locke, J. 1689–1700/1975. An Essay Concerning Human Understanding, ed. P. Nidditch. Oxford: Clarendon Press. Martin, R., and J. Barresi. 2006. The Rise and Fall of Soul and Self: An Intellectual History of Personal Identity. New York: Columbia University Press. Nadler, S. 2006. Spinoza’s Ethics: An Introduction. Cambridge: Cambridge University Press. Priestley, J. 1782/1818. Disquisitions Relating to Matter and Spirit. In The Theological and Miscellaneous Works of Joseph Priestley, ed. J.T. Rutt, vol. 3, 2nd ed. London.

34

 My thanks to Christian Coseru for some most helpful advice.

86

G. Strawson

Regius, H. 1647/1985. An Account of the Human Mind, or Rational Soul, Which Explains What It Is and What It Can Be. In The Philosophical Writings of Descartes, vol. 1. Trans. J. Cottingham et al. Cambridge: Cambridge University Press. Rozemond, M. 1998. Descartes’s Dualism. Oxford: Oxford University Press. Russell, B. 1914/1919. The Relation of Sense-Data to Physics. In Mysticism and Logic, 145–179. London: Longmans, Green and Co. ———. 1950/1956. Mind and Matter. In Portraits from Memory, 145–165. New York: Simon and Schuster. Siderits, M. 2007. Buddhism as Philosophy. Indianapolis: Hackett. Spinoza, B. 1663/1985. Principles of Cartesian Philosophy. In The Collected Works of Spinoza. Trans. and ed. E. Curley. Princeton: Princeton University Press. ———. 1677/1985. Ethics. In The Collected Works of Spinoza. Trans. and ed. E. Curley. Princeton: Princeton University Press. Strawson, G. 1994. Mental Reality. Cambridge, MA: MIT Press. Strawson, G. 2003/2008. What is the relation between an experience, the subject of the experience, and the content of the experience?. In G. Strawson Real Materialism and Other Essays Oxford: Oxford University Press. ———. 2010/2017. Fundamental Singleness: How to Turn the 2nd Paralogism into a Valid Argument. In The Subject of Experience, ed. G. Strawson. Oxford: Oxford University Press. ———. 2011. Selves: An Essay in Revisionary Metaphysics, Revised edition. Oxford: Oxford University Press. ———. 2021. Identity Metaphysics. The Monist 104: 60–90. ———. forthcoming. Descartes’s Mind 2. von Rospatt, A. 1995. The Buddhist Doctrine of Momentariness: A Survey of the Origins and Early Phase of this Doctrine up to Vasubandhu. Stuttgart: Franz Steiner Verlag. Weinberg, S. 1997. Before the Big Bang. In The New York Review of Books 44/10, June 12.

The Imperfect Reality of Persons Jonardon Ganeri

In this chapter, my interest is primarily in the thought of the great Ābhidhārmika Buddhist intellectual Vasubandhu (c. 360  CE), author of the compendious Abhidharmakośabhāṣya. He has famously been described by Mark Siderits as a reductionist about the self, influenced by the seminal work of Derek Parfit. What I will argue is that the identification needs to be handled with care.1

1 Suffering: The Cause, and the Cure If the Buddha is indeed a psychological doctor, struggling to cure us of our pain and suffering with the medicine that is his philosophy, it would seem that he is more like an official of the World Health Organisation than a general practitioner in a local surgery: his concern is with the global eradication of suffering, rather than with tending the needs of particular individuals. We ought not conclude, however, that the Buddha’s interest in persons is merely statistical, and this is because of another claim he makes, namely, that the path to the removal of suffering rests in freedom from mistaken conceptions of self, or mistaken ideas about what having this concept

 This chapter reproduces part of Chapter 6 of my book The Concealed Art of the Soul (Oxford: Oxford University Press, 2007). 1

J. Ganeri (*) University of Toronto, Department of Philosophy, Toronto, ON, Canada e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_5

87

88

J. Ganeri

consists in.2 So while the Buddha’s ultimate goal is characterisable in wholly impersonal terms, the means by which the goal is realised is necessarily tailored to the individual. For the Buddha’s teachings about the self are ineliminably pragmatic or ‘skilful.’ To any given audience, he will teach such a doctrine about self as is calculated best to divest that particular audience of its false sense of self. Indeed, to an audience made up of moral sceptics and hedonists, the Buddha is willing to declare that there is indeed a permanent self, because this will instil in that audience a sense of moral commitment and responsibility. Fostering a sense of self in those who do not have it may well be the best way to encourage a concern for their own future pain, and a concern for one’s own pain could well be the necessary precondition for a concern about the pain of others.3 And yet these prudential concerns bring with them distinctive kinds of suffering, anxieties about one’s own future and survival, not to mention a whole panoply of desires that in one way or another presuppose a belief in the ‘enduring essence’ of things. To combat those sufferings in an audience who already has a prudentially grounded moral sense, the Buddha will teach a more refined doctrine, that there is no enduring self, that the endurance of a self can’t be what really matters in questions of survival, moral responsibility and quality of life. But whatever doctrine the Buddha teaches, the aim is always tactical and pragmatic: to divest that audience of the sense of self that perpetuates suffering, so as to reduce the total amount of suffering globally and impersonally conceived, and to redirect the attention of the audience in the direction of ‘what matters’. It is therefore only with a great deal of caution that we can speak of the Buddha himself as having and advancing a theory of self or no-self at all.4

 nairātmya-darśanaṃ śreyo-mārgam. As Richard Gombrich observes, this is “the most fundamental doctrine of Buddhism: that Enlightenment consists in realizing that there is no soul or enduring essence in living beings,” in ‘Review of J. T. Ergardt,’ Numen 26 (1979), p. 270. The final section of the Snake Sutta rehearses this view. It is perhaps worth stressing that the claim is not that we labour under a self-deception. Self-deception has to do with wilful errors in my de se judgements about my own motivations and desires; the error to which the Buddha refers has to do with mistakes about the possession of a concept of self as such. It might nevertheless be the case that an error about the nature of the self is necessarily involved in any act of self-deception; Richard Holton argues this in “What is the role of the self in self-deception?” Proceedings of the Aristotelian Society 101 (2000/01), pp. 53–69. 3  “Likewise, each person holds himself most dear; hence one who loves himself (attakāmo) should not harm others.” Saṃyutta Nikāya i 172. The Connected Discourses of the Buddha, trans. Bhikkhu Bodhi (Boston: Wisdom Publications, 2000), p. 171. 4   See also Lambert Schmithausen, “Ich und Erlösung im Buddhismus,” Zeitschrift für Missionswissenschaft und Religionswissenschaft 53 (1969), pp. 157–170, “Spirituelle Praxis und philosophische Theorie im Buddhismus,” Zeitschrift für Missionswissenschaft und Religionswissenschaft 57 (1973), pp.  161–1186; Klaus Oetke, “Ich” und das Ich: analytische Untersuchungen zur buddhistisch-brahmanischen Atmankontroverse (Wiesbaden: Franz Steiner, 1988), pp. 57–242. For an overview of the debate: Alex Watson, The Self’s Awareness of Itself: Bhaṭṭa Rāmakaṇṭha’s Arguments against the Buddhist Doctrine of No-Self (Vienna: Publications of the De Nobili Research Library, 2007), §1.1. 2

The Imperfect Reality of Persons

89

Errors to do with the place the notion of self has in our mental lives run too deep to be corrigible merely through conceptual analysis alone. The ‘cure’ for these misconceptions is never simply to announce the right view. On the other hand, we have also seen that later Buddhist philosophers, participating in the Sanskrit intellectual world, did seek to articulate theories about the constitution and role of the concept of self; and they were also willing to engage with their philosophical opponents in substantive ‘truth-directed’ debate about the self.5 They are concerned, especially, to locate the source of error that underwrites the false thinking about self to which we are habitually attached. This project incurs two distinct philosophical commitments. One is to identify false constitutive accounts of the concept of self, and provide instead a true account of what our concept of self consists in. The other is to consider whether the concept of self, correctly described, has any place in a properly constituted mental life.6 ‘No Self!’, understood as a rallying cry, can refer to either project: it can be a call for the rejection of a mistaken understanding of what the concept of self commits us to; and it can be a call for that concept, even properly understood, to be shown the door.7

 For a discussion of the “Buddhist turn to Sanskrit” around the beginning of the first millenium: Sheldon Pollock, The Language of the Gods in the World of Men: Sanskrit, Culture, and Power in Premodern India (Berkeley: University of California Press, 2006), pp. 56–59. 6  One might, for example, compare arguments against the inclusion of the concept self and the term “self” with arguments for a corresponding claim about racist concepts and terms. Michael Dummett, for instance, argues that such terms introduce ‘non-conservative’ extensions in the language. Michael Dummett, Frege: Philosophy of Language (London: Duckworth, 1981). Alternatively, one might base an argument for the claim on the idea that the concept of self has outrun its original evolutionary value, as does Mark Leary, The Curse of the Self: Self-Awareness, Egoism and the Quality of Human Life (New York: Oxford University Press, 2004). 7  Compare Jon Elster, “Nobody at home,” London Review of Books 2–15 June 1983, pp.  9–11: “Buddhism offers three doctrines with respect to the self. First, it contains a theoretical critique of the notion of an enduring self, together with a constructive analysis of the actual unity and continuity of the person. …Next, [it] offers an account of the emergence of the illusionary belief in the self. …Finally, and crucially, [it] proposes a way of overcoming this spontaneously arising illusion.” 5

90

J. Ganeri

2 Was the Buddha the First Bundle Theorist? Everyone knows that the Buddha talked about five streams of psycho-physical elements (skandha-santāna), and said that it is a mistake to think about self.8 But did he mean that it is a mistake to think that the self is something over and above the elements, or did he mean that it is a mistake to think that the self is anything at all, when in fact there are only the elements? Is the idea that there is a right and a wrong way for the notion of self to figure in our mental lives, and what the Buddha does is to help us remove our attachment to false ways of thinking about self? Or is the point that we can give a perfectly correct account of the concept of self, but we go wrong in allowing that concept any place in our mental lives at all? Of necessity, Buddhist discussion of the self addresses both these questions. David Hume said that he could never catch himself without a perception, nor “observe anything other than the perception.” He claimed that we resort to the notion of a soul, and a self, and substance, “to disguise the variation” in the succession of our perceptions, a notion that is “a fiction, either of something invariable and uninterrupted, or of something mysterious and inexplicable.”9 The mind, he said, is really a ‘bundle,’ or a ‘commonwealth,’ or a ‘theatre’ of perceptions, and it is the bundle that settles questions about the identity of the person over time. He clearly does not think that the self consists in a bundle of perceptions, but it is less certain that he takes resorting to the fiction that there is a self to be necessarily an ill from which we must be cured. Derek Parfit rejects two claims: (i) that “[w]e are, or partly consist in, souls, or immaterial substances;” and (ii) that “[t]here are no persons, thinkers, or agents. There are only persisting bodies and related sequences of thoughts, experiences, and acts.” The first of these he calls the Cartesian View, the second, the Buddhist

 Richard Sorabji points to evidence that even the Greek and Roman philosophers in the first century CE knew it, because “the Buddhist reason for not fearing death is found both in the Stoic Seneca [c. 4 BCE–64 CE] and in the Platonist Plutarch [c. 46–127 CE], but in a way that remains incompatible with the rest of what they say, which suggests that it is an alien growth.” He is referring to Seneca Letters 24, 19–21; 58, 22–3, and Plutarch Plutarch On the E at Delphi 392 C–E. See his Self: Ancient and Modern Insights about Individuality, Life, and Death (Oxford: Clarendon Press, 2006), chapter 5, and his Emotion and Peace of Mind (Oxford: Clarendon Press, 2002), pp.  247–248. We might note here that Clement of Alexandria (c. 150–215  CE) mentions the Buddha in his Stromateis, saying, “Some, too, of the Indians obey the precepts of Boutta, whom, on account of his extraordinary sanctity, they have raised to divine honours” (Strom. 1.71.6). Clement of Alexandria, Stromateis, trans. John Ferguson (Washington: The Catholic University of America Press, 1991), p. 76. Ferguson comments that “[t]he reference to the Buddha is exceptionally interesting: Pantænus may have travelled in the East.” Pantænus was the teacher of Clement and the founder of the catechetical school in Alexandria; Eusebius (c. 260–340 CE) informs us that he went to India. It is certainly possible, too, that there were Buddhists in Alexandria then. 9  David Hume, A Treatise of Human Nature, edited by L.  A. Selby-Bigge (Oxford: Clarendon Press, 1964), pp. 252–253. 8

The Imperfect Reality of Persons

91

View.10 He has defended a third view, Reductionism, that “[o]ur existence consists in the existence of a body, and the occurrence of various interrelated mental processes and events. Our identity over time consists in physical and/or psychological continuity.”11 A reductionist claims that where there had seemed to be two things, we find that in fact we have only one, but does not deny the reality of either.12 In Reasons and Persons, Parfit relates Reductionism to the claim that there are indeed such things as persons, even though we could give a complete description of reality without mentioning persons;13 but sustained criticism of the ‘impersonal description’ thesis has led him to retreat from it.14 The idea behind that thesis is that talk of persons introduces no new entities. Talk of sequences and bundles of experiences is the basis for the introduction of persons into discourse, just as talk of bounded flows of water is the basis for our talk of rivers; but this talk introduces no new entities: what it does is to allow the old ones to be spoken of in new ways. However, it has seemed to many critics that talk of flows of experience depends on talk of persons, that one cannot talk about the first and not the second. An important consequence of Parfit’s work has been the considerable recent attention to the idea that personal identity is not what matters to survival. What I wish to consider now is Parfit’s discussion of what he has called the Buddhist View. Parfit refers to a passage which he claims shows that the Buddha himself said that there are no persons but only streams of experience.15 The text in question is a Nikāya passage from the Sutta on Ultimate Emptiness, which is quoted by Vasubandhu in his famous “Refutation of the Theory of a Self” (ātmavāda-­ pratiṣedha), a text also going by the name “Examination of the Person” 10  Derek Parfit, “Experiences, Subjects, and Conceptual Schemes,” Philosophical Topics 26 (1999), pp. 217–270; at p. 260. 11  “Experiences, Subjects, and Conceptual Schemes,” p. 218. 12  “Reductionist accounts aim to show that where we thought we had two sets of concepts, entities, laws, explanations, or properties, we in fact have only one, which is most perspicuously characterized in terms of the reducing vocabulary.…[but] there is little attempt to effect a translation of one discourse into another.” Kathleen Lennon and David Charles eds, Reduction, Explanation and Realism (Oxford: Clarendon Press, 1992), p. 2. 13  Derek Parfit, Reasons and Persons (Oxford: Clarendon Press, 1984), p. 212. 14  “[The impersonal description thesis], as I shall admit here, was a mistake.” “Experiences, Subjects, and Conceptual Schemes,” p. 218. The critics include John Campbell, Past, Space, & Self (Cambridge, Mass.: MIT Press, 1994), and Quassim Cassam, Self and World (Oxford: Clarendon Press, 1997). 15  Derek Parfit, “Experiences, Subjects, and Conceptual Schemes,” p. 260. In an interview, he says, “I do not accept the Buddhist no self view, since I believe that persons exist.” Cogito 9 (1995), p. 123. Parfit concedes that the claim might be simply an emphatic way of rejecting souls; but that is not actually what is going on either, as we will see. A passage that does seem emphatic in just that sense is quoted by Buddhaghoṣa: “For there is suffering, but none who suffers; there is doing but no doer; extinction, but no one extinguished; although there is a path, but no traveller (dukkham eva, na koci dukkhato, kārako na, kiriyā va vijjati/atthi nibbuti, na nibbuto pumā, maggam atthi, gamako na vijjatī” (16.90). Visuddhimagga of Buddhaghośācārya, critically edited by Swami Dwarikadas Shastri (Varanasi: Baudha Bharati, 1977), p.  433. Trans. Bhikkhu Ñyāṇamoli, The Path of Purification (Colombo: Semage, 1964), p. 587.

92

J. Ganeri

(pudgala-viniścāya).16 Parfit reads from an English version Theodore Stcherbatsky had prepared from Jinamitra’s Tibetan translation: “There are acts, and also their consequences, but there is no agent who acts …There is no person, it is only a conventional name given to a set of elements.”17 A comparison with the now available Sanskrit text18 reveals that something has gone wrong here. First, the term saṅketa has been read by the Tibetan translator according to the more technical meaning “conventional name” rather than the more natural and modest “(causal) agreement.”19 Second, and more seriously, the sense of the original ‘anyatra + ablative’ construction has been entirely lost. A better translation is provided by Duerlinger: “There is action and its result, but no agent is perceived that casts off one set of aggregates and takes up another elsewhere apart from the elements20 agreed upon (to arise interdependently).” So the passage cited does not deny that there are persons: it says that there no persons that are not comprised of streams. Nor does the text provide any support for the claim that the Buddha thought that persons are fictions, or that personal names are merely conventional (if by that what is meant is that they are

16  It is an appendix to his Abhidharmakośabhāṣya. Abhidharmakośa and Bhāṣya of Vasubandhu, critically edited by Dwarikadas Shastri (Varanasi: Bauddha Bharati, 1973), pp. 1189–1234. For translations, see James Duerlinger, Indian Theories of Persons: Vasubandhu’s “Refutation of the Theory of Self” (London: RoutledgeCurzon, 2003), and Matthew T. Kapstein, Reason’s Traces: Identity and Interpretation in Indian and Tibetan Buddhist Thought (Boston: Wisdom Publications, 2001), pp. 347–374. 17  Derek Parfit, “Experiences, Subjects, and Conceptual Schemes,” p. 260. The quotation is from Theodore Stcherbatsky, “The soul theory of the Buddhists,” Bulletin de l’Académie des Sciences de Russie (1919), Series VI, Vol. XIII, pp. 823–854 and 937–958; at p. 845. Parfit cites the same passage in Reasons and Persons, p. 502. There he says, “I claim that, when we ask what persons are, and how they continue to exist, the fundamental question is a choice between two views. On one view, we are separately existing entities, distinct from our brain and bodies and our experiences, and entities whose existence must be all-or-nothing. The other view is the Reductionist View. And I claim that, of these, the second view is true. As [the text cited] shows, Buddha would have agreed.” (p. 273). The Buddha may or may not have agreed, but Vasubandhu probably would have; and, ironically, if Parfit had had a better translation available to him, he might not have changed his own mind. 18  asti karma, asti vipākaḥ, kārakas tu nopalabhyate. ya imāṃś ca skandhān nikṣipati, anyāṃś ca skandhān pratisandhāty anyatra dharmasaṅketāt. Abhidharmakośabhāṣya, p.  1208. Compare Duerlinger, p. 88; Kapstein, p. 362. 19  Thus Poussin: “The meaning of the expression anyatra dharmasaṅketāt is not in doubt. The Vyākhyā [Yaśomitra’s commentary] explains dharmasaṅketād iti pratītyasamutpādalakṣanāt: ‘outside of the combination of the dharmas, that is to say, outside of the successive causation of the dharmas;” and elsewhere (ad iii.18): saṅketa = hetuphalasaṃbhandavyavasthā [agreement in the relation of cause and effect]. But Paramārtha [the Chinese translator; Jinamitra does the same] understands saṅketa as ‘metaphorical designation,’ from whence the translation, ‘One does not maintain the existence of an agent…except when, conforming to worldy usage, one says that the dharmas are a pudgala [person].” Louis de la Vallée Poussin, Abhidharmakośabhāṣyam, English trans. by Leo M. Pruden (Berkeley: Asian Humanities Press, 1988–90), Vol. IV, p. 1369. 20  Given that one objection to reductionism is that experiences are not ‘ontologically prior’ to persons, this use of the term ‘element’ for skandha should be taken with a light touch.

The Imperfect Reality of Persons

93

metaphorical or fictional designations), i.e. that the Buddha’s attitude towards persons was the same as Hume’s attitude towards selves. Vasubandhu’s point, in citing the passage, is simply that he has no quarrel with someone who speaks of a ‘person’ (pudgala), as long as they mean nothing more by that than the stream; and in particular, that they do not mean that what they call the ‘person’ is something ontologically independent of the stream. It is no accident that Vasubandhu makes this point as forcefully as he can here. In one of his sermons, The Burden, the Buddha spoke about “the burden, the carrier of the burden, the taking up of the burden and the laying down of the burden.”21 He then explains that the burden are the streams of psychological and physical elements; the carrier of the burden is the person (pudgala), “this venerable one of such a name and clan;” taking up the burden is craving, and laying it down is the complete cessation of craving. The view is summed up in verse: The five aggregates are truly burdens, The burden-carrier is the person. Taking up the burden is suffering in the world, Laying the burden down is blissful. Having laid the heavy burden down Without taking up another burden, Having drawn out craving with its root, One is free from hunger, fully quenched.

This sermon became a key text for one group of Buddhists, the Pudgalavādins, who maintained that the ‘person’ (pudgala) was something distinct from the aggregates.22 These are the people whom Vasubandhu is trying to refute in his “Examination of the Person.” Richard Gombrich says that this sermon “caused a good deal of trouble in the history of Buddhist thought… a whole school of Buddhists, the pudgala-vādin, took this text as their main authority for claiming that the person was a sixth entity, separate from the five khandha, since it was that person, not the five khandha, which was subject to craving and so picked the khandha up.”23 Vasubandhu, however, interprets the statement quoted by Parfit as saying that there is no harm in talking about a person, as indeed the Buddha himself does here, but

 Saṃyutta Nikāya III, 25–6. The Connected Discourses of the Buddha, trans. Bhikkhu Bodhi (Boston: Wisdom Publications, 2000), p. 871. 22  They numbered more than a quarter of all the Buddhist monks in India in the seventh century, according to the testimony of Hsüan-tsang (602–664 CE), the Chinese pilgrim-scholar who spent twenty years in India before returning to China in order to translate many Indian Buddhist texts. See Leonard C.  D. C.  Priestley, Pudgalavāda Buddhism: The Reality of the Indeterminate Self (Toronto: Centre for South Asian Studies, 1999); Thich Thiên Châu, The Literature of the Personalists of Early Buddhism (Delhi: Motilal Banarsidass, 1999), and “Les réponses des Pudgalavādin aux critiques des écoles bouddhiques,” Journal of the International Association of Buddhist Studies 10 (1987), pp. 33–53. It is possible that there is at least one living Pudgalavādin: Richard Sorabji. See his Self: Ancient and Modern Insights about Individuality, Life, and Death (Oxford: Clarendon Press, 2006). 23  Richard Gombrich, How Buddhism Began (London: Athlone, 1996), pp. 67–68. 21

94

J. Ganeri

only on the understanding that it is not something ‘separate from’ the five streams of aggregates.24 If neither the Buddha nor Vasubandhu is claiming that “there are no persons, thinkers or agents,” is it right to describe either of them a reductionist? And what about the Pudgalavādins? They are still Buddhists, after all, and like all Buddhists reject what Parfit calls the Cartesian View. A helpful metaphor here, one exploited by all Buddhists, is the metaphor of fire and fuel. The Buddha himself, in The Fire Sermon, said that the fires of lust, hatred and delusion burn away in the sense-­ faculties as well as in the perceptions and feelings they produce, whether pleasant or painful.25 Furthermore, the word for fuel is upādāna, a word whose other meaning in both Pāli and Buddhist Hybrid Sanskrit is ‘grasping, clinging, addiction, attachment.’ Related is the term upādāya, whose core meaning is ‘taking to oneself, assuming,’ and which is used in both languages in the sense of ‘on the basis of, making use of, depending on.’ This is also the very term that the Pudgalavādins (as reported by Vasubandhu) use to state their view, and it is immediately elaborated in terms of the fire-fuel metaphor: [The Pudgalavādins assert that] a person is not substantially real (dravyataḥ) or real by way of conception (prajñaptitaḥ), since he is conceived in reliance upon (upādāya) aggregates which pertain to himself, are acquired, and exist in the present … A person is [conceived] in the way in which fire is conceived in reliance upon fuel. Fire is conceived in reliance upon fuel, it is not conceived unless fuel is present and cannot be conceived if it either is or is not other than fuel… Similarly, a person is not conceived unless aggregates are present. If he were other than aggregates, the eternal transcendence theory [that a person is substantially real] would be held, and if he were not other than the aggregates, the nihilism theory [that a person does not exist at all] would be held.26

The Pudgalavādin is using the metaphor of fire and fuel to explain a sense in which there can be relations of conceptual and ontological dependence between two things. Parfit distinguishes, among various sorts of ontological dependence, between adjectival and compositional ontological dependence.27 The first is illustrated by the relation between a dent and a surface, the second by the relation between a tree and the cells of which it is composed. The difference between the Pudgalavādin and Vasubandhu might then lie in this: the Pudgalavādin thinks that persons are adjectivally dependent on streams of physical and psychological elements, while Vasubandhu thinks that the relation is one of compositional dependence (a person,  See also Steven Collins, Selfless Persons: Imagery and Thought in Theravāda Buddhism (Cambridge: Cambridge University Press, 1982), pp. 164–165. Collins notes that the Pali word hāra is better translated as ‘the act of carrying or bearing’ than as an agent noun, ‘the carrier or bearer,’ and remarks, “[t]the idea, then, is that the ‘person’ is a state created by the act of ‘picking up’ the burden of the khandhā, through desire, a state which simply consists in the act of ‘bearing the burden’.” 25  Saṃyutta Nikāya IV, 28. The Connected Discourses of the Buddha, p. 1143. 26  Abhidharmakośabhāṣya, p. 1163; trans. Duerlinger, pp. 73–74. Candrakīrti uses the same ‘neither…nor…’ argument, but to very different effect. 27  “Experiences, Subjects, and Conceptual Schemes,” p.  226. For a related distinction, see Vātsyāyana under Nyāyasūtra 3.1.1 (quoted below). 24

The Imperfect Reality of Persons

95

Vasubandhu says, is not other than the stream). For fire is essentially in the fuel it is burning, as dents are essentially in or of surfaces. We cannot say that the dent is something other than the surface, or that it is just the surface. If this way of distinguishing the views is right, then both have given textually admissible interpretations of the Buddha’s words, and neither endorses the Cartesian View; but only Vasubandhu is a reductionist. None of our Buddhists, though, holds to what Parfit has called ‘the Buddhist View.’ Indeed this is ucchedavāda, nihilism, one of the two extremes between which the Buddha sought a Middle Way, the other being śāśvatvāda, eternalism, the Cartesian View.28 Not that either of our Buddhists is yet out of the woods. The Pudgalavādin says that, distinct from yet dependent on the psycho-physical stream, the person is ‘inexplicable’ (avyaktavya); but that claim itself requires explication. And Vasubandhu seems to assume that it is possible to talk about the psycho-physical stream (skandha-santāna) without mentioning persons, and also seems to assume that the stream is not itself ontologically dependent, either adjectivally or constitutively, on persons; but we have come to appreciate how problematic both those assumptions are.

3 Why Persons Are Not Merely Conceptual Fictions One recent interpreter has said that “[t]he Buddhist Reductionist [i.e. Vasubandhu] claims that ‘person’ is a mere convenient designator for a complex causal series of impermanent, impersonal psychological elements. That is, ultimately there are no persons, only physical objects, feelings, perceptions, volitions and consciousness.”29 Another interpreter has said, however, that “[w]e need to realize that Vasubandhu does not reject the view that persons ultimately exist. For he too believes that conventionally real persons ultimately exist by reason of being the same in existence as collections of aggregates.”30 Vasubandhu himself says, famously, that “persons are real with reference to conception, real in the same sense that such things as heaps and streams are.”31 The distinction between what is real with reference to conception (prañjapti-sat) and what is real substantially (dravya-sat) is one I will return to in the next section, where I will argue that it is a distinction between two concepts of objectivity (and therefore not a distinction between the subjective and the objective, or between mere appearance and reality). Vasubandhu’s claim sounds, in any case, pretty good for persons. Heaps are made out of grains, but are real nevertheless; streams are made out of water, but are still streams. Vasubandhu does not seem  Accusing another Buddhist of having a view that implied a denial of persons is a common enough dialectical tactic among Buddhists; see Duerlinger, pp. 9, 23. 29  Mark Siderits, Personal Identity and Buddhist Philosophy: Empty Persons (Aldershot: Ashgate, 2003), pp. 24–25. 30  James Duerlinger, Indian Theories of Persons: Vasubandhu’s “Refutation of the Theory of Self” (London: RoutledgeCurzon, 2003), p. 21. 31  prajñapti-sat pudgalo rāśidhārādivat. Abhidharmakośabhāṣya, p. 1205. 28

96

J. Ganeri

to mean that persons are merely conceptual, that they are internal or subjective fictions. He seems to mean that reference to them is supported by a concept-mediated, fluid and aggregative model of classification, that the delimitation of the psycho-­ physical raw materials out of which persons are made into individual streams is not simply a matter of cutting nature at its own hard natural joints. In fact, Vasubandhu does not say either that persons are ‘ultimately real’ (dravya-sat), or that they are not ultimately real; and his silence is well-motivated. For to ask of something which has been agreed to be constituted out of other things whether it ‘really’ exists or not seems to be one of those questions Parfit describes as empty: we are in possession of all the facts; there is no further fact of the sort the question seems to ask for. It has been suggested that the import of his statement is that the term ‘person’ is, for Vasubandhu, a fictional term, and indeed that persons are merely for him conceptual fictions. We have already seen some of the difficulties that translation of the word saṅketa has given rise to, and similar difficulties accompany the word prajñapti. This word is rendered as “nominal” by Stcherbatsky and Matilal, “conventional” by Duerlinger, “relative” by Poussin, “conceptual” by Kapstein, and “convenient designator” by Siderits. Stcherbatsky’s translation leads him to represent Vasubandhu as thinking that persons are merely nominal entities, and the translation of prajñapti as “convenient designator” clearly runs the same risk.32 If ‘person’ is a mere name, like “Hamlet” or “the hare’s horn”, then it does not have a real referent, and persons are illusions of language and conception. It is a short step from there to agreeing with Hume that some fiction is involved. (I believe that Candrakīrti does indeed think that the term ‘self’ lacks a reference, but his reason is that the term is not a referring expression at all; it has a quite different linguistic role.) Does the claim that the reality of persons is dependent on concepts imply that persons are mere fictions? The idea, that because a term is relative to our interests and ways of classifying it is also a fiction, is reminiscent of a view associated with the early Nietzsche. Nietzsche claimed that language and concepts never reach the truth, because all concepts are merely abstractions from sense-data, and all words are but arbitrary assignments: If truth alone had been the deciding factor in the genesis of language, and if the standpoint of certainty had been decisive for designations, then how could we still dare to say “the stone is hard,” as if “hard” were something otherwise familiar to us, and not merely a totally subjective stimulation! We separate things according to gender, designating the tree as masculine and the plant as feminine. What arbitrary assignments! How far this oversteps the canons of certainty! We speak of a “snake”: this designation touches only upon its ability to twist itself and could therefore also fit a worm. What arbitrary differentiations! What one-­ sided preferences, first for this, then for that property of a thing! The various languages placed side by side show that with words it is never a question of truth, never a question of adequate expression.33  Kapstein discusses the problems with Stcherbatsky’s translation; Reason’s Traces, p. 348.  Friedrich Nietzsche, “Truth and lies in the non-moral sense,” in Daniel Breazeale, ed. and trans., Philosophy and Truth: Selections from Nietzsche’s Notebooks of the Early 1870s (Amherst, NY: Humanity Books, 1979), pp. 79–91; see esp. p. 81–82.

32

33

The Imperfect Reality of Persons

97

Nietzsche also said that “truth is an illusion we have forgotten is an illusion,”34 that we can come to know that our concepts fictionalise and distort the real nature of things, the implication being that we can somehow sneak round the side of the veil of ideas and compare our conceptual scheme with reality perceived from some impartial vantage-point. What the early Nietzsche forgot is that our concepts might be the only window we have on reality; there is no catching them from the side-on. And so, as Bernard Williams says, “it is trivially true that ‘snake’ is a human concept, a cultural product. But it is a much murkier proposition that its use somehow falsifies reality – that ‘in itself’ the world does not contain snakes, or indeed anything else you might mention.”35 Similarly here; even if person is a human concept, a humancentred way of grouping and classifying experience, it would not follow that ‘in itself’ the world does not contain such things as persons, that a person is “an illusion we have forgotten is an illusion.” The Buddhists need follow Nietzsche only part of the way in his saying that “there is no ‘being’ behind doing, effecting, becoming; ‘the doer’ is merely a fiction added to the deed  – the deed is everything.”36 They might say that the doer is the deed. Nor indeed need they say that because there is something human-centred in what we decide to count as a heap, it follows that heaps are ‘merely fictions.’37 Indeed it is their Nyāya critics who are the ones persistently to press these Buddhists into an admission that the identity of persons is, for them, an illusion like that of the circle of light generated by a swinging lamp. This is a pressure they just as consistently resist. Their insistence on the objectivity of persons, albeit different in kind from that of more fundamental sorts of entity, is unwavering. The admitted fact that persons are not objective in some absolute sense (paramārthasat) does not preclude their having objectivity in some other degree (saṃvṛti-satya). The idea behind the claim that persons are mere fictions is that the person is the sort of illusion which results from standing too far back from the facts of personal identity, namely the psycho-physical stream, just as a cloud or a table looks from a distance like a solid object, but dissolves into vapour or atoms on closer inspection. When it comes to settling questions about what there really is in the world, the assumption is that the ‘view from close up’ is the proper one to assume. We do not, however, typically regard clouds and tables as optical illusions simply because they disappear when we zoom in. Why should we not say, to the reverse, that when we look at an impressionist painting close-up and see only dots of colour, the disappearance of the painting is the illusion and the ‘real’ painting is the one seen from far off? It is for such reasons as these that we need to find a different anchor for the distinction invoked by Vasubandhu.  “Truth and lies,” p. 84.  Bernard Williams, Truth & Truthfulness (Princeton: Princeton University Press, 2002), p. 17. See also Maudemarie Clark, Nietzsche on Truth and Philosophy (Cambridge: Cambridge University Press, 1990). 36  Friederich Nietzsche, The Genealogy of Morals, trans. Walter Kaufmann (New York: Vintage Books, 1967), p. 45. 37  For an argument against the elimination of persons but in favour of the elimination of objects, however: Trenton Merricks, Objects and Persons (Oxford: Clarendon Press, 2001). 34 35

98

J. Ganeri

4 The Objectivity of Person-Involving Conceptual Schemes Vasubandhu himself has something more to say about the imperfect reality of persons. Something is real in the cloaking and concealing (samvṛti-sat) or conceptual (prajñapti-sat) sense if it no longer cognised when it is divided, either physically or in the mind. Something is real in the ultimate (paramārtha-sat) or foundational (dravya-sat) sense if it is still cognised even when so divided. We must remember, while reading his explanation, that in Vasubandhu’s world-view, everything is constituted out of primitive ‘components’ or elemental ‘qualities’ (dharma), some of which are physical (rūpa) and others psychological (nāma): That of which one does not have a cognition when it has been broken is real in a concealing way (samvṛti-sat); an example is a pot. And that of which one does not have a cognition when other [elemental qualities (dharma)] have been excluded from it by the mind is also conventionally real; an example is water. That which is otherwise is ultimately real (paramārtha-sat). (AK 6.4)

Vasubandhu elaborates: That of which one does not have a cognition when it has been broken into parts is real in the concealing way. An example is a pot, for when a pot is broken into shards, one does not have a cognition of it. And that of which one does not have a cognition when other elemental qualities (dharma) have been excluded (apohya) by the intellect (buddhi) is also to be known as real in the concealing way. An example is water, for when one mentally excludes its [physical] form (rūpa) and so on, one has no cognition of water. That is, the ‘concealing’ designation is applied to those things such as pots and water. Hence, when we say, within the scope of the concealing, “There is a water-jug and water,” we have spoken the truth (satyam); we have not uttered a falsehood. Hence, it is called a “concealing truth.” The existence of things in a way other than that is ultimate reality. That of which one still has a cognition even when it has been broken is ultimately real. And that of which one still has a cognition even when other elemental qualities are mentally excluded is also ultimately real. An example is [physical] form (rūpa). For when that form is broken into infinitesimal particles, one still has a cognition of that real thing (vastu) [namely, form]. And when other elemental qualities such as taste are mentally excluded, one still has a cognition of that whose nature is form. One should see that this is also the case with sensation (vedanā) and so on. Something is said to be ultimately real [or true (satyam)] because it exists ultimately.38

These definitions suggest that what is really at issue is not so much the existence of social practices or conventional fabrications, but what we might call ‘stability under analysis.’ An impressionist painting ceases to represent anything when investigated  Abhidharmakośabhāṣyam of Vasubandhu, ed. P. Pradhan and revised by Aruna Haldar (Patna: K. P. Jayaswal Research Institute, 1975), p. 334. Trans. John Dunne, Foundations of Dharmakīrti’s Philosophy (Boston: Wisdom Publications, 2004), p. 41, slightly modified. See also B. K. Matilal, Perception: An Essay on Classical Indian Theories of Knowledge (Oxford: Clarendon Press, 1986), pp. 246–247; John B. Buescher, Echoes from an Empty Sky: The Origins of the Buddhist Doctrine of the Two Truths (Ithaca: Snow Lion Publications, 2005), pp. 70–72. Poussin notes that the Chinese translator Paramārtha again departs from the original, with “Things like ‘a pot’ exist only as a metaphorical designation (prajñapti) of shape.” Poussin, Abhidharmakośabhāṣyam, English trans. Leo M. Pruden, Vol. III, p. 1045, n. 43. 38

The Imperfect Reality of Persons

99

too closely: the perceptual image is unstable under investigation. Vasubandhu introduces this idea in a slightly strange way. The problem is that, while we can well enough understand the way in which a pot is no longer perceived when it is divided into pieces, the argument does not work for water, which, being a sort of stuff, can be divided up without disappearing. So Vasubandhu says that things like water disappear from cognitive view when they are divided up in the mind. The text here, however, is ambiguous: does he have in mind a mental division of water into further constituents, or a mental division of water from other things; and if the latter, is the idea that water is mentally divided from things other than water, or is the idea rather that water is divided from its own properties? I think the first reading has to be the correct one. On the second reading, I divide water from things other than water, but there is no reason to think that I thereby lose my cognitive grip on water; in fact, just the opposite – distinguishing a thing from what it is not is a way to individuate it in thought.39 The last reading will mean that Vasubandhu is giving an argument similar to Aristotle’s “stripping argument” (Metaphysics 7.3). I try to imagine the thing without first one and then another of its properties, until eventually I find I am not thinking about anything at all. The trouble is that, on that interpretation, the definitions simply don’t connect with the discussion about the composition of persons, and their putative reduction to streams.40 Vasubandhu’s argument, as I take it, is not that we cannot retain our grasp of the idea of a person if in thought we strip away all the properties associated with being a person, but rather that the person slips out of cognitive focus when we ‘zoom in’ and divide the person, in thought, into the component elements of the stream. In fact, however, there is a simple reason why Vasubandhu’s argument looks like a “stripping argument,” and that is that the basic Abhidharma ontology is an ontology of tropes or property-particulars. So dividing something into its constituents just is a matter of dissolving it into its (particular) properties.41 In the Simile of the Lute in the Saṃyutta Nikāya, the story is told of a minister who hears the sound of a lute for the first time, and who is entranced by it. He asked  It is true that Vasubandhu seems to allude to a standard Buddhist theory of conceptual analysis, the so-called ‘exclusion from others’ (anyāpoha), a technique used to explain both the hierarchical arrangement of concepts and the semantic reach of general terms. The term “tree,” for example, excludes from its semantic range other kinds of vegetation, as well as animals, minerals, and so on, but not oaks, elms, and so on. However, that theory was only developed in this form by his successor, Dignāga. 40  Matilal does see a “stripping” argument here: “We cannot have any awareness (perceptual) of the water-body, if we strip it of all the sensible and other properties.” Perception, p. 246. But he then muddies the water, saying that “the designata of such mass terms … are cognitively or conceptually analysable into fundamental qualities or quality-atoms. When such qualities are taken away by analytic process, nothing substantial remains.” ibid., pp. 247–248. So he seems to see two distinct kinds of conceptual analysis in play. In support of the reading I prefer: Paul Williams, The Reflexive Nature of Awareness: a Tibetan Madhyamaka Defence. Richmond, Surrey: Curzon (1998), p. 117; “On the abhidharma ontology,” Journal of Indian Philosophy 9 (1981), pp. 227–257 at pp. 237–238. 41  Vasubandhu says of water, that in its common use, the term “water” designates (particular) colours and shapes (AK 1.13), that is, rūpa. Its uncommon use is, of course, to denote one of the four primary elements (mahābhūta). 39

100

J. Ganeri

first for the lute to be brought to him; but then he asks for the sound by itself. When he is told that it is the various parts of the lute together with the effort of the musician that give off the sound, the minister breaks the lute into pieces, splinters each of the pieces, burns all the splinters and scatters the ashes in a strong wind. “A poor thing indeed is this that they call a lute,” he says. Explaining the simile, the Buddha said: “So too, bhikkhus, as [a bhikku] investigates [the skandha], whatever notions of ‘I’ or ‘mine’ or ‘I am’ had occurred to him before no longer occur to him”.42 Without this explanation, the simile might seem to suggest that there is a relation of ontological dependence between the sound and the parts of the lute; but the Buddha’s interpretation makes of it rather a conceptual dependence: when one investigates a person into their component skandha, the concept of an “I” falls away. Vasubandhu’s statement that persons are ‘real with reference to conception’ is to be taken as saying that one can think in a person-involving way only as long as one does not analyse or ‘mentally divide’ the person into a flow of experience. As soon as one entertains the analysis, one no longer thinks in terms of a conceptual scheme that involves persons; one no longer sees the world that way. So Vasubandhu’s view is not that there are no persons, but that person-involving conceptual schemes are unstable, in the sense defined. Putting the matter this way would certainly explain why the Buddha is held to think that philosophical knowledge can lead to transformation – knowledge of the analytical reduction of persons results, if this claim is correct, in a reorientation of mind, a switch in perspective from a person-involving to an impersonal conceptual scheme. The Buddhists’ Nyāya critics urge that the Buddhist concept of a person is like the idea we form of a single continuous circle of light when we see a rapidly circling torch. They ascribe to the Buddhists the view that a person is the cognitive equivalent of an optical illusion. I think, however, that Vasubandhu’s understanding of the relationship between the impersonal and the person-involving conceptual schemes enables him to resist that pressure.43 I have already mentioned J. L. Austin’s observation that it is in general a mistake to treat reflections as if they are illusions, for, unless there is some devious mirror-work involved, they are perfectly normal parts of everyday experience. This is so even though reflections are conditional on facts about the observer, such as their position. Amartya Sen has argued, in a similar spirit, that adopting an objective view does not require that the interpreter seek to assume what Thomas Nagel has called the ‘view from nowhere.’ There is, he claims, a clear sense in which positional views  – views from somewhere  – achieve

 Saṃyutta Nikāya iv 197–8. The Connected Discourses of the Buddha, trans. Bhikkhu Bodhi (Boston: Wisdom Publications, 2000), p. 1254. 43  For a similar comment, see also John B. Buescher, Echoes from an Empty Sky, pp. 61–62. The view of Vasubandhu’s critical commentator Saṃghabhadra (c. 380 CE), author of the Nyāyānusāra, should also be considered. It seems that Saṃghabhadra was firmer about this than Vasubandhu himself. The Nyāyānusāra is extant only in Chinese; for discussion see Louis de la Vallée Poussin, “Documents d’Abhidharma,” Bulletin de l’école française d’extrême-orient, 30 (1930), pp.  247–298; Collett Cox, Disputed Dharmas: Early Buddhist Theories of Existence (Tokyo: International Institute for Buddhist Studies, 1995). 42

The Imperfect Reality of Persons

101

objectivity.44 To give a simple analogy, an observation that the sun and moon appear the same size is objective (that is, true for any observer, irrespective of subjective differences between observers), but positional (dependent on the observer’s having a certain position in space, on the earth’s surface). It would be quite wrong to describe the observation as an illusion. In other words, we need to be able to distinguish between positionality and subjectivity; not every perception or concept that depends on subject-specific parameters is to be assimilated to a subjective illusion.45 There is another reason to distance the idea of that which is ‘real with reference to conception’ from the idea of an illusion. It is well known that experience displays a certain resistance to, or ‘independence’ from, belief. The experience of an illusion persists even when it is known to be an illusion. In the Müller-Lyer illusion, for example, the experience of one line as longer than the other is not dispelled even when the subject measures the lines for themselves. Similarly, the effect of a trompe l’oeil can survive the discovery that it is a trick of clever paintwork; the same is true of half-immersed sticks seen as bent. This is what Gareth Evans has called the ‘belief-independence’ of informational states: “the subject’s being in an informational state is independent of whether or not he believes that the state is veridical.”46 If Vasubandhu is right in his analysis, however, then positionally objective ways of thinking differ from illusions in just this respect. The switch to a view from nowhere undermines positionally objective ways of thinking. Those ways of thinking are, nevertheless, real ways of thinking (ways of thinking about what is real); they are not merely subjective. They are real, but they are not ideal – the problem with these ways of thinking about the world is not that they are false, but that they are flawed, imperfect. The existence of much better ways to think, these Buddhists say, does not falsify the ways we do habitually think. The point is lost if we are tempted to look at the distinction between the ‘two truths’ after the fashion of Plato’s cave. The distinction between the ‘two truths’ is not a distinction between reality and mere appearance, or between fact and fiction. It is a distinction between better and worse but nevertheless objective ways of thinking.47 We are

 See Amartya Sen, “Positional objectivity,” Philosophy and Public Affairs 22 (1993), pp. 126–145.  There is a further distinction, within the category of the positionally objective, between positionally objective illusions, like the bent-stick or the mirage, and positionally objective non-illusory observations. This distinction would be drawn formally by certain Mādhyamika authors, who distinguish two varieties of conventional truth (saṃvṛti-satya), one ‘objective’ or ‘real’ (tathya), the other ‘erroneous’ (mithyā). See Malcolm Eckel, Jñānagarbha’s Commentary on the Distinction between the Two Truths: An Eighth Century Handbook of Madhyamaka Philosophy (Albany: State University of New York Press, 1987), pp. 75, 123 n. 41; C. W. Huntington, “The system of the two truths in the Prasannapadā and the Madhyamakāvatāra: a study in Mādhyamika soteriology,” Journal of Indian Philosophy 11 (1983), pp. 77–106. 46  Gareth Evans, The Varieties of Reference (Oxford: Clarendon Press, 1982), p. 123. 47  There are many other attempts to draw the distinction in the Indian Buddhist literature. Dharmakīrti (PV 2.3) provides a definition of the distinction that is in marked contrast with that of Vasubandhu (AK 6.4). He says that only the ultimately real is capable of causal influence, all the rest is merely saṃvṛti-sat. See also Mervyn Sprung ed., The Problem of Two Truths in Buddhism and Vedānta (Dordrecht: Reidel, 1973); John B. Buescher, Echoes from an Empty Sky, pp. 55–83. 44 45

102

J. Ganeri

not in error when we think of the world in a person-involving way; it is just that we could do better – ethically better – by thinking of it in some other way altogether, by standing in a different cognitive relationship with the world. One implication of the distinction is that the organisation of experience into persons is artificial rather than natural, that it has the character more of an achievement rather than a given, and that it is at least notionally within our power, and so within the orb of moral concern, not to organise and divide experience that way. Bringing persons within the realm of human achievement opens a possibility that is unavailable to the metaphysicians of self: that it might be better not to classify experience that way; indeed, that it might be better not to classify experience any way at all.

5 Two Problems for Vasubandhu’s Reductionism Criticisms of Reductionism in recent literature have concentrated on the problem of showing that the concept of a stream of experience is either conceptually or ontologically prior to that of a person. In this section I want to look at two problems that arose for Vasubandhu within the Indian literature, problems respectively to do with the notions of ownership and causal sequence. In the next section I will consider two arguments, again from the Indian literature, that there must be more to someone than a stream. An objection to any reduction of the idea of person to that of causally connected flow of experience appears to follow from considerations of ownership. Consider what is involved in the idea of ownership, specifically in the idea that we own our experiences. The question is whether facts about ownership can be reduced to facts about causal connection. It happens that Vasubandhu himself considers just such an objection to the Reductionist view. If a self does not exist, [they ask,] whose is this memory? [They say that] the meaning of the use of the genitive case [indicated by the use of “whose”] is ownership. It [the self] is the owner of a memory in the way that Caitra owns a cow. [In their view,] a cow cannot be used for milking or for carrying anything [by someone] and so on unless it is owned [by them].48

Vasubandhu’s way of dealing with the objection is less than convincing. He asks us to consider what is meant in saying that Caitra owns a cow, and his answer is that it is just a matter of causal connections between two streams, the stream that is Caitra and the stream that is the cow: [In your example,] what is called “Caitra” is called the owner of a cow because we are aware of a single continuum of a collection of [phenomena] causally conditioning [other] phenomena [within the same continuum] and assume a causal connection to the occurrence of changes of place of, and alterations in, the cow. But there is no one thing called “Caitra,” or a cow. Therefore there is no relation between the owner and what it owns over and above that between a cause and its effect.  Abhidharmakośabhāṣya, p. 1217; trans,. Duerlinger, Indian Theories of Persons, pp. 97–98; cf. Kapstein, Reason’s Traces, p. 368. 48

The Imperfect Reality of Persons

103

This rejoinder fails to meet the point of the objection: it is true that because Caitra owns the cow, it is he and nobody else who milks it and carries things with it; but these facts are not what his ownership of the cow consists in. Later Nyāya philosophers who analysed the notion of being owned (svatva) said that it consists in the capacity of the object to be used as one wishes (yatheṣṭhaviniyogayojyatva).49 But a capacity might go unexercised, so there may be no causal connections corresponding to it. What about the idea that we own our own experiences? I do not mean by this simply the idea that experiences must belong to an owner, i.e. a subject of experience. The view described by Strawson as the “no-ownership” view denies this, but Strawson rightly found it unintelligible.50 The view I have in mind is that I own my experiences in the way Caitra owns his cow, that I have capacities with regard to them that nobody else has, even if those capacities are never exercised.51 For example, I have the capacity to stop having one of my experiences, e.g. by shutting my eyes during a particularly gruesome episode in a film. The objection is that the Reductionist lacks the resources to make sense of this notion of capacity, right or entitlement. Nor is it easy to see how the Reductionist can explain a disanalogy between ownership of experiences and ownership of property, namely that I do not have the power of transfer with regard to the experiences I own. Another problem for Vasubandhu’s position52 is due to the later Vaiśeṣika philosopher, Śrīdhara (c. 990 CE). He says that if Reductionism were true, then we could form no conception of causal sequence. The claim is that the very possibility of having a notion of causal sequence – and this is of course a notion to which the Reductionist is absolutely committed – requires that Reductionism is false. [Buddhist:] As a result of there being a causal connection, a later memory [is a memory] of what was experienced at an earlier moment. The son does not, however, remember what was experienced by the father; this is because there is no causal connection between the cognitions of father and son, and their bodies, though admittedly so [connected], are not [themselves constituted of] consciousness.

 For a review of the later theory: J. Duncan M. Derrett, “An Indian contribution to the study of property,” Bulletin of the School of Oriental and African Studies 18 (1956), pp.  486–498; esp. pp.  481–483. The metaphysical status of property was the cause of a major controversy in the seventeenth century, between realists, following the precedent of Raghunātha Śiṛomaṇi, who held that it “had an objective reality of its own independently of a particular individual’s consciousness,” and anti-realists, who thought instead that it “was a special figment or condition of the brain, and that without consciousness of ownership … property did not exist.” (Derrett; the terminology is mine). According to the realists, the facts of ownership are irreducible: svatva is its own padārtha. According to the anti-realists, the facts of ownership are judgement-dependent. Ethan Kroll is editing and translating the principal textual sources. 50  Peter F. Strawson, Individuals: An Essay in Descriptive Metaphysics (London: Methuen, 1959), pp. 95–98. 51  On the phenomenology of ‘having’: Richard Zaner, The Problem of Embodiment (The Hague: Martinus Nijhoff, 1964), esp. 21–35; Gabriel Marcel, “Outlines of a phenomenology of having”, in Being and Having, trans. Katharine Farrer (London: Dacre Press, 1949); see also his Metaphysical Journal, trans. Bernard Wall (London: Rockliff, 1952). 52  For still further critical analysis: Oetke, “Ich” und das Ich, pp. 194–242. 49

104

J. Ganeri

[Śrīdhara:] This is not well-reasoned, for in the absence of self, there would be no determinate notion (niścaya) of a causal connection. At the time of the cause, the effect has yet to occur, and when its time comes, the cause has gone. Aside from the two of them, some unitary perceiver is denied; so who would observe the causal connection between those two things occurring in sequence?53

Śrīdhara’s argument is that in order to form the idea of causal sequence I must be able to experience sequences of events as sequences. But if Reductionism were true, there would be just flows of fleeting experience, and no one experience would be in a position to witness a drawn out sequence of events. In other words, if Reductionism were true, it would be inconceivable; but it is conceivable, therefore it is false. If Śrīdhara is right, then Hume is more consistent than Parfit, for Hume said both that persons are mere bundles of experience and that there are no real causal connections. If Śrīdhara is right, we must either be fully Humean, or else be fully Realist; there is no middle way.54

 Praśastapādabhāṣyam of Praśastapāda with the commentary Nyāyakaṇḍalī by Śrīdhara Bhaṭṭa, edited by Durgādhara Jhā (Varanasi: Sampurnanand Sanskrit University, 1997), p.  170. The Kashmiri Naiyāyika Jayanta Bhaṭṭa (c. 850–910 CE) argues similarly against the Buddhists on the grounds that there must be temporally extended but single perceptions; the argument is found in his play, the Āgamaḍambara 1.170. Csaba Dezső ed. and trans., Much Ado About Religion, (New York: New York University Press, 2005), p. 76. 54  Substantially the same argument is also to be found in one of Śrīdhara’s contemporaries, the Kashmiri Śaiva philosopher Rāmakaṇṭha (c. 950–1000 CE). See his Nareśvaraparīkṣāprakāśa, ed. M. K. Shastri (Srinagar: Kashmir Series of Texts and Studies, Vol. XLV, 1926). pp. 71–72. I am grateful to Christopher Bartley and Alex Watson for this information. See further Alex Watson, The Self’s Awareness of Itself: Bhaṭṭa Rāmakaṇṭha’s Arguments against the Buddhist Doctrine of No-Self (Vienna: Publications of the De Nobili Research Library, 2007), chapter 4. Ratnakīrti (c. 1070 CE), for the Buddhists, attempts a response in his Kṣaṇabhaṅgasiddhi: a momentary cognition represents itself as having been produced by another earlier one. See Haraprasad Shastri ed., Six Buddhist Nyāya Tracts in Sanskrit (Calcutta: The Asiatic Society, 1989), p. 32. For discussion: Satkari Mookerjee, The Buddhist Philosophy of Universal Flux (Delhi: Motilal Banarsidass, 1975), pp. 66–68. Ratnakīrti repeats the argument in another of his works, the Santānāntaraduṣaṇa, a work in which he is led to the conclusion that there are no apparent boundaries between streams, and so in fact no distinct streams at all. 53

Eliminating Selves, Reducing Persons Monima Chadha and Shaun Nichols

Mark Siderits has been one of the sharpest, clearest philosophers working on Buddhism in the last several decades. His work has also been strikingly wide-­ ranging. In this chapter, we will focus on two themes in his work that we find particularly interesting. First, Siderits makes a strong case that Abhidharma Buddhists promote mereological nihilism – the view that only simple entities are ultimately real, and aggregates (like a chariot or a heap) are at best useful fictions. Mereological nihilism grounds an argument against the existence of selves, which of course, is central to the tradition. But Siderits draws out a further, bracing conclusion from mereological nihilism – that consciousness isn’t ultimately real. We think this conclusion so antithetical to other commitments of Buddhism that it provides reason to doubt whether Buddhists should sustain the commitment to mereological nihilism. There is another important argument for the no-self view, however, Vasubandhu’s causal-efficacy argument; on this argument, real things must enter in cause-and-­ effect relations, and when we turn to cause-and-effect, there is no place where the self plays a necessary role. This argument informs our exploration of the second theme from Siderits – his defense of the “personalist” view that person (unlike self) is a useful fiction. We challenge this claim, first by noting that Vasubandhu himself uses the causal-efficacy argument against both selves and persons, and second by arguing that selves and persons are also on a par when it comes to whether they are useful fictions.

M. Chadha (*) Monash University, Department of Philosophy & Monash Centre for Consciousness and Contemplation Studies, Melbourne, VIC, Australia e-mail: [email protected] S. Nichols Cornell University, Sage School of Philosophy, Department of Philosophy, Ithaca, NY, USA © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_6

105

106

M. Chadha and S. Nichols

1 Mereological Nihilism and Two Truths The chariot provides what is perhaps the most famous example in Buddhist philosophy. As part of an argument that there is no self, Nagasena convinces King Milinda that the notion of a chariot is merely conventional – the chariot isn’t real. Here is part of the exchange: Nagasena: “Is the banner-staff the chariot?” Milinda: “Nay, verily, bhante.” Nagasena: “Is the yoke the chariot?” Milinda: “Nay, verily, bhante.” Nagasena: “Are the reins the chariot?” Milinda: “Nay, verily, bhante.” Nagasena: “Is the goading-stick the chariot?” Milinda: “Nay, verily, bhante.” Nagasena: “Pray, your majesty, are pole, axle, wheels, chariot-body, banner-staff, yoke, reins, and goad unitedly the chariot ?” Milinda: “Nay, verily, bhante.” At this last step, Milinda concedes immediately without protestation. But that seems altogether too quick. Of course the reins aren’t the chariot and the yoke isn’t the chariot. It’s even plausible that the parts of a chariot dispersed in a field does not make for a real entity. Indeed, it doesn’t make for a chariot! But when those parts are composed into a vehicle that conveys the king from place to place, at that point it seems that the chariot is indeed a real thing, not just a convention. Presumably there is a missing premise here. One premise that would do the required work is to claim that aggregates aren’t real. Obviously there is no thing that is both simple and the chariot. Siderits promotes this way of thinking about chariots as flowing from Abhidharma views, according to which, “the real is the concrete particular, and aggregation of particulars is the mark of the mental” (2014, 437). And this applies to the case of the chariot in particular: …a chariot is actually not a real thing. The parts are real, but the whole that is made up of those parts is not. The whole can be reduced to the parts, it isn’t anything over and above the parts. This is the view known as ‘mereological reductionism’. (2007)

More recently, Siderits dubs the view “mereological nihilism” (2015). Mereological nihilism seems incredible since it denies all partite things. Most of the concepts we use involve partite things: persons, planets, neurons, molecules, mountains, rivers, tables, computers, and so on. According to mereological nihilism, none of these things exist. Abhidharma Buddhists, according to Siderits, embrace a more thoroughgoing nihilism that advocates that there are no enduring substances at all (2015, 97). On this reading of the Abhidharma texts, the only things

Eliminating Selves, Reducing Persons

107

that are ultimately real are individual dharmas (elements or tropes).1 Substances, on this Abhidharma Buddhist view, are analysed into a bundle of tropes. So, Siderits claims, “there will ultimately be no elementary particles, nor will the dualist or idealist be able to claim that there are at least minds” (ibid.). Those who respond to mereological nihilism with a furrowed brow, will be certainly tempted to give the Abhidharma Buddhist philosopher an incredulous stare. (David Lewis enriched the technical vocabulary of the phenomenology of the reception of philosophical theories by the addition of the useful phrase “the incredulous stare” but at the same time he reminded us that an incredulous stare is not an argument against a philosophical theory.) To save the Abhidharma Buddhist against incredulous stares Siderits invokes the doctrine of “two truths”. According to Siderits, Abhidharma Buddhists acknowledge the reality of some composites in a “loose and popular” sense even while they maintain composites are not ultimately real. Siderits summarizes the Ābhidharmika position thus: The Abhidharma term for this ‘loose and popular’ sense is ‘conventionally existent’ (saṁvṛiti-sat); things that exist in this way are termed conceptual constructions (they are prajñapti-sat). Abhidharma thus posits a two-tier ontology: ultimately real entities (the dharmas) that are genuinely impartite or non-composite; plus the ontological back-­ benchers, those composite objects to which we express ontological commitment in our everyday speech and thought (2014, 437).

Armed with the two truths, the mereological nihilist, Siderits thinks, can explain why ‘I’ does not refer to any real entity, and also explain our ordinary thought and talk about composites. The Abhidharma Buddhist claims that strictly speaking chariots do not exist and that “chariot” is just a convenient designator for dharmas arranged chariot-wise. Given the demands of everyday life, it is just easier for us to talk about the existence of ordinary enduring middle-sized objects. The ease reflects our interests and cognitive limitations, but that may not reflect how things really are. Siderits notes that “to allow those interests and limitations to colour our representations about how things are can sometimes lead to cognitive error and erroneous beliefs can lead to harm” (2015, 98). Belief in the existence of pots and chariots cannot cause serious harm, but the tendency to posit ordinary objects as referents for our shorthand linguistic devices can lead us to think there is one concrete particular where there are in fact many concrete particulars arranged pot-wise or chariot-wise. In the case of ‘I’, however, these Buddhists maintain that this tendency does serious harm, for positing a referent leads to existential suffering.  There are significant differences among Buddhists themselves about the interpretation and the role of dharmas. In the canonical sūtra texts, dharma is used in the sense of “law,” “truth,” or “teaching”. This is in contrast to the various translations and meanings of dharma in Abhidharma Indian Buddhist philosophy, where it is glossed as factor, thing, element, constituent, phenomenon, event, datum, property, quality, fundamental existent, reality, or not infrequently left untranslated. In the early Abhidharma tradition the primary concern is soteriological, thus the dharma analysis focuses on cultivating wholesome dharmas and abandoning unwholesome ones. However, as the Abhidharma tradition evolves the emphasis shifts from categorisation to ontological status of the individual analytical products. For an informative discussion of the evolution of dharma in the early Abhidharma tradition see (Cox 2004). 1

108

M. Chadha and S. Nichols

Siderits distinguishes between two positions about the status of composites that the nihilist may adopt: Eliminativism and Reductionism. He spells out the distinction in the following way: Characterized semantically, then, non-reductionism about Ks will be the claim that Ks will be mentioned in our final theory about the ultimate nature of reality. Both reductionists and eliminativists deny this claim, but they disagree over whether continued talk of Ks will have any utility in the light of our final theory. The eliminativist, of course, proposes eliminating all talk of Ks, both in our final theory and in its ordinary- language adjuncts. It might be thought, then, that the reductionist sees matters thus: while the term “K” is in principle eliminable from our language (since we can give a complete description of reality without mentioning Ks), its continued use is both tolerable (because of translatability), and of some utility given our interests. (1997, 457)

Siderits’ terminology here might be a source of confusion. For in some sense, all Buddhists are eliminativists about the self. Moreover, on one typical use of “reductionism” in philosophy of science, a successful reduction is fully vindicating and not in the least eliminativist. If thermal properties like heat can be reduced to properties in statistical mechanics, that helps to secure thermal properties in science – thermodynamics and statistical mechanics describe the same thing. To say that heat is nothing “over and above” mean molecular kinetic energy is not to say that heat is nothing. Rather, it is to say what heat is. Importantly, if a neuroscientist claims that the self is nothing over and above the brain, this would not be enough to satisfy the Buddhist. The Buddhist wants to maintain that the self is nothing. We think that the distinction Siderits is drawing is better characterized in terms whether or not an Eliminativist view is Uncompromising. Uncompromising eliminativism about the self denies the existence of the self and doesn’t propose any replacement notion. There are various gentler alternatives, one of which is the idea that one can be an Eliminativist about a certain kind while also maintaining that it’s appropriate to continue talking about the kind as a useful fiction. The category hive mind might be an example of a useful fiction. A swarm of bees doesn’t have a mind, but it is often useful to speak as though it does. In this way, a Buddhist might maintain that even though chariots don’t exist, it’s useful to continue to use the language of chariots. Siderits argues that the Ābhidharmika have this gentler kind of eliminativism in mind, which motivates the distinction between the two truths. And once we have the two truths in place, Siderits argues: The analytic entailments holding between ‘there is a chariot’ and ‘there are these parts arranged in such-and-such a way’ reflect application conditions and reidentification conditions for ‘chariot’ that grow out of the customary linguistic practices of speakers of the language. Once these are fixed there are cases where it is true to say that there is a chariot. In such cases, it will also be true to say that there are these parts arranged in such-and-such a way. … The existence of chariots is vindicated. …This is the point Buddhist Reductionists make when they call the claim that there are chariots ‘conventionally true’: while chariots do not belong to our final ontology, in our daily lives we may and should continue to employ the folk theory according to which they exist. Since—given the way the world mind-independently is – our interests and cognitive limitations make the theory useful for creatures like us, the posit of chariots should be retained (2015, 111)

Eliminating Selves, Reducing Persons

109

Chariots are useful fictions, so we have the license to talk about them. By analogy persons are useful fictions too. What about selves? This chapter addresses this question.

2 Problems with Mereological Nihilism Siderits’ version of Buddhist reductionism has a striking conclusion – mereological nihilism--that the only things that are ultimately real are fundamental elementary tropes, the dharmas. All composites are at most useful fictions. Chariots aren’t dharma, nor are planets, organisms or even neurons. These are all useful fictions. These particular implications may not worry Abhidharma Buddhists, but there’s one implication that will be concerning. What about conscious experiences? This question is not an easy one to answer since the treatment of consciousness is equivocal in Abhidharma Buddhist literature. Consciousness in this literature is sometimes conceived of as a single dharma, or a conscious experience which is an assembly of dharmas, or one of the psychophysical aggregates (skandhas). Citta, which is usually translated as consciousness, sometimes signifies a single conscious dharma but is also used to signify what we would call an ordinary conscious experience. For example, envy (īrṣyā) is unwholesome consciousness (akusalacitta) which is consciousness accompanied by hatred (one of the unwholesome roots). Ordinary conscious experience of envy involves negative feelings (vedanā), intention (cetanā) to act in a certain way towards the envied object and so on, but these factors (cetasikas) are not ordinarily experienced as separate factors, they form the whole experience of feeling envious. Ronkin (2018) presents a model of the Abhidharma account of citta: The archetype of the operation of consciousness is citta as experienced in the process of sensory perception that, in Abhidharma (as in Buddhism in general), is deemed the paradigm of sentient experience. Citta can never be experienced as bare consciousness in its own origination moment, for consciousness is always intentional, directed to a particular object that is cognized by means of certain mental factors. Citta, therefore, always occurs associated with its appropriate cetasikas or mental factors that perform diverse functions and that emerge and cease together with it, having the same object (either sensuous or mental) and grounded in the same sense faculty. Any given consciousness moment—also signified by the very term citta—is thus a unique assemblage of citta and its associated mental factors such as feeling, conceptualization, volition, or attention, to name several of those required in any thought process. Each assemblage is conscious of just one object, arises for a brief instant and then falls away, followed by another citta combination that picks up a different object by means of its particular associated mental factors.

The key claim here is that a conscious dharma (e.g. a sensation arising from contact between a sensory organ and an object) can never be experienced in its own right. Ordinary conscious experience, on this view, is a unique assembly of various mental factors (feeling, discrimination, etc.) that arise and cease together with the sensation. Thus, although the mental factors (dharmas) constituting it are simultaneously present in the same temporal moment an ordinary conscious experience is not itself

110

M. Chadha and S. Nichols

a single dharma, (See Ganeri 2012, 127) and Chadha (2015, 548–9). I.  The Abhidharma Reductionist seems to be committed to the following inconsistent triad: 1. Everything that is a composite is unreal. 2. Conscious experiences are real. 3. Conscious experiences are composites. The Abhidharma Reductionist, according to Siderits, must accept 1 and so must reject either 2 or 3. It’s not always clear whether Siderits thinks Buddhists should reject 2 or 3. On the one hand, Siderits notes that mereological nihilism does not stand in the way of affirming the ultimate existence of conscious dharmas (2011, 314–5), which suggests rejecting 3. However, there is a problem with rejecting 3 because even though conscious dharmas are simple, they are not ordinarily experienced as simple dharmas. For example, although hiri is a simple conscious dharma, we do not experience it as such. Hiri is accompanied by other dharmas, positive feeling (vedanā) and intention to avoid evil acts (cetanā) and these dharmas are part and parcel of the experience of hiri, they are not experienced as separate dharmas. Ordinary conscious experiences are composites of multiple dharmas. Indeed, rejecting 3 also requires Siderits to reject that ordinary sensory perceptions, are ultimately real. For the Abhidharma, however, sensory perception is the paradigm of conscious experiences. The mere coming together of an object, sense faculty and its corresponding consciousness is not sufficient for a conscious experience to arise, it is always accompanied by universal mental factors, e.g., contact (sparśa), attention (manasikāra), feeling (vedanā), identification (samjñā̄), andintention (cetanā̄).The senses always process a steady stream of sensory impressions; some clusters of these impressions become a concrete object conscious experience only when attention is directed toward specific regions of the perceptual field. Attention, thus, is necessary but not sufficient for conscious experience and similar analysis can be offered for the other factors. Siderits can choose to focus on other examples from the tradition. But, the onus is on him to offer an alternative analysis of conscious experiences as simple dharmas on behalf of the Abhidharma which does not appeal to sensory perceptions or any other ordinary conscious experiences. On the other hand, if Siderits recommends rejecting 2, then the Buddhist must maintain that conscious experiences are at most useful fictions. Siderits seems not concerned by this implication. He suggests that Buddhist Reductionists do not have a good explanation of the phenomenal aspect of our conscious states.2 Siderits raises the following question: How can the Buddhist Reductionist explain the phenomenality of conscious experiences? He offers the following as a solution to Reductionist: By way of an abductive inference from the global availability of objects. Since global availability is just an aggregation of distinct causal pathways, it will come as no surprise that consciousness is a conceptual fiction, a single entity posited in order to simplify the task of data management. Just as it seems to us that there is a chair when the parts are assembled in a certain way, so it seems to us that there is the conscious state of seeing a chair in my path  Vasubandhu, unlike later philosophers, like Dignāga and Dharmakīrti, does not want to subscribe to the self-illumination or reflexivity thesis. 2

Eliminating Selves, Reducing Persons

111

to the door when the chair is made available not only to the action-guidance system (as in blind-sight) but also to the memory system, the speech system, etc. We can thus understand why it is that, despite there being no such thing as what-it-is-like-ness, it would seem to us that there is such a thing. Subjectivity is a useful fiction. (2011, 326–7)

Buddhist Reductionism thus implies that ultimately there is no subjectivity, and no conscious experiences. These are just useful fictions, like persons, pots and chariots. Siderits writes: The Buddhist Reductionist stance on consciousness and the self depends crucially on mereological reductionism, the view that the composite entities of our folk ontology are conceptual fictions, conventionally but not ultimately real. (2011, 315)

Siderits might not be worried by this implication, but we think that the Ābhidharmikas would want to backtrack and reconsider the commitments that led them to the conclusion that phenomenal conscious states and ordinary conscious experiences are just useful fictions. We think the Abhidharma Buddhists should consider rejecting 1. There are key elements of the Abhidharma tradition that run strongly against the idea that conscious experiences are merely useful fictions. Conscious mental states are responsible for guiding behaviour and action; for example, my conscious experience of hiri partly consists in my forming the intention to avoid evil acts and partly in feeling good about forming that intention and perhaps more. Such an experiential state is a mereological whole of many dharmas (hiri, vedanā, cetanā, etc.) but this whole is causally efficacious. Therefore, it must be regarded as real by the criterion for reality adopted by the Ābhidharmikas (Gold 2018). This criterion, as we shall see in the next section, is used by Vasubandhu to argue for the non-existence of the self and persons. Since conscious experiences are causally efficacious, they are good candidates for being real. Furthermore, Abhidharma is distinctive insofar as it aims to provide the theoretical counterpart to the Buddhist practice of meditation and, more broadly, a systematic account of conscious experience (Ronkin, SEP 2018). If conscious experiences are not ultimately real, it’s unclear why the Abhidharma should be so committed to providing a careful first-person description of conscious experience. Siderits’ version of Buddhist Reductionism might be taken to force the Abhidharma to its logical conclusion and deny the reality of conscious experiences, but we doubt that the Abhidharma would want to go that far. Rather, they may want to re-examine the analysis that has lead them to this conclusion and be cautious about endorsing a thoroughgoing mereological nihilism. If Buddhists don’t endorse mereological nihilism, this robs them of one central argument against the self. The chariot analogy, offered in the Visuddhimagga, is after all offered in service of the no-self view. But, as is well known, there are other arguments for this view in the Buddhist traditions. We want to draw attention to the argument in offered by Vasubandhu in the Abhidharmakośabhāṣya (Commentary on the Treasury of Abhidharma). The reason we want to draw attention to this argument is that Vasubandhu uses the same argumentative strategy and the same principle, i.e., the Causal Efficacy Principle, to argue against the existence of persons. This will put us in a position to address Siderits’ work on the notion of person as a useful fiction.

112

M. Chadha and S. Nichols

3 Vasubandhu Against Selves The final chapter of the Abhidharmakośabhāṣya offers a defence of the Buddhist doctrine of no-self. Vasubandhu’s master argument targets the Nyāya-Vaiśeṣika (Hindu) view of the self. The argument is based on the Causal Efficacy Principle: everything that is real or substantial (dravya) is causally efficient, having specifiable cause-and-effect relations with other entities (see, e.g., Gold 2018). Everything else is merely a conceptual construct, a mere convention (prajñapti) and thus should be rejected as unreal. The so-called “selves” must be one or the other— part of a causal series and thus real or conceptually constructed and thus unreal. Vasubandhu argues that selves are not causally efficacious. The master argument against the self can be reconstructed thus: S1: The self is presupposed to be a persisting conscious substance that exists independently of the psycho-physical aggregates. (Nyāya premise) S2: Everything that is real or substantial (dravya) is causally efficacious, having specifiable cause-and-effect relations with other entities (Abhidharmakośa-­ bhāsya, 130, 27; Gold 2018 SEP). S3: The self (as presupposed by the Naiyāykas) is causally inefficacious, since it does not have specifiable effects (Abhidharmakośa-bhāsya 472,16-478-13). C: Thus, there is no self (as presupposed by the Naiyāykas). Vasubandhu’s Nyāya opponent objects to S3 by claiming that such a self is implicated as the owner of experiences and the agent of actions. The Naiyāykas ask: How can we make sense of agents of physical actions and that of subject of experience and knowledge without there being a self? The list goes on: cognition, happiness and pain are qualities had by a substratum, what is the substratum of these qualities; who is the referent of the notion of “I”; who is the one who is happy or unhappy; and, finally who is the agent of karma and the enjoyer of the results of karma? None of these questions Vasubandhu argues, forces us to accept the existence of selves over and above the collections of psycho-physical atoms or aggregates. Each of these claims about agents of action, subjects of experience, knowers, and bearers of moral responsibility, etc. can be explained in ways that refer only to causal series of psychophysical elements (dharmas). For example, the Naiyāykas ask if there is no owner of experiences, whose is the memory? Vasubandhu’s response is to say that there is no owner of memories just as there is no owner of cognitions more generally. Perceptions arise on account of their appropriate causes—that is, sense-faculty, object, and attention. Vasubandhu claims that ownership is just a matter of causal relations between two streams. Regarding agency, Vasubandhu argues that there is no need to postulate an agent for bodily actions like eating, bathing, walking, etc. When we examine these activities closely, we can see that there is no need to infer a self as a cause. This is illustrated in the following exchange with a hypothetical Nyāya objector:

Eliminating Selves, Reducing Persons

113

Objector: There being no self, who is the doer of these deeds? And who is the enjoyer of their fruit? Response: What is the meaning of “doer” Objector: The doer is the one who does, the enjoyer is the one who enjoys. Response: Synonyms have been offered, not meaning. Objector: The grammarians say the this is the mark of the “doer”: “the doer is autonomous”. Response: Now whose autonomy is there with respect to what effects? Objector: In the world, Devadatta is seen with the respect to bathing, sitting, locomotion, etc. Response: But what is it that you term “Devadatta”? If the self, then that remains to be proven. Then the five bundles? In that case, those are the doer. Moreover, this action is threefold: action of body, speech, mind. There, with respect to the actions of body [speech], they are undertaken depending upon the mind. And those of mind are undertaken depending upon a proper cause with respect to body [and speech]. That being so in that case, nothing at all has autonomy. For all things coming about depending on conditions. …Therefore, no doer characterized [as autonomous] is apprehended. That which is the dominant cause of something is called its doer. But the self is not found to be a cause, with respect to anything. Therefore, there cannot thus be a self. For from recollection there is interest; from interest consideration; from consideration willful effort; from willful effort vital energy; and from that, action. So, what does the self do here? [Abhidharmakośabhāṣya, Trans. adapted from Kapstein 2001, 373]. A self contributes nothing to the arising of an action, for the desire to eat, say a mango, arises from a memory of enjoying a mango in the past, from this desire arises a consideration as to how to satisfy this desire, and from this consideration arises an intention to move the body for the sake of satisfying the desire and then from this movement, say of the hand to acquire and cut a mango, which finally leads to the action of eating a mango. There is no need to invoke the self as an owner of experiences or as an agent of actions at any point in this explanation. For the Abhidharma Buddhist the self is an ontological dangler without a causal role or an explanation. Vasubandhu says that by the very fact that we cannot apprehend the capacity of the self, any more than a capacity of the various chants uttered by a quack doctor (e.g., Phut! Svaha!), when it is established that the effect has been brought about by the use of certain herbs, we must conclude that the hypothesis of the self is problematic (Pruden 1988, 1347). The point of these explanations is not just that there is a better alternative explanation of phenomena like memory and action, but that these alternative explanations show there is no need to postulate or infer a self to explain these phenomena.

114

M. Chadha and S. Nichols

4 Vasubandhu Against Persons The principle of causal-efficacy discussed in the last section is also used to argue against the Vātsīputrīyas, also called Pudgalavādins or “Personalists”  – self-­ described Buddhists who deny the existence of the “self” but accept the reality of a so-called “person” (pudgala). Vasubandhu begins by raising a dilemma for the Personalists. The person is either caused or uncaused. As Buddhists, Personalists must reject the second horn of the dilemma. Everything in the Buddhist’s universe obeys the law of dependent origination. Persons cannot be uncaused, since nothing in the Buddhist universe is without a cause (Pruden 1988, 1314). So, the Personalists must opt for the other horn of the dilemma and specify the cause of persons. In response, the Personalists offer the analogy of fuel and fire. Persons “depend on” the aggregates just in the way that fire depends on the fuel, but is neither the same nor different from the fuel. Persons are not identical with the aggregates because the notion of person is of a thing that persists over time whereas aggregates change. If we grant the Personalists this much, they next need to maintain that persons are not distinct from aggregates. But the Abhidharma would reject this claim in light of the causal-efficacy argument. The Abhidharma argue that anything that persons do is in virtue of the properties of the parts. Fire is hot in virtue of the properties of the fuel-­ particles. So, they argue that the fire is not distinct from the fuel. The same applies to persons and aggregates. The only option, Vasubandhu insists, is to say that persons are not real, they exist in name only, they are mere conceptual constructions. The argument can be reconstructed thus: P1: The person is presupposed to be a persisting being that exists “depending on” the psycho-physical aggregates (Pudgalavādin premise) P2: Everything that is real or substantial (dravya) is causally efficacious, having specifiable cause-and-effect relations with other entities (Abhidharmakośa-­ bhāsya, 130, 27; Gold 2018 SEP). P3: The person (as presupposed by the Pudgalavādin) is causally inefficacious since it does not have specifiable effects (Abhidharmakośa-bhāsya, 461–472, 14). PC: Thus, there is no person (as presupposed by the Pudgalavādin). Vasubandhu’s opponents object to P3 by claiming that there are many scriptural texts and orthodox Buddhist positions expressed in those texts that imply the existence of persons. The Pudgalavādins argue that we need to postulate persons to explain moral responsibility, rebirth, and so forth: If there is no person how can we make sense of the talk of “bearer of the burden” (468,1–9). The questions continue. If there is no person, how can we make sense of claims that assert transmigration (471,24–472,3), and again what are we to make of Buddha’s claims that he remembered being this or that person in a previous existence (472,3–7). It should be emphasized that each of these points – about rebirth, memory, responsibility – are exactly the kinds of things that the self is posited to do in Indian philosophy. The Personalist seems to be trying to use the notion of person to salvage much of the

Eliminating Selves, Reducing Persons

115

practice that a more revolutionary Buddhist regards as arising from the dangerous notion of self. Vasubandhu, who is a revolutionary Buddhist, argues that none of the statements from the canonical Buddhist texts (about responsibility, rebirth, memory, soteriology, salvation and the nature of the liberated saint) is to be taken as referring to or implying that there are persons. Each of these can and should be interpreted so that it they refer only to aggregates (skandhas). In the discussion, the hypothetical Pudgalavādin opponent quotes the “sūtra of the burden-bearer” (Bhārahārasūtra) and asks if the aggregates were the person, how do we make sense of [the Buddha’s] talk of the burden bearer: Objector: You hold the bundles alone to be three-timed [belonging to past, present, and future], not the person; but if the bundles were the person, then why did [the Buddha] say this: O monks, I will teach the burden, the taking up of the burden, the casting off of the burden, and the bearer of the burden. Response   : Why shouldn’t this have been said? Objector   : Because the burden itself is not rightly the burden bearer. Response   : Why so? Objector   : Because this is not seen. Response   : Neither is it rightly ineffable. Objector      : Why so? Response: Because this is not seen. It is then implied that the one who takes up the burden is not subsumed by the aggregates. The Fortunate Lord taught the meaning of the burden-bearer: The one who is that long-living one, with such-and-such a name… up to… with such-and-such a longevity, with such-and-such an end to life. What he has made known thus, should not be known otherwise, i.e., as “permanent” or “ineffable”. The aggregates are painful in nature, thus they receive the name “burden”; the later aggregates are caused by the former aggregates; thus, the later ones receive the name the “bearer of the burden” [Abhidharmakośabhāṣya, Trans. adapted from Kapstein 2001, 360–1]. Vasubandhu’s strategy is to translate the talk of burdens and burden-bearer into a linear, causal series: it is because the former skandhas (the burden is the cause), torment the next ones (the burden-bearers are their effects), that they are called burden-­ bearers. The rebuttal of the Pudgalavādin’s objection is to say that the person (the burden-bearer) is no more independent from the skandhas (the burden) than is craving (the taking up of the burden). After offering similar explanations for karma, moral responsibility, rebirth and transmigration, Vasubandhu concludes that none of these phenomena offer any reason to postulate persons. The point of these explanations is not just that there is a better alternative explanation of the phenomena, but that these alternative explanations show there is no need to postulate or infer a person to explain these phenomena.

116

M. Chadha and S. Nichols

5 Selves and Persons Now that we have reviewed Vasubandhu’s arguments against selves and persons, we want to return to Siderits. Siderits would agree with Vasubandhu that persons don’t ultimately exist. Mereological nihilism suffices to generate that result. According to Siderits’ favoured version of Buddhist Reductionism, only dharmas exist ultimately. However, Siderits thinks there is a key role for appealing to persons as part of the conventional true. He writes: Early Buddhism and Abhidharma have often mistakenly been seen as Eliminativist. The error arises through attending solely to what is said at the ultimate level of truth, and failing to appreciate the relation between the ultimate and conventional levels of truth. Eliminativism is not simply the view that talk of persons may in principle be eliminated. Both Reductionist and Eliminativist maintain that ultimately there are no persons. But the Eliminativist urges in addition that the claim that there are persons be seen as conventionally false as well, since the Eliminativist maintains that our commonsense theory of persons is incoherent, or at least so misleading as to be more troubling and confusing than theoretically useful. By contrast, the Reductionist holds that while unquestioning adherence to the commonsense theory of persons does result in misguided views about how we should live our lives, the theory does have its uses, which fact requires explanation; hence it is conventionally true, though ultimately false, that there are persons. (1997, 465)

The foregoing gives the Buddhist Reductionists the license to use the expression ‘person’ in the language because it is useful for creatures like us to talk about persons. By labelling something as merely conventionally existent, we are claiming that it is not part of our fundamental ontology. It works as a convenient shorthand and thus is to be regarded as a useful fiction. Are persons useful fictions? We think that this question needs to be examined in light of the Buddhist views about selves more broadly. There are important differences between selves and persons; for instance, selves are simple and exist independently of the aggregates whereas persons are composites that exist dependently of the aggregates. Nonetheless selves and persons are afforded the same treatment by various classical Buddhist schools. As we saw, in Sects. 3 and 4, using the same strategy and the same principle Vasubandhu argues that persons and selves do not ultimately exist. Furthermore, as we saw in Sect. 4 persons and selves have the same job description. In addition, ordinary folk do not distinguish between self and person, perhaps because they have the same job description. Given that the notion of useful fiction is tied to folk usage of terms like persons and selves, we argue that folk usage does not distinguish between the two and therefore either both are useful fictions or none are. There are, no doubt, other grounds, to push for an asymmetry between selves and persons but not on grounds Siderits offers that one is a useful fiction but the other isn’t. There are numerous places in the sūtras where the Buddha does speak of “person”. Buddhists do not want to say that the Buddha was lying or unwittingly making false statements. (Siderits cites this as one of the reasons that the Buddhist must account for persons). Different schools in the Indian Buddhist tradition have different strategies for making sense of this thought and talk of persons. Both the

Eliminating Selves, Reducing Persons

117

Abhidharma as well as Madhyamaka traditions held that persons exist conventionally. There are some differences though. According to Madhyamakas the pragmatic usefulness of person talk is important as a catechetic device for teaching the doctrine to the uninitiated. All talk about persons and other continuing entities are preliminary and merely provisional teachings are meant to offer a transition between the adhesion to worldly beliefs and the doctrine of universal emptiness (Eltschinger 2014, 470). The Ābhidharmikas, on the other hand, did not think of “person” as a useful pragmatic device or skillful means but emphasized that it is nothing but a conventional designation, a mere name for the group of aggregates (the collections of dharmas). Persons are nothing over and above series of aggregates which in turn are collections of dharmas. No Buddhist wants to say that selves are ultimately real, but what about the conventional existence of selves? Classical Buddhist philosophers readily admit that selves too exist conventionally. Madhyamaka philosophers, for example, Nāgārjuna, would have no trouble speaking about selves as conventionally existent, as long as we do not think that this talk of selves is grounded in the ontologically significant substance-property distinction. It is nothing more than, as Westerhoff puts it, “a shifting coalition of psycho-physical elements” (SEP 2019). Just like persons. Abhidharma philosophers, like Vasubandhu, also admit that persons are conventionally real. In a discussion not often cited in the contemporary literature in Chapter 3 of the Abhidharmakośabhāṣya, Vasubandhu opens the discussion of dependent origination (pratītya-samutpāda) by raising an objection from the opponents who believe in self (ātman). The Buddhist belief in rebirth, according to the opponent, proves the existence of the self. Vasubandhu responds by saying that there is no self that takes up the aggregates at the beginning of life, abandons them at death, and takes on new aggregates in the next life. There is no internal agent of actions that abandons the aggregates at the end of this life and takes up new aggregates in the next one. The opponent then asks, is there any kind of self (ātman) that is you do not deny. The reply here is revealing: “We do not deny an ātman that exists through designation, an ātman that is only a name given to the aggregates” (Pruden 1988, 399). The point simply is that Vasubandhu thinks that selves exist conventionally, as long as we are clear that the term ‘self’ is just a name given to the aggregates. Just like persons. So, it seems that for these classical Buddhists selves have as much reality (conventionally speaking) as do persons. Next, we return to the question of useful fictions. As we saw, the Madhyamakas (for example, Nāgārjuna) regard both persons and selves as useful fictions, the Ābhidharmikas (for example, Vasubandhu) do not regard either selves or persons as useful fictions. Rather, persons and selves are mere names, convenient designators that are nothing more than a shorthand for the aggregates (and ultimately collections of dharmas). Again, it seems that for these philosophers either both persons and selves are useful fictions or neither. Siderits’ position differs from these classical Buddhist philosophers. He wants to say persons but not selves are useful fictions. In some sense, Siderits’ version of Buddhist Reductionism combines the Abhidharma and Madhyamaka strategy. Siderits says that the right way to think about the Buddhist Reductionist position is

118

M. Chadha and S. Nichols

to say that term “person” is both “tolerable (because of translatability), and of some utility given our interests” (1997, 457). How does Siderits justify drawing a wedge between the notions of selves and persons? Ordinary language doesn’t provide a promising approach. In ordinary language, “person” seems to be synonymous with “self”, or at any rate, there is no clearly marked semantic distinction in ordinary English. It is important to note that the target of the Buddhist no-self theory, in particular Abhidharma versions, is not just to reject some given theory of self in the Indian debate, but to diagnose a deep mistake in our ordinary primitive conceptual scheme (Ganeri 2012, 64). Ordinary ways of thinking about the self assume that the self, whatever it is, persists within a lifetime and possibly across lifetimes. And since psychological and physical properties do not persist over a lifetime, let alone beyond a lifetime, the self must be conceived of as independent of psychological and physical properties. What about the ordinary concept of person? Mostly, the concept of self and person are used interchangeably in the language, except that we hardly ever use the word ‘self’. We use reflexive pronouns like “myself” and “yourself” to claim identity with the thing that is referred to by the ‘I’ or ‘you’. What is that thing? A person. There is no settled common sense view about the nature of this self/person except that it undergirds our identity over time. In other words, our ordinary talk of “I” does not distinguish between ātman and person. Thus, it’s hard to see how we can use common sense as motivation for justifying that persons are a useful fiction, but selves aren’t. Furthermore, the standard Buddhist arguments for no-self derive from the fact of impermanence or momentariness (See Siderits 2011). But the notion of person posited by the Personalists seems to be caught up precisely in persistence beyond the momentary. Persons as postulated by Personalists continue to exist during a lifetime and possibly across lifetimes and they are bearers of karma, thus are persisting loci of agency and responsibility. Thus, the Personalist’s notion of person seems to share the very features that lead Buddhists to view selves with suspicion. Persons are just as ethically problematic as selves because they can be equally responsible for our sense of being the same thing within and across lifetimes. The central Buddhist teaching is to extirpate the false sense of ‘I’ and in that endeavour the conventional posit of person can’t be a useful fiction, it will be an obstacle. We think it’s useful to consider this issue in the context of Buddhist soteriology, which aims at escape from existential suffering. The core Buddhist idea is that we suffer because we mistake there to be a persisting being which is the referent of “I” and is sole the locus of meaning and value. The Buddhist insight is that there is no such persisting thing, everything that exists is transitory. If there is no persisting being, there is no locus of value and meaning, irrespective of whether we think it is a simple self (that exists independently of the aggregates) or a composite person (that exists dependently of the aggregates). This also grounds an argument for compassion. Ultimately, we are collections of physical and mental dharmas causally related to future and past dharmas within a series and with other series of dharmas. This way of looking at the world undermines the basis “me” and “mine” and “others” that are distinct from me. Thus there is no basis for discriminating between “my” suffering and that of “others”. The aim of the Buddhist teaching is to reduce

Eliminating Selves, Reducing Persons

119

all suffering, impersonally and selflessly. The self posit is to be rejected as it likely to distort this Buddhist vision, create artificial distinctions between “me” and “others” and thus increase suffering. It seems that the person posit will do the same, and should be rejected for the same reasons. Persons are fictions to be sure, but for the Buddhist they seem to be no more useful than selves. We’ve argued that in critical respects, selves and persons are on a par for Buddhists. Both selves and persons are causally inefficacious; thus, both are conceptual constructs and so ultimately not real. In addition, we’ve argued that, at least for Abhidharma Buddists, selves and persons seem to be equally problematic from a soteriological perspective: they are a result of flawed ways of thinking about the world and these ways of thinking create suffering. So, we think, contra Siderits, eliminativism about persons and selves is the right way to think about Abhidharma Philosophy. For the Buddhist, persons, like selves should be regarded as dangerous fictions, not useful ones.

References Chadha, M. 2015. Time-Series of Ephemeral Impressions: The Abhidharma-Buddhist View of Conscious Experience. Phenomenology and Cognitive Sciences 14 (4): 543–560. Ganeri, Jonardon. 2012. Buddhist No-Self: An Analysis and Critique. In Hindu and Buddhist Ideas in Dialogue: Self and No-Self, ed. Jonardon Ganeri, Irina Kuznetsova, and C.  Ramprasad, 63–76. London: Ashgate. Gold, Jonathan C. 2018. Vasubandhu. In The Stanford Encyclopedia of Philosophy, Summer 2018 edition, ed. Edward N.  Zalta. URL = https://plato.stanford.edu/archives/sum2018/entries/ vasubandhu/. Kapstein, Matthew. 2001. Reason’s Traces: Identity and Interpretation in Indian and Tibetan Buddhist Thought. Boston: Wisdom Publications. Pruden, L. 1988. Abhidharmakośabhāsyam. English translation of Poussin, Louis de la Vallée (1923–1931 [1980]), L’Abhidharmakośa de Vasubandhu, 6 vols. Bruxelles: Institut Belge des Hautes Études Chinoises. Berkeley: Asian Humanities Press. Ronkin, Noa. 2018. Abhidharma. In The Stanford Encyclopedia of Philosophy, Summer 2018 edition, ed. Edward N.  Zalta. URL = https://plato.stanford.edu/archives/sum2018/entries/ abhidharma/. Siderits, Mark. 1997. Buddhist Reductionism. Philosophy East and West 47 (4): 455–478. ———. 2007. Buddhism as Philosophy: An Introduction. Aldershot: Hackett. ———. 2011. Buddhas as Zombies: A Buddhist Reduction of Subjectivity. In Self, No Self?: Perspectives From Analytical, Phenomenological, and Indian Traditions, ed. Mark Siderits, Evan Thompson, and Dan Zahavi, 308–332. Oxford University Press. ———. 2014. Causation, ‘Humean’ Causation and Emptiness. Journal of Indian Philosophy 42 (4): 433–449. ———. 2015. Personal Identity and Buddhist Philosophy: Empty Persons. Aldershot/Burlington: Ashgate. Westerhoff, Jan Christoph. 2019. Nāgārjuna. In The Stanford Encyclopedia of Philosophy, Spring 2019 edition, ed. Edward N. Zalta. URL = https://plato.stanford.edu/archives/spr2019/entries/ Nāgārjuna/.

The Curious Case of the Conscious Corpse: A Medieval Buddhist Thought Experiment Robert H. Sharf

1 Preamble There is now considerable enthusiasm for dialogue between the fields of cognitive science, philosophy of mind, and Buddhism. Dozens of books and articles have appeared of late exploring areas of convergence, notably analogs between theories of embodied (aka extended, embedded, or enactive) cognition emerging from cognitive science and psychology, and Abhidharma and Yogācāra models of cognition drawn from medieval Buddhism. One early trailblazer in this field was The Embodied Mind: Cognitive Science and Human Experience, by Francisco Varela, Evan Thompson, and Eleanor Rosch (1991). Here a neuroscientist, a philosopher, and a cognitive psychologist team up to show, among other things, how Buddhist insights can help make sense of new data coming out of cognitive science. Varela worked with the Dalai Lama to establish the Mind & Life Institute, which would come to play a crucial role, both intellectually and financially, in the support and dissemination of research that spans Buddhism and Western science. And there are now several other organizations involved in fostering such research as well, including the Center for Healthy Minds at the University of Wisconsin-Madison, the Contemplative Sciences Center at the University of Virginia, and the Center for Compassion and Altruism Research and Education at Stanford University. The recent interest in Buddhist theories of mind is fueled, in part, by critical developments in the fields of neuroscience and cognitive psychology. Research suggests that cognition is distributed widely throughout unimaginably complex neural networks, parts of which communicate only indirectly with other parts. We have R. H. Sharf (*) University of California, Berkeley, Department of East Asian Languages and Cultures, Berkeley, CA, USA e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_7

121

122

R. H. Sharf

direct conscious access to only a small fraction of what is going on “under the hood,” and this can leave us mere bystanders when it comes to our own inner lives. Our experience of free will, agency, and self-possession is, at least in some respects, illusory, and the illusion is maintained by a natural propensity for confabulation—I unwittingly impose a narrative arc onto my fractured experience that mitigates cognitive dissonance, and this involves writing a singular and continuous but ultimately chimerical “me” into the storyline. Moreover, scientists claim to have discovered these things not through introspection and philosophical analysis but through empirical research. These findings (and many more) threaten the common-sense belief in a unitary and autonomous agent or self or cogito that perdures through or stands astride the moment-to-moment flux of experience. In place of this Cartesian picture, researchers now proffer models of embodied, embedded, and/or enactive cognition that approach consciousness as an emergent phenomenon distributed throughout a system that includes both the subject and object of experience. In short, these theories attempt to do away with an autonomous “self” or cogito. That the Buddhists were advancing sophisticated, non-self models of cognition some two millennia ago suggests, at least to some, that they may have something useful to contribute to the topic.1 Unsurprisingly, this interdisciplinary dialogue has provoked controversy. Many of the neuroscientists and psychologists working in the area are themselves Buddhist practitioners, and this raises questions about their motives and objectivity. At the same time, scholars of Buddhism complain that the scientists are perpetuating ill-­ informed and deracinated depictions of Buddhism. And there is some truth to this: Westerners, including scientists, are sometimes drawn to Buddhism out of the naïve conviction that, once stripped of its cultural accretions, Buddhism is a rational, atheistic, and eminently empirical science of the mind. In his 2008 book, Buddhism and Science: A Guide for the Perplexed, Donald Lopez chronicles earlier apologetic attempts, stretching back some 150 years, to cast Buddhism as scientific—attempts that now seem misconceived if not daft. Lopez believes that, given time and critical distance, the more recent efforts to present Buddhism as a science of mind will fare no better. That book was followed by The Scientific Buddha: His Short and Happy Life (2012), in which Lopez does more than ridicule the notion that Buddhism is compatible with science. Here he argues that approaching Buddhism as empirical and scientific forestalls the possibility of a more valuable engagement with its teachings, one that focuses on Buddhism’s trenchant critique of the philosophical and ethical assumptions that underly and sustain contemporary scientific research, including research in the areas of evolutionary biology, cognitive science, and

 It should be noted that the critique of a singular and perduring self has a long history in the West that predates the intervention of cognitive science. Its modern philosophical articulation is associated closely with Derek Parfit (1984), whose analysis of self is often compared with that of the Buddhists. 1

The Curious Case of the Conscious Corpse: A Medieval Buddhist Thought Experiment

123

psychology. In short, critics like Lopez believe that Buddhist materials are being read out of context, and that scientizing Buddhism runs the risk of dumbing down the science and denaturing the Buddhism. I too put little stock in the conceit that Buddhists solved, long ago, the major philosophical and existential problems that befuddle us today. I appreciate the dangers involved in wrenching Buddhist materials from their historical and cultural contexts, and agree that researchers in this area tend to be more critical in assessing the claims and authority of modern science than they are when assessing the claims and authority of Buddhism. Be that as it may, the critiques are typically predicated on assumptions about historical and cultural alterity—the notion that there is an unbridgeable gap between our modern, secular, scientific worldview, and the enchanted worldview of medieval Buddhists. And I believe that this assumption can be pushed too far. In my own forays into this area I have focused on some rather peculiar products of the medieval scholastic imagination, including (1) nirodha-samāpatti, which is a meditative state in which all cognition has temporarily ceased, leaving the yogi in what is tantamount to a vegetative coma; (2) the “beings without conception” (asaṃjñika-sattvāḥ), who are creatures in a supernal realm who have bodies but no minds; and (3) the indigenous Chinese Buddhist doctrine of the buddha-nature of insentient objects (wuqing foxing 無情佛性), which holds that objects such as grass and walls and roof tiles are inherently awakened. Such doctrines might be regarded as the musings of credulous exegetes with too much time on their hands, but this, I believe, is a hasty and uncharitable misreading. On closer look, these notions turn out to be sophisticated “thought experiments” (Gedankenexperiment) bearing on existential conundrums related to insentience, death, and the problem of thinking about non-conceptual experience. In short, the philosophical quandaries that motivated these medieval thought experiments are akin to those that continue to perplex philosophers today (Sharf 2014). Thought experiments are typically used to elicit an intuitive or even visceral response that will, it is hoped, throw light on a particular conceptual puzzle. Insofar as they aim to engage deep-seated or pre-reflective sentiments, they may better survive the journey across temporal, linguistic, and cultural distances than do other registers of philosophical discourse. Or at least, this is one thesis to be entertained below. The use of thought experiments goes back to the very beginnings of philosophy. If people recall only one thing from Plato, it is likely the allegory of prisoners chained in a cave able to see only shadows on the cave wall. Descartes’s “evil demon” (aka the “deceiving god”) is an equally memorable image, used, not unlike Plato’s cave, to distinguish truth from the way things merely seem. Today philosophers are more apt to draw their thought experiments from science fiction than from theology. In order to elicit and explore epistemological skepticism, philosophers now talk of brains-in-vats, or living in the Matrix or in a holodeck. To hone their thinking about mind-body duality, they contemplate brain transplants and

124

R. H. Sharf

teleportation. “Philosophical zombies”—beings like us in every way except for their lack of an inner life or subjective awareness—take center stage in discussions of the nature of mind and consciousness. The Turing test, which posits a machine that so perfectly emulates human responses that it seems sentient, is used to the same effect, as are the androids and cyborgs that populate films like Terminator, Bladerunner, Westworld, Ex Machina, and so on. (At what point should we consider such beings sentient or conscious? Is there a difference in kind between them and us, or is the difference one of degree?) Thought experiments are particularly popular in debates over qualia. Daniel Dennett proposed a “brainstorm machine” that wires the subjective experience of one person to another, while Ned Block conjures an “inverted earth”—a planet exactly like the earth except that the colors are reversed. In “Mary’s room,” Frank Jackson imagines a talented color scientist, Mary, who has spent her entire life mastering the neurology of color vision while living in a black and white room. (When she finally exits the room and experiences color for the first time, does she learn something new?) Another one of Block’s thought experiments—the “China brain”—envisions each and every person in China assuming the role of a single neuron such that, connected by walkie-talkies, they collectively simulate the activity of a single brain. (Could this collectivity be considered “conscious”?) These are, of course, wildly implausible and contrived scenarios, but that is precisely the point. Daniel Dennett calls them “intuition pumps”—they are intended to draw upon, fine-­ tune, and sometimes overturn our intuitions pertaining to difficult problems—problems of free-will and determinism, of self-identity and mind-body dualism, of the nature of consciousness, and of the epistemic warrants for our beliefs. Medieval Buddhist thought experiments reveal an interest in many of these same questions. This is not to say that the Buddhists resolved anything—Buddhists, like philosophers today, gave rise to a welter of competing arguments and positions. Rather, my point is that, despite contemporary academic fashions that foreground difference and alterity if not incommensurability, many of the underlying existential issues that preoccupy philosophers today are the same ones that preoccupied Buddhist philosophers centuries ago. And this claim finds support in the medieval Buddhist use of thought experiments that engage, in a matter immediately accessible to us, our embodiment—the jarring sense in which we inhabit, and at the same time are, our bodies. In claiming that Buddhist notions of nirodha, or the mindless gods, or the buddha-­ nature of roof tiles were intended as thought experiments, I open myself to the familiar criticism that I am merely finding new ways to wrench Buddhist dogmas out of context in order to make them seem sophisticated, topical, and philosophically salient. Some are wont to insist that the Buddhists took such notions literally, and that I am ignoring the cosmological, soteriological, and exegetical interests of the monks who wrote on these subjects. Which is why I was so excited when I came upon the tale I will discuss here—the man whose body was replaced with a corpse. I hope that this example will convince the skeptics. But to make sense of my reading of the corpse-man narrative, I need to retrace the steps that led me here.

The Curious Case of the Conscious Corpse: A Medieval Buddhist Thought Experiment

125

2 Zahavi’s Critique In an edited volume on the philosophical notion of no-self (Siderits et al. 2011), the phenomenologist Dan Zahavi contributed the only chapter explicitly critical of the Buddhist position. While he recognizes areas of overlap between Buddhist thought and contemporary currents in cognitive science and philosophy of mind, he is not convinced by the Buddhist doctrine of no-self (anātman). His critique, in short, is that the Buddhist construal of the ātman is so narrow as to render its denial of little philosophical import. That is to say: the Buddhist understanding of ātman as a metaphysical “soul” that is eternal, unchanging, and hence not subject to causation, is too easy a target, and thus the refutation of such an ātman contributes little to contemporary phenomenology. Zahavi’s alternative is what he calls a “minimal self,” which he believes is necessary to account for my first-person perspective—a perspective that is, in the end, mine if only insofar as it doesn’t belong to anyone else. In making his case he borrows Sartre’s notion of ipseity (selfhood, from the Latin ipse), as well as Husserl’s notion of Meinheit (mineness) to denote the irreducible and prereflective nature of consciousness experience—that which makes all experience appear to me rather than to someone else. He insists that this is not a quality or characteristic or discoverable content of experience. It is rather the “distinctive givenness or how of experience. It refers to the first-personal presence of experience. It refers to the fact the experiences I am living through are given differently (but not necessarily better) to me than to anybody else” (Zahavi 2011: 59). Zahavi’s “minimal self” is a sort of middle ground between postulating an unchanging soul substance on the one hand, or a “manifold of interrelated changing experiences” on the other. This minimal self is intended to foreground the “first personal self-­ givenness” of experience; in Zahavi’s words, “conscious mental states are given in a distinct manner, with a distinct subjective presence, to the subject whose mental states they are, a way that in principle is unavailable to others” (Zahavi 2011: 59–60). Zahavi’s critique of the Buddhists is thus simple: granted that the “self” I imagine is in some sense illusory, it is, nonetheless, my illusion, rooted in my sense of a body. Zahavi’s “minimal self” is intended to denote this unique, continuous, and ineliminable perspective. Zahavi is certainly not the first to raise this objection. The history of Indian Buddhist philosophy is, in part, the history of Buddhists defending their controversial if not confounding anātman doctrine against a host of objections, including ones similar if not identical to the one raised by Zahavi. Take, for example, the Nyāya philosopher Vātsyāyana and his later commentator Uddyotakara. They critique the Buddhist no-self theory as incapable of explaining either the synchronic or the diachronic unity of our phenomenal experience. (Synchronically, the anātman theory, at least in its early Ābhidharmika formulations, fails to explain cross-modal binding; diachronically, it fails to account for memory, personal identity, and other phenomena that require temporal continuity. See below.) In short, to the critics, the Buddhist claim that we are simply a series or continuum of causally linked

126

R. H. Sharf

momentary events with nothing that perdures from one instant to the next cannot accommodate the fact that I seem to have a unique perspective on my own continuum. Even some within the Buddhist fold found the anātman doctrine challenging. The Pudgalavādins, for example, raised concerns not unlike those of the Nyāya, and they proffered the notion of a pudgala or “person” to take up some of the slack. The pudgala, which is a sort of emergent entity that is neither identical to nor separate from the five aggregates, serves much the same explanatory role as does Zahavi’s minimal self: it accounts for my ineliminable sense that whatever happens in my mental stream happens to me and not to another.2 While positions analogous to Zahavi’s can already be found in the Indian tradition, I still found his articulation of the critique in phenomenological terms effective, and I felt that the efforts of others in the same volume to refute Zahavi, notably George Dreyfus, never fully hit the mark. Dreyfus defends the Buddhist position, arguing that while I may insist that I don’t believe in an eternal and unchanging soul, psychologically I still behave as if I do. And it is this reified psychological self, not a metaphysical soul, that is the source of clinging. But this doesn’t respond to Zahavi’s philosophical point, which is to give due attention to the fact that I don’t confuse my first-person experience with that of somebody else. (This position is known in philosophy as “immunity to error through misidentification.”) As touched on above, critiques of anātman generally proceed along two broad axes: one synchronic and one diachronic. The synchronic issue is known in cognitive psychology as the “binding problem.” How is information that emerges from different sense faculties and that is processed in widely distributed areas of the nervous system synthesized so as to create the unified phenomenal world of our experience? This was a serious problem for the Buddhists: if there is no overarching self, why do we seem to be singular selves rather than six parallel but independent selves, each associated with one of the six sense registers? Some Buddhist systems manage this by extending the functions of the sixth sense, namely mind. In Sarvāstivāda systems, for example, in addition to its role as an independent sense faculty with its own immaterial sense objects (dharmas), the mind faculty (manovijñāna) also manages the “impressions” (ākāra) transduced from the five material senses (Sharf 2018: 829–831). It is perhaps to clear up the resulting ambiguities that later Yogācāra theorists offload some of the cognitive burden carried by the mind faculty to the faculty of “mentation” (manas, aka “defiled mentation,” kliṣṭa-manas). It is the manas that ultimately binds together the six sense consciousnesses, creating the illusory impression of a unitary self, together with the psychological attachments and afflictions that invariably attend it. (Note that even in later Yogācāra there can be a blurring of the functions of the mano-vijñāna and the manas.) So it would seem that, despite their adherence to no-self, some Buddhist exegetes felt compelled to posit entities like the pudgala or the manas—entities whose function is similar to that of Zahavi’s “minimal self”—to account for cognitive binding.  There is a vast literature on the Buddhist anātman theory; for an insightful overview that includes a discussion of both the external (Nyāya) and internal (Pudgalavāda) critiques, see Kapstein (2001: 28–177), as well as the discussions in the chapters by Coseru and Ganeri in this volume. 2

The Curious Case of the Conscious Corpse: A Medieval Buddhist Thought Experiment

127

The problem with anātman is not synchronic binding alone. There is also the problem of diachronic identity, which is at the heart of Zahavi’s critique. Why do I not confuse my mind-body continuum with that of someone else? Why do my past and my memories seem to belong to me rather than to another? That Buddhists repeatedly felt compelled to respond to these issues suggests that the critique had traction.

3 The Corpse Man in the Da zhidu lun At the same time as I was reading Zahavi, I happened to be going through the “Perfection of Charity” (tan boluomi 檀波羅蜜, dāna-pāramitā) chapter of the Da zhidu lun 大智度論 (*Mahāprajñāpāramitopadeśa, “Great Treatise on the Perfection of Wisdom”) along with a visiting scholar, Stefano Zacchetti, and our graduate students.3 Among other things, the “Perfection of Charity” chapter contains an extended defense of the non-self doctrine. The interlocutor, in an attempt to refute the Buddhist anātman position, presents what I take to be the thrust of Zahavi’s critique. Question: By what means do we know that there is no self? All people give rise to the notion of a self with regard to their own bodies, but they don’t give rise to the notion of self with regard to other bodies. If there is no self within my own body, and yet I falsely perceive there to be a self, then as for the non-self of other bodies, I ought to give rise to the false perception of myself with regard to them as well. 問曰:何以識無我?一切人各於自身中生計我、不於他身中生我；若自身中無我、 而妄見為我者、他身中無我、亦應於他身而妄見為我. (T.1509: 25.148b11-14)

This would seem to be precisely the objection raised by Zahavi. And while I had encountered versions of it elsewhere, I might not have appreciated its cogency had I not been reading Zahavi at the time. In any case, without pausing, the interlocutor then raises the problem of how, given the anātman doctrine, the Buddhists account for the discrimination of things like colors. “Moreover, if internally there is no self, and if form and consciousness arise and pass away from moment to moment, how then do we discriminate between the colors green, yellow, red, and white? 復次、若內無我、色識念念生滅、云何 分別知是色青、黃、赤、白” (T.1509: 25.148b14-16). In other words, for perception and discrimination to occur, there must be some retention of what is not immediately available to sense experience, as well as a central “work space” or “holding area” capable of juxtaposing and evaluating what is present vis-à-vis what is not. This is related to the problem of memory, any solution to which might seem to require something functionally akin to Zahavi’s minimal self. Finally, the questioner raises a familiar critique of the anātman doctrine:

 A French translation can be found in Lamotte 1944–80: 2.735–750.

3

128

R. H. Sharf

Moreover, if there is no self, and an individual’s momentary awareness is constantly arising and passing away, then when one’s life comes to an end this momentary succession also ends. To whom then do all the sinful and meritorious actions (karma) follow and accrue? Who experiences pain or pleasure? Who is liberated? Given the various kinds of causes and conditions, we therefore know there is a self. 復次、若無我、今現在人識、漸漸生滅、身命斷時亦盡、諸行罪福、誰隨誰受? 誰受苦樂?誰解脫者?如是種種因緣故知有我. (T.1509: 25.148b16-19)

The first Buddhist rejoinder in the Da zhidu lun is clever but arguably sophistic: Answer: The difficulty redounds onto you. Were you to regard another body as yourself, then there would be even more reason to ask, “How is it that you give rise to the notion of self in a body that is not yours?” Because it arises from the causes and conditions of the five aggregates, it is empty and without self. The twenty views of personhood arise from the causes and conditions of ignorance. This belief in a self emerges from the continuous succession of the five aggregates. And because they emerge from these five aggregates, we construe these five aggregates as the self; they are not located in the body of another because of habitual conditioning. 答曰:此俱有難!若於他身生計我者、復當言「何以不自身中生計我」?復次、五 眾因緣生故空無我。從無明因緣生二十身見。是我見、自於五陰相續生。以從此五 眾緣生故、即計此五眾為我。不在他身、以其習故. (T.1509: 25.148b19-24)

The Buddhist then continues by focusing specifically on the issue of why we do not locate our own self within the bodies of others. Moreover, if you had a soul 神, it might then be possible to impute one’s own self to another. But you are not yet clear on whether you have a soul or not, so how can you ask about taking another as yourself. This is like someone asking you about the horns of a hare, and you reply that they are similar to the horns of a horse. If the horns of a horse really existed, then they might help us understand the horns of hares. But though we are even less clear about horns of horses, you still want to appeal to them to understand the horns of hares. Moreover, because one gives rise to the notion of a self within one’s own body, you claim there is a soul. You then claim that this soul pervades everywhere. But then you ought to consider the bodies of others as yourself! Therefore, you can’t claim: “Within my own body there arises what I consider to be a self, but in the bodies of others it does not arise, and therefore I know there is a soul.” Moreover, there are indeed people who give rise to a self with regard to other things. When non-Buddhists sit in meditation and engage in the “all-pervading earth-element contemplation,” they perceive the earth as if it were the self, and the self as if it were the earth. And they do the same with water, fire, wind, and space. As they are mixed up, they regard the bodies of others as themselves. 復次、若有神者、可有彼我。汝神有無未了、而問彼我!其猶人問兔角、答似馬 角。馬角若實有、可以證兔角；馬角猶尚未了、而欲以證兔角。復次、自於身生我 故、便自謂有神。汝言「神遍」、亦應計他身為我。以是故、不應言「自身中生計 我心、於他身不生、故知有神」。復次、有人於他物中我心生、如外道坐禪人、用 地一切入觀時、見地則是我、我則是地、水、火、風、空、亦如是。顛倒故、於他 身中亦計我. (T.1509: 25.148b24-c4)

We then come to the pièce de résistance, the case of the corpse-man, in which someone really does locate their self within the body of another. Moreover, there are instances in which the self arises with respect to another person. There once was a man who undertook a distant journey on assignment, and he spent a night alone

The Curious Case of the Conscious Corpse: A Medieval Buddhist Thought Experiment

129

in an empty hut. In the middle of the night a demon arrived carrying a corpse and placed it down in front [of the hut]. Then another demon arrived chasing after the first, and began shouting angrily at him: “That corpse is mine! How did you end up with it here?” The first demon replied: “It is mine! I brought it here myself!” The second demon said: “In fact I was the one who carried this corpse here.” The two demons each grabbed one hand [of the corpse] and struggled over it. The demon said: “Here is someone we can ask.” The second demon then asked the man: “Who brought this corpse here?” The man thought to himself: “These two demons are pretty strong. If I speak the truth I am certain to end up dead, but if I speak falsely I’ll also end up dead. In either case there is no escape, so what is the point in lying?” So he replied: “The first demon carried the corpse here.” The second demon was furious, grabbed the man’s arm, pulled it off, and threw it to the ground. The first demon took one arm from the corpse and affixed it back on the man. In this fashion, both arms, both legs, the head, flanks, and indeed the [man’s] entire body were changed [into the body of the corpse]. Thereupon the two demons joined in eating the [original] body of the man who had been transformed, wiped their mouths, and departed. The man thought to himself: “I just saw, with my own two eyes, two demons eat up my body—the body to which my mother gave birth. Do I actually now have a body or not? If I do, then it is wholly the body of another. If I don’t, then what is this body?” Thinking like this made him horribly confused, like a madman. The next morning he found the road and departed, going until he reached the previous country. There he saw a Buddhist temple and a group of monks. He was obsessed, and asked repeatedly whether his body existed or not. The monks asked, “What kind of human are you?” He answered: “I don’t even know if I am human or not.” The monks discussed the affair at length, and they all concurred: “This person knows non-self; it should be easy for him to attain liberation.” And so they said to him: “Your body, from the very beginning to the present time, has always been without self. It is not something that just happened now. It is only the coalescence of the four elements that we consider to be ‘my body.’ There is no difference between your original body and the one you have today.” These monks led him to the path, his defilements were extinguished, and he attained the stage of the arhat. Hence at times other bodies can also be taken to be one’s self, and thus you can’t take that as evidence that there is a self. 復次、有時於他身生我、如有一人、受使遠行、獨宿空舍。夜中有鬼擔一死人 來著其前、復有一鬼逐來瞋罵前鬼:「是死人是我物、汝何以擔來?」先鬼言:「是我 物、我自持來。」後鬼言:「是死人實我擔來!」二鬼各捉一手爭之。前鬼言:「此有 人可問。」後鬼即問:「是死人誰擔來?」是人思惟:「此二鬼力大、若實語亦當死、 若妄語亦當死、俱不免死、何為妄語?」語言:「前鬼擔來。」後鬼大瞋、捉人手拔 出著地、前鬼取死人一臂拊之即著。如是兩臂、兩腳、頭、脅、舉身皆易。於是二 鬼共食所易人身、拭口而去。其人思惟:「我人母生身、眼見二鬼食盡、今我此身 盡是他肉。我今定有身耶?為無身耶?若以為有、盡是他身；若以為無、今現有身。 」如是思惟、其心迷悶、譬如狂人。明朝尋路而去、到前國土、見有佛塔眾僧、不 論餘事、但問己身為有為無?諸比丘問:「汝是何人?」答言:「我亦不自知是人、非 人?」即為眾僧廣說上事。諸比丘言:「此人自知無我、易可得度。」而語之言:「汝 身從本已來、恒自無我、非適今也。但以四大和合故、計為我身、如汝本身、與今 無異。」諸比丘度之為道、斷諸煩惱、即得阿羅漢。是為有時他身亦計為我。不可 以有彼此故謂有我. (T: 1509: 25.148c5-28)

Among other things, this fanciful tale illustrates the point made at the beginning of the passage, namely, that even if one did in fact suddenly experience oneself in the body of another, rather than constituting evidence of a perduring self, it would actually undermine it!

130

R. H. Sharf

We could pause here to consider the strengths and weaknesses of the Da zhidu lun defense of the anātman doctrine.4 But as so much ink has been spilled on the philosophical coherence of the Buddhist anātman theory already, I would prefer to move directly to the issue at hand: Did the author(s) of the Da zhidu lun (whomever they may have been) take the corpse-man tale literally—as historical evidence of an actual body swap? Or was it intended as a somewhat playful thought experiment? The answer, I believe, can be found in looking at their sources.

4 The Genealogy of the Corpse-man Tale The Da zhidu lun, traditionally considered an Indic commentary on the Prajñāpāramitā-sūtra by Nāgārjuna, was rendered into Chinese by Kumārajīva between 402 and 406. The provenance, dating, and authorship of the Indic original remains an open question, but the current consensus among scholars is that the attribution to Nāgārjuna cannot be trusted, and that while the Chinese text is based on an Indian commentary, it is replete with interpolations and glosses that were likely introduced by Kumārajīva and his translation team.5 Nevertheless, the section to which the corpse-man story belongs bears all the marks of Indian philosophical disputation, and there is no reason to doubt its Indic provenance. We are fortunate to have access to several alternative versions of the corpse-man tale, notably those found in two Chinese translations of the “Biography of King Aśoka” (Aśokāvadāna, Aśokarājāvadāna). The textual history of this avadāna is complex, and the Chinese versions contain considerable material not found in the surviving Sanskrit text which is transmitted as part of the Divyāvadāna. The corpse-­ man tale is among the material not found in the Divyāvadāna version; nor is it found in the legends of King Aśoka from other Sanskrit or Pali sources. Another complicating factor is dating: scholars generally place the composition/compilation of the Aśokāvadāna to the second or third century A.D., but for reasons outlined below this date turns out to be problematic. Be that as it may, evidence suggests that the corpse-man tale that survives in the first Chinese translation of the Aśokāvadāna precedes, and represents a likely source for, the version found in the Da zhidu lun.6 The first translation—the “Biography of King Aśoka” (Ayuwang zhuan 阿育王 傳, T.2042)—was supposedly completed in 306 by An Faqin 安法欽 during the

 For a brief reflection on its philosophical import, see Ganeri (2007: 212–215), and Ganeri (2012:115–117). 5  See esp. the extended discussion in Lamotte 1944–80: 3.viii–xliv, who concludes that the original text likely dates to the early fourth century, as well as the convenient summary in Zacchetti (2015: 190–191). 6  On the various versions of the Aśokāvadāna see Deeg (2009: 12); Lamotte (1958: 261–272); Mukhopadhyaya (1963); Palumbo (2010: 20); Przyluski (1923); Strong (1983); and Wille (2000). Translations of the corpse-man tale from the Chinese versions can be found in Chavannes (1911:72–77); Li (1993: 164–165); Przyluski (1923: 381–382); and Strong (1992: 128–129). 4

The Curious Case of the Conscious Corpse: A Medieval Buddhist Thought Experiment

131

Western Qin Dynasty. However, Antonello Palumbo has disputed this attribution, which he argues is based on the much later and notoriously unreliable Lidai sanbao ji 歷代三寶紀 of 598. On the basis of his analysis of early catalogues, along with internal evidence, Palumbo believes the translation was actually produced around the time of Kumārajīva, that is, at roughly the same time as the Da zhidu lun.7 The second translation, the “Scripture of King Aśoka” (Ayuwang jing 阿育王經, T.2043), was completed in 512 by Sengqiepoluo 僧伽婆羅 (Saṃghabhara? Saṃghavara? Saṃghavarman?, 460–524). In both translations, the corpse-man story figures in a series of anecdotes related to Aśoka’s teacher, the famous Buddhist master Upagupta. Each of the short episodes centers on Upagupta’s pedagogical gifts—his ability to identify and help the disciple overcome the specific impediment that prevents him from making headway. The impediments include lust, gluttony, greed, lethargy, pride, and so on, and in each case Upagupta devises a clever ploy, often utilizing his supernatural powers, to help the student quell his affliction and attain liberation.8 The corpse-man story appears in the case of a disciple whose overwhelming obstruction is his vanity. The narrative in the first Chinese translation, the Ayuwang zhuan, runs as follows. In the Kingdom of Mathura there lived the son of a noble family, who, taking leave of his father and mother, wished to go to Reverend Upagupta to ordain [as a monk]. Although he left lay life, he was still very vain and remained attached to his body, and so he wished to return home. So he went to the Reverend to take his leave so as to return home. The Reverend said: “Then you should depart tomorrow.” The next day the man paid obeisance to the feet of the Reverend and took his leave. Midway on his journey he saw a shrine and had [second] thoughts about it: “If I return home, my mother and father will make a big deal out of it, so it would be better were I to spend the night in this shrine and then return directly to the Reverend tomorrow. That night the Reverend magically produced an ogre (yakṣa) who arrived [at the shrine] carrying a corpse. He then [produced] another ogre who arrived empty-handed. The two demons started arguing. The first said: “I came carrying this corpse.” The second said: “I came carrying the corpse.” The former demon said: “I have a witness. This man saw me arrive carrying the corpse.” At this point the man thought to himself: “Today I’ll certainly die, so I might as well just tell the truth.” So he said to the second demon: “This corpse was brought here by the first demon, so it is not yours to claim.” The second demon flew into a rage and yanked off the man’s shoulder. The first demon then took the shoulder from the corpse and affixed it to the man as it was before. The second demon then yanked off one arm. The first demon then took an arm from the corpse and affixed it to the man as it was before. The second demon then yanked off his two legs. The first demon used the legs of the corpse to restore him as before. Then the two demons joined in eating the fresh pieces of meat that had been yanked from the man, and summarily departed. Thereupon, the man’s attachment to his body completely vanished. Afterwards, arriving at the Reverend’s resi Palumbo’s argument is based in part on (1) the dates for the dissemination of the Aśoka legend in China, which is not attested until the late fourth century; (2) the absence of any mention of the text in the more reliable Chu sanzang ji ji 出三藏記集 of 515; (3) the fact that the Lidai sanbao ji attribution appears to be drawn from a lost catalogue, the Jin shi zalu 晉世雜錄, which is itself a likely forgery; and (4) the translation terminology of the Ayuwang zhuan, which is typical of that which became standard only after the circulation of Kumārajīva’s translations in the early fifth century (Palumbo 2010: 20 n.31). 8  For a full discussion see esp. Strong (1992:118–144). 7

132

R. H. Sharf

dence he completed his monastic ordination, and receiving [Upagupta’s] instructions in the dharma he attained the stage of an arhat. He then took a counting stick and placed it in the cave.9 摩突羅國有一族姓子、辭父母欲向尊者鞠多所求欲出家。既出家已、極愛著身 故復欲還家。便往尊者所辭欲還家。尊者語言:「且往明日。」明日禮尊者足即欲 還去。道中見天寺而作是念:「若還向家、父母或能為我作大事。不如即住此天寺 宿、明日當還詣尊者所。」尊者即夜化作一夜叉擔死人來。更有一夜叉空手而來。 二鬼共諍。一言:「我擔死人來。」第二者言:「我擔死人來。」前一鬼言:「我有證 人。此人見我擔死人來。」時此人念言:「我今畢定死竟應作實語。」語後鬼言:「 此死人者前鬼擔來、非是汝許。」後鬼大瞋拔其一髆、前鬼以死人髆還續如故。後 鬼復拔一臂、前鬼更拔死人臂還復補處。後拔其兩腳、前鬼悉以彼死人腳補之如 本。如是二鬼共食所拔新肉即時出去。於是愛身之心即便都滅。後至尊者所度使出 家、為說法要得阿羅漢。便令便令擲籌著於窟中. (T.2042: 50.123b3-20)

There are a number of conspicuous differences between this tale and the one found in the Da zhidu lun. Recall that the Da zhidu lun provides nothing by way of a backstory. We are told only that a man on an unspecified journey encounters the demons while spending the night at an empty road-side hut. But in the Aśokāvadāna version, we learn that the man in question is a disciple of Upagupta, and there is nothing accidental about his encounter with the demons—the demons are conjured intentionally by Upagupta to teach the young man a lesson. But the most striking difference, for our purposes, is that in the Da zhidu lun story the man’s body is replaced entirely with that of the corpse, while in the Aśokāvadāna, it is only a shoulder, an arm, and two legs that are replaced. This apparently is all that is needed to cure the student of his delight in his body. (Who wouldn’t be cured!) As such, this tale is not about an existential identity crisis but rather about overcoming vanity. Note also how the shift in context alters the significance of the demons devouring the man’s flesh. In the Aśokāvadāna version, seeing the demons feasting on his severed limbs serves to heighten the novice’s disgust for his own flesh, which just a moment before was a source of pride. In the Da zhidu lun tale, the image of the two demons feasting on the poor man’s entire body intensifies the existential confusion—the odd sensation of watching from afar as one’s body is being devoured. Moving now to the 512 translation of the Aśokāvadāna, we find some curious variances that, despite its later date, warrant our attention. In the Kingdom of Mathura there was a young man who sought permission from his parents to leave home as a monk. He traveled to Upagupta’s place, and upon arriving paid obeisance to his feet and said, “Venerable: may I have permission to leave home to be a monk and receive full ordination in accord with the Buddha’s teachings? I want to practice chastity in accordance with the teachings of the World Honored One.” Upagupta discerned that the young man was still bound by attachment to his body. So he said: “Welcome. I will bestow the precepts on you.” The man, having heard this, paid obeisance to his feet, but wanted [first?] to return to his family. Along the way he had this thought: “Should I return home, I may encounter obstacles and be unable to become a monk.” Alongside the road there was a shrine where he decided to spend the night. Upagupta used his supernatural powers to create two ogres (rākṣasa). One entered into the shrine carrying a dead corpse, while the other

 On the placing of the counting stick (śalākā) into the cave, which occurs whenever one of Upagupta’s disciples becomes an arhat, see Strong (1992: 141–143). 9

The Curious Case of the Conscious Corpse: A Medieval Buddhist Thought Experiment

133

arrived empty-handed. When they entered the shrine they began to argue over the corpse. One said: “I got the corpse!” The other said: “I got the corpse!” They argued in this way among themselves, and being unable to agree, they asked the man: “Which one of us came to this shrine bearing the corpse?” The man thought to himself: “If I tell the truth the ogre who came empty-handed will certainly kill me. But then if I don’t tell the truth, the ogre who arrived carrying the corpse will most certainly kill me. In either case I end up dead, so I ought not lie. He then said to the ogres [pointing to the first one]: “It was that one who came with the corpse.” Thereupon the ogre who came empty-handed pulled at [the man’s] arm wanting to eat it. Then the ogre who came with the corpse started pulling as well in an attempt to free him. Then [the empty-handed ogre] pulled at his foot wanting to eat it. The ogre with the corpse started pulling as well in an attempt to free him. This went on for a long time until the sun came up. After two days passed he went back to Upagupta’s place, and arriving there he became a monk. He fervently cultivated the way and attained the stage of an arhat. He then took a counting stick and deposited it in the cave. 摩偷羅國有一男子、啟其父母求欲出家。往優波笈多處。至已禮足、白言:「大 德、我得佛法中出家作比丘受具足不。我欲於世尊法中修行梵行。」優波笈多見其 於身為愛所縛、語言:「善來。我當與汝出家。」其人聞已禮長老足欲還其家。即 於中路作是思惟:「我若至家或有留難不得出家。」於其路中有一神廟便在中宿。 優波笈多即以神力作二羅剎。一持死尸入於廟中。一則空往。既入廟已、共諍死 尸。一言:「我得此尸。」一人言:「我得此尸。」於是二羅剎互共相諍。既不自決 而問此人:「誰將此尸來入廟耶?」此人思惟:「若我實言彼空來者必當殺我。若不實 語將尸來者復應殺我。乃可受死、不應妄語。」即語鬼言:「是彼將來。」時空來 鬼即牽其臂而欲食之。將尸鬼者助其牽掣令得免脫。又牽其脚而欲食之。將尸鬼者 復助牽掣令得免脫。如此良久遂至日出。經二日後往優波笈多處。至已為其出家、 精進修道即得阿羅漢果。乃至取籌置石室中.10

There are a number of differences between the two Chinese renderings of the Aśokāvadāna, including the fact that, in the later text, the young man had not yet ordained when he encountered the demons. But the most notable change is that, in the later version, the second demon doesn’t succeed in yanking off and devouring the man’s limbs. Rather, the two demons engage in a tug of war with the man’s body, with one trying to tear it apart so as to eat it, and the other trying to set him free. This terrifying ordeal goes on throughout the night, and in the end the demons leave without eating anything at all. Something seems a bit odd with this later rendition. For one thing, it lacks much of the punch of the two earlier versions. Gone is the horror of imagining what it would be like to witness one’s body parts exchanged with those of a corpse, and then to watch as demons devour one’s flesh—ordeals that, in the early Aśokāvadāna translation, are concocted by Upagupta specifically to target the man’s vanity. Note also how, in the Da zhidu lun and the Ayuwang zhuan, the corpse itself plays a critical role in the story, as it is needed to harvest replacements for the man’s limbs as they are torn off. In the later Ayuwang jing, however, the corpse is inconsequential—it is the source of the disagreement between the demons, but they could just as well be arguing over something else. It is difficult to know what to make of these changes, but one possibility is that it is the result of a later redactor trying to remove an incongruity. Note that, unlike the Da zhidu lun version, the demons in the Aśokāvadāna tale are imaginary beings conjured by Upagupta. One might then 10

 T.2043: 50.165b11-c2; cf. translation in Li (1993: 164–165).

134

R. H. Sharf

wonder how these imaginary creatures, bearing what is presumably an imaginary corpse, could effect a real and lasting limb transplant. So it is possible that a literal-­ minded editor replaced the limb replacement with an easier-to-swallow magical apparition of two demons engaged in a tug of war over the disciple’s body. In any case, the question at hand is the genealogical relation between the Da zhidu lun account and the earlier Chinese translation, both of which appear in China around the same time, and both of which involve the transplant of body parts. The Indian Aśokāvadāna has been dated to the second or third century A.D. by a number of scholars,11 but this is based in part on the received dating of the first Chinese translation, which, as we have seen, is problematic. We have also seen that the corpse-man tale appears in a linked series of anecdotes celebrating the prodigious pedagogical skills of Upagupta, and that this sequence does not appear in the extant Aśokāvadāna included in the Divyāvadāna. I would note, however, that Sanskrit fragments belonging to this sequence of tales have been found among some small palm-leaf fragments in the Schøyen collection, so there is little question that the corpse-man tale associated with Upagupta did in fact circulate in India.12 In terms of genre—a popular “Hīnayāna” avadāna versus a technical Mahāyāna sūtra commentary—one might assume that the Aśokāvadāna antedates the Da zhidu lun. But given the fluid nature of the content of the received texts, genre alone doesn’t help establish a relationship between the corpse-man tales that appear therein. Nevertheless, on the basis of content I would suggest that the corpse-man in the Ayuwang zhuan represents an earlier stage in the historical development of the tale. Note that it figures in a sequence of stories associated with Upagupta that are related by both structure and content. These stories, each of which illustrate Upagupta’s superlative teaching style, form a coherent, structurally and thematically related set, and the palm-leaf fragments found in the Schøyen collection attest to the circulation of this set in India. It seems plausible that the corpse-man tale found in the avadāna and associated with the Sarvāstivāda teacher Upagupta antedates, and was the inspiration for, the version found in the Mahāyāna scripture commentary. There is yet one more reason to believe that the version of the corpse-man in the Da zhidu lun descends from that in the Aśokāvadāna rather than vice versa. The two stories disagree on whether it is only the limbs of the man that are replaced by the limbs of the corpse, or his entire body. It seems most likely that the evolution would have been from the partial limb replacement to the more dramatic whole-body transformation, rather than the other way around. In sum, the available internal and external evidence suggests that the authors/compilers/editors of the Da zhidu lun recognized the philosophical potential of the Upagupta corpse-man tale which was already in circulation. Borrowing it from an earlier source, possibly the Aśokāvadāna, they deftly tweaked it to suit their purposes.

 See, for example, Mukhopadhyaya (1963: lx), and Strong (1983: 27).  Wille (2000). According to Sander (2000: 293–295), the fragments are written in a “Gilgit/ Bamiyan Type 1” script, which would suggest a date of no later than the seventh century. 11 12

The Curious Case of the Conscious Corpse: A Medieval Buddhist Thought Experiment

135

Before concluding, I should mention one more Buddhist tale in which we encounter two demons squabbling over a corpse. This is found in the Pali “Mahāsīlava-jātaka,” which is arguably earlier than, and possibly a distant ancestor to, the versions we have been discussing so far.13 As this is a long and complex tale, I will merely summarize the key points here. The benevolent king of Benares, Mahāsīlava (the Buddha in a former life), is an exceedingly righteous and compassionate monarch. One of his ministers (Devadatta in a former life) is exiled for misdeeds in the king’s harem, and ends up as chief minister to the king of Kosala. The treacherous minister convinces the king of Kosala that it would be easy to conquer Benares, since their king, Mahāsīlava, is a pushover who would refuse to fight. Indeed, when Kosala crosses the border with his army and challenges Mahāsīlava, Mahāsīlava, against the entreaties of his ministers, refuses to take up battle and surrenders himself along with his court. The king of Kosala then takes Mahāsīlava and his one-­ thousand ministers to a cemetery and buries them alive up to their necks, leaving them for the jackals to devour. Even then, Mahāsīlava does not object and harbors nothing but good will toward his captors. At midnight the jackals show up. Mahāsīlava manages to clutch the chief jackal firmly in his teeth, and this keeps the other jackals at bay. As the chief jackal struggles to escape Mahāsīlava’s hold, he stirs up the earth around the king allowing Mahāsīlava to break free. Mahāsīlava then frees the rest of his ministers. Mahāsīlava, still in the cemetery, then comes upon an exposed corpse along with two demons who are arguing over how to divide up the corpse so they can eat it. The demons, knowing that Mahāsīlava is a fair and righteous monarch, ask for his assistance in divvying up the corpse. Mahāsīlava agrees but only if the demons first help him bathe, eat, and recover his royal sword from the usurper king of Kosala. The demons oblige, using their magical powers to lavishly bathe Mahāsīlava and bedeck him with unguents, flowers, and fine clothes, feast him on fine food, and retrieve his sword. Mahāsīlava uses the sword to cut the corpse in half, and gives each demon his fair share. The demons, satiated and happy, go on to help Mahāsīlava recover his kingdom.

There is little linking the demons arguing over a corpse in the Mahāsīlava-jātaka to the Aśokāvadāna or the Da zhidu lun. That demons hang about in cemeteries and abandoned shrines and bicker over corpses may have been a stock motif in Indian folklore. But it does show that the motif was known to early Buddhist storytellers, and had gained a toehold in an important Buddhist paracanonical collection that may predate the texts discussed above.

5 Final Thoughts There are in fact countless Indian folk tales (kathā) featuring demons, powerful yogis, and corpses that playfully engage issues of self-identity and embodiment. Most notable are tales about “vampires” (vetāla) that are able to possess and

 Fausbøll 1877: 261–268 (jātaka 51 and its commentary); translation in Chalmers (1895: 128–133). There is some evidence that portions of this story, specifically the incident with the jackals, may be related to northern European folklore (Tawney 1883: 120–121). My thanks to Sean Kerr for bringing this jātaka to my attention. 13

136

R. H. Sharf

animate corpses as well as living persons. The Vetālapañcaviṃśati (“Twenty-five Tales of a Vampire”) collection includes an arresting story about a woman who mistakenly swapped the heads of her husband and her brother; the question is then posed: which one should she now consider her husband?14 There are also many stories of yogis who use their supernatural powers to enter and control the bodies of others, much as do the vampires (Bloomfield 1917). Medieval Indian literature provides abundant imaginative grist for the philosopher’s mill. Our own investigation has focused more narrowly on Buddhist tales involving demons arguing over a corpse. In one—the Mahāsīlava-jātaka—the demons seem to be included largely for their entertainment value. They provide a bit of comic relief following the burial ordeal of king Mahāsīlava, and they drive the plot line forward by allowing the king, who had already forsworn his warriors, to recover his kingdom. The authors of the later versions will tap the bickering-demon motif for its moral and/or philosophical potential. The Aśokāvadāna version is associated with the wise and canny Upagupta, who uses the quarrelsome demons to transplant the limbs of a corpse onto his disciple as a kind of shock therapy. This, I believe, is the probable source for the form of the tale that finds its way into the Da zhidu lun. The author of this later version evidently felt free to refashion the tale to suit his own purposes, namely, his multi-pronged defense of the Buddhist no-self theory. More specifically, the author of the Da zhidu lun will use the demon story to refute the interlocutor’s claim that our inability to mistake ourselves for others speaks to the existence of a self. On the surface then, the Da zhidu lun introduces the corpse-man tale as evidence that it is indeed possible to mistake oneself for another. But that our author has brazenly doctored the narrative suggests that he was not personally invested in the historical veracity of the tale, and one must presume that his intended audience was no less sophisticated. The intent, rather, is to imagine, as a thought experiment, the psychological and philosophical consequences of finding oneself in someone else’s body. Doing so will presumably help one appreciate the Buddhist anātman theory—that, on close inspection, the “self” turns out to be an empty signifier, and that the misguided belief in a self arises from our tendency to project a self onto what is in fact a temporally discrete but causally connected sequence of cognitive events that have no necessary relationship with any particular material form. In short, the image of a person’s body being replaced piecemeal with that of a corpse is intended to undermine my sense of who and what I am by showing the degree to which my sense of myself as a perduring agent is rooted in my identification with my body. In this sense it recalls the use of teleportation and

 The best known version of the story is found in the twelfth book of the Kathāsaritsāgara compiled by Somadeva in the eleventh century; see the translation in Tawney (1880–84: 2:261–264). The tale was borrowed by Thomas Mann for his 1940 novella “Die vertauschten Köpfe.” Mann’s story, which is set in India, is dedicated to the Indologist Heinrich Zimmer, through whose writings Mann was introduced to the tale. The head-swapping story then returns to India via the play Hayavadana (1971) by the contemporary South Indian playwright Girish Karnad, who knew the story through Mann’s adaptation (Mahadevan 2002). Thanks to Phyllis Granoff for this reference. 14

The Curious Case of the Conscious Corpse: A Medieval Buddhist Thought Experiment

137

brain-­transplants in contemporary philosophical debates over self identity.15 But there is a key difference: both the teleporter and brain-transplant scenarios presume a close connection between mind and brain; the conceit is that the brain that is replicated or transplanted will continue to serve as physical support for the self, and this is what generates the quandary.16 For medieval Buddhists, however, neither the brain nor any other part of the physical body serves as the singular seat of consciousness, so pondering teleportation or a brain transplant would not elicit the intended effect. Buddhist Ābhidharmika metaphysics is closer to dual-substance ontology, wherein the mental continuum is ontologically distinct from, but (at least in the case of beings in the desire and form realms) still associated with, a material body. Mind or consciousness can and does move in toto from one body to another, as happens at death. (Even in the case of those Buddhists, such as the Sarvāstivāda, who believe in an “intermediate existence” [antarābhava] between death and rebirth, a subtle material [rūpa] body is still required during the interregnum; see Kritzer 2000.) How then might an Ābhidharmika induce the conceptual vertigo that we experience when we contemplate a brain transplant or teleportation? The tale of the vain novice provided the answer: the body-swap could be effected piecemeal. Following the lead of the Upagupta version, the author of the Da zhidu lun has the body transformed one bit at a time, such that the spacial locus of self-identity remains in place while everything around it changes. In this sense, the physicalism presumed in the corpse-man tale differs from that in the brain-transplant and teleportation scenarios, in which the continuity of one’s mental stream is integrally tied to the structural or isomorphic continuity of a physical substrate. The success of a thought experiment depends on how cleverly it draws upon our metaphysical intuitions—the very point of the experiment is to expose such intuitions to considered refection and analysis. We may be unclear about how a neurosurgeon would go about a brain transplant or how a teleporter actually works, and medieval Buddhist monks may have been equally fuzzy about how a demon would go about grafting a limb. In each case the audience might be skeptical that such a feat is even possible. But that is beside the point. The goal is to elicit the

 The use of teleportation in philosophical discussions of identity became popular following the discussion in Parfit (1984: 197 ff.), although he was not the first to come up with the idea. In Parfit’s version, a person travels to Mars via a machine on Earth that scans the traveler’s atomic makeup and relays the information to a machine on Mars that instantly produces an exact replica. Meanwhile, back on Earth, the scanner kills the original person. There are many variants, including ones in which the original is not killed. (Extended discussions of Parfit’s theory can be found in the chapters by Coseru and Ganeri in this volume.) See also Daniel Dennett (1978), who relates an amusing tale in which his brain is removed from his body and placed in a vat; his brain remains connected to his body via radio transmitters, but when his original body is accidentally destroyed it is replaced with that of another. 16  The conceptual quandary is akin to that explored in the “rubber-hand illusion” (Botvinick and Cohen 1998; Tsakiris and Haggard 2005), and “full-body illusion” (Lenggenhager et  al. 2007; Blanke and Metzinger 2009). In both cases, a clever experimental setup is used to induce subjects to identify or locate themselves within a surrogate body (or body part) that lies at some remove from their physical bodies. 15

138

R. H. Sharf

philosophical astonishment that results from imagining oneself in the body of another. (Consider Śāriputra’s bewilderment upon finding himself suddenly in the body of a goddess in Chapter 6 of the Vimalakīrti-nirdeśa.) Judging from its inclusion in later anthologies, the Da zhidu lun body-swap tale appears to have been effective. It is included in the Zhongjing zhuanza piyu 眾經撰 雜譬喻 (Miscellany of Parables from the Scriptures), a short two-fascicle work containing forty-seven stories compiled by Dao Lüe 道略 in the early fifth century.17 The corpse-man tale also appears in a later compilation, the Jinglu yixiang 經律異 相 by Sengmin 僧旻, Baochang 寶唱, et al., completed in 516. This is a collection of miracle stories culled from the scriptures intended to attest to the power of Buddhism, and the authors explicitly mention the Da zhidu lun and Ayuwang jing as their sources for the versions of the corpse-man tale that appear therein.18 Whatever its popularity, it is not clear that the corpse-man tale is an effective response to the Nyāya critique. Nor is it likely to convince Zahavi. The man who finds himself in someone’s else’s body may be confused, but Zahavi could respond that the confusion is still his. That is, there is still a phenomenological continuity—a distinct and singular point-of-view that survives the transformation from one body to the next. Indeed, Zahavi might insist that without presuming the existence of this inner “witness” the tale loses its evocative power. The Ābhidharmika will respond that the point of the tale is to expose this witness as an illusory construct, and that phenomenal appearances, however deeply rooted, are not to be trusted. Thought experiments like the corpse-man tale may not in themselves decide such issues, although they can help us tease out critical nuances, including the subtle ways in which Buddhist intuitions about mind-body duality differ from our own. But this is only possible once we recognize the tales for what they are: playful but philosophically compelling thought experiments well suited to survive the journey across vast temporal and cultural distances. Acknowledgements  Earlier versions of this paper were delivered at the Institut für Indologie und Tibetologie, Ludwig-Maximilians-Universität München (May 16-18, 2012), at the K.J. Somaiya Centre For Buddhist Studies, University of Mumbai (7th Biennial International Conference on “Cross-Cultural Transmission of Buddhist Texts: Translation, Transliteration and Critical Edition,” December 5-7, 2012), and at National Taiwan University, Taipei (5th International Sheng Yen Education Foundation Conference, June 29, 2014). I would like to thank the organizers of those events, as well as Christian Coseru, Phyllis Granoff, Sean Kerr, Nir Feinberg, Alexander von Rospatt, Elizabeth Horton Sharf, and Evan Thompson for their invaluable comments and suggestions.

 T.208: 4.531c25-532a17; cf. Chavannes (1911: 72–74). Little is known about this collection or its compiler, but it is based on materials translated by Kumārajīva. See the discussions in Chavannes (1911: 1–3); Sugiyama (1992); and Yost (2013). 18  T.2121: 53.92c22-93a5 and 241a23-b9. 17

The Curious Case of the Conscious Corpse: A Medieval Buddhist Thought Experiment

139

References Blanke, Olaf, and Thomas Metzinger. 2009. Full-Body Illusions and Minimal Phenomenal Selfhood. Trends in Cognitive Sciences 13 (1): 7–13. Bloomfield, Maurice. 1917. On the Art of Entering Another’s Body: A Hindu Fiction Motif. Proceedings of the American Philosophical Society Held at Philadelphia for Promoting Useful Knowledge (April 13, 1916), 56 (1): 1–43. Botvinick, Matthew, and Jonathan Cohen. 1998. Rubber Hands ‘Feel’ Touch that Eyes See. Nature 391 (6669): 756. Chalmers, Robert. 1895. The Jātaka or Stories of the Buddha’s Former Births. Vol. 1. Cambridge: Cambridge University Press. Chavannes, Édouard. 1911. Cinq Cents Contes Et Apologues: Extraits Du Tripitaka Chinois Et Traduits En Français. Vol. 2. Paris: Ernest Leroux. Deeg, Max. 2009. From the Iron-Wheel to Bodhisattvahood: Aśoka in Buddhist Culture and Memory. In Aśoka: In History and Historical Memory, ed. Patrick Olivelle, 109–144. Motilal Banarsidass. Dennett, Daniel C. 1978. Where Am I? In Brainstorms: Philosophical Essays on Mind and Psychology, 310–323. Montgomery: Bradford Books. Ganeri, Jonardon. 2007. The Concealed Art of the Soul: Theories of Self and Practices of Truth in Indian Ethics and Epistemology. Oxford: Oxford University Press. ———. 2012. The Self: Naturalism, Consciousness, and the First-Person Stance. Oxford: Oxford University Press. Kapstein, Matthew T. 2001. Reason’s Traces: Identity and Interpretation in Indian and Tibetan Buddhist Thought. Boston: Wisdom Publications. Kritzer, Robert. 2000. Rūpa and the Antarābhava. Journal of Indian Philosophy 28 (3): 235–272. Lamotte, Étienne. 1944-1980. Le Traits de La Grand Vertu de Sagesse de Nāgārjuna (Mahāprajñāpāramitāsāstra). Vol. 5. Louvain: Institut Orientaliste. ———. 1958. Histoire du Bouddhisme Indien. Des Origines à Lʼère Śaka. Louvain: Institut Orientaliste. Lenggenhager, Bigna, Tej Tadi, Thomas Metzinger, and Olaf Blanke. 2007. Video Ergo Sum: Manipulating Bodily Self-Consciousness. Science 317 (5841): 1096–1099. Li, Rongxi. 1993. The Biographical Scripture of King Aśoka. Berkeley: Numata Center for Buddhist Translation and Research. Lopez, Donald S. 2008. Buddhism & Science: A Guide for the Perplexed. Chicago: University of Chicago Press. ———. 2012. The Scientific Buddha: His Short and Happy Life, The Dwight Harrington Terry Foundation Lectures on Religion in the Light of Science and Philosophy. New Haven: Yale University Press. Mahadevan, Anand. 2002. Switching Heads and Cultures: Transformation of an Indian Myth by Thomas Mann and Girish Karnad. Comparative Literature 54 (1): 23–41. Mukhopadhyaya, Sujitkumar. 1963. The Aśokāvadāna. Sanskrit Text Compared with Chinese Versions. Delhi: Sahitya Akademi. Palumbo, Antonello. 2010. From Constantine the Great to Emperor Wu of the Liang: The Rhetoric of Imperial Conversion and the Divisive Emergence of Religious Identities in Late Antique Eurasia. Paper Presented at the Mellon Foundation Sawyer Seminar Conversion in Late Antiquity: Christianity, Islam, and Beyond. University of Oxford, Corpus Christi College, 16 January 2010. Parfit, Derek. 1984. Reasons and Persons. Oxford: Clarendon Press. Przyluski, Jean. 1923. La légende de I’Empereur Açoka (Açoka-avadāna) dans les textes indiens et chinois. Paris: Paul Geuthner. Sander, Lore. 2000. A Brief Paleographical Analysis of the Brāhmī Scripts in Volume 1. In Buddhist Manuscripts Volume I, Manuscripts in the Schøyen Collection I, ed. Jens Braarvig, 285–300. Oslo: Hermes Publishing.

140

R. H. Sharf

Sharf, Robert H. 2014. Is Nirvāṇa the Same as Insentience? Chinese Struggles with an Indian Buddhist Ideal. In India in the Chinese Imagination: Myth, Religion, and Thought, ed. John Kieschnick and Meir Shahar, 141–170. Philadelphia: University of Pennsylvania Press. ———. 2018. Knowing Blue: Early Buddhist Accounts of Non-Conceptual Sense Perception. Philosophy East and West 63 (3): 826–870. Siderits, Mark, Evan Thompson, and Dan Zahavi, eds. 2011. Self, No Self?: Perspectives from Analytical, Phenomenological, and Indian Traditions. Oxford: Oxford University Press. Strong, John S. 1983. The Legend of King Aśoka: A Study and Translation of the Aśokāvadāna. Princeton: Princeton University Press. ———. 1992. The Legend and Cult of Upagupta: Sanskrit Buddhism in North India and Southeast Asia. Princeton: Princeton University Press. Ryūshō, Sugiyama 杉山龍清. 1992. 『衆経撰雑譬喩』と『大智度論』との関係について [A Study of the Zhongjing zhuanza piyu and the Da zhidu lun.] Journal of Indian and Buddhist Studies 40 (2): 584–586. Tawney, Charles Henry. 1880–84. The Kathá Sarit Ságara; or, Ocean of the Streams of Story. Vol. 2. Calcutta: J. W. Thomas. ———. 1883. Indian Folklore Notes from the Pali Jātakas and the Kathā Sarit Sāgara. Journal of Philology 12 (23): 112–126. Tsakiris, Manos, and Patrick Haggard. 2005. The Rubber Hand Illusion Revisited: Visuotactile Integration and Self-attribution. Journal of Experimental Psychology: Human Perception and Performance 31 (1): 80. Varela, Francisco J., Evan Thompson, and Eleanor Rosch. 1991. The Embodied Mind: Cognitive Science and Human Experience. Cambridge: The MIT Press. (A revised edition was published in 2017.). Wille, Klaus. 2000. Fragments from the Aśoka legend. In Buddhist Manuscripts Volume I, Manuscripts in the Schøyen Collection I, ed. Jens Braarvig, 219–231. Oslo: Hermes Publishing. Yost, Tyson Joseph. 2013. Anthologizing Buddhists: A Study of Avadāna Narratives and the Communities that Read Them in Early Medieval China. Ph.D. dissertation, Cornell University. Zacchetti, Stefano. 2015. Prajñāpāramitā Sūtras. In Brill’s Encyclopedia of Buddhism, ed. Jonathan A. Silk, vol. 1, 171–209. Leiden: Brill. Zahavi, Dan. 2011. The Experiential Self: Objections and Clarifications. In Self, No Self?: Perspectives From Analytical, Phenomenological, and Indian Traditions, ed. Mark Siderits, Evan Thompson, and Dan Zahavi, 56–78. Oxford: Oxford University Press.

How Can Buddhists Account for the Continuity of Mind After Death? Jan Westerhoff

1 Introductory Remarks When the relation between Buddhism and contemporary natural science is discussed there is usually at least one elephant in the room: the Buddhist conception of rebirth.1 This appears to constitute a clear example of a situation where Buddhism asserts the existence of something that science considers to be simply not there.2 The reason for this is obvious. If we accept the widespread contemporary belief that the mind is what the brain does, or, somewhat more cautiously, that the human mind could do nothing without the human brain also doing something at the same time then the destruction of the human brain during and after the process of death must

 I believe that there are some smaller, subsidiary elephants as well. One is the Buddhist conception of supernatural powers (iddhi, siddhi) that the practitioner is supposed to achieve at a relatively low level of meditational accomplishment. These include feats such as being able read another’s mind, or transforming one’s body into that of a god or an animal. See Bhadantācariya Buddhaghosa (1976: 1: 409–410, 444). 2  For a comprehensive and up-to-date bibliography of the literature on the debate about the Buddhist conception of rebirth see Bhikkhu Anālayo (2018: 237–271). 1

J. Westerhoff (*) Lady Margaret Hall, University of Oxford, Oxford, UK e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_8

141

142

J. Westerhoff

mean the cessation of the human mind.3 There are a number of strategies for addressing this tension. They can be grouped into four types: 1. Dogmatism. This suggests that one should accept that rebirth obtains because the Buddha, your teacher, or some other figure regarded as trustworthy has told you so. 2. Naturalism. This argues that the current scientific conception of how the mind works is essentially correct, so Buddhism’s belief in rebirth must be wrong. This position can then take two forms, a negative and a positive form.

(a) Eliminativist naturalism claims that because the belief in rebirth is central to Buddhism, its rejection entails that Buddhism as a whole cannot be true. (b) Reconstructivist naturalism claims that there are essential parts of Buddhism that do not rely on belief in rebirth, and that a reconstructed version of Buddhism can be put together that does not postulate rebirth.

3. Revisionism. This claims that the Buddha was not really serious when he taught rebirth and that this teaching does not form an integral part of Buddhism. 4. Validationism. This sets out to provide support for the claim that rebirth actually takes place. Validationism can take various forms. It can be based on empirical observations,4 or on theorizing about more general features of the world.5 Neither the dogmatist nor the eliminativist naturalist will be particularly interested in debating the tension between the Buddhist and the scientific view of rebirth since for both parties the tension has already been resolved: for one by rejecting the contemporary view of how the mind works, for the other by rejecting Buddhism’s potential for presenting a true account of the world. Revisionism does not seem to be a particularly attractive position, as detailed research into the earliest Buddhist materials make it hard to deny that the Buddha taught the theory of rebirth, and that he indeed meant what he said when he did so, that is, that his teaching of rebirth was not simply a concession to contemporary Indian cultural sensibilities that would have expected such claims to form part of soteriological teachings.6

 Note that this does not presuppose the mistaken idea that the Buddhist conception of rebirth requires a form of mind-body dualism (Batchelor 2011: 38; Edwards 1996: 14–15. Batchelor 2015: 300 considers the Buddhist belief in rebirth as based on a form of substance dualism). Early Buddhist sources like the Mahānidāna-sutta explicitly speak about the mutual conditioning of the physical and the mental (Bhikkhu Anālayo 2018: 9, 58), something that the metaphysical independence of the material and the mental which is characteristic of various forms of dualism does not allow. The Buddhist conception of mind is compatible with functionalism, however, unless we either assume that there can be non-physical bases for the realization of mental activity, or that entire minds could somehow be moved from one physical basis to another it could not provide an account of rebirth on such a functionalist basis. 4  Stevenson (1974), Bhikkhu Anālayo (2018), chapter 4. 5  Franco (1997: 128–132), Arnold (2012). 6  For a discussion supportive of this point see Bhikkhu Anālayo (2018), chapter 1. 3

How Can Buddhists Account for the Continuity of Mind After Death?

143

This leaves us with reconstructivist naturalist and validationist positions. The latter has received quite a bit of discussion in the recent literature,7 and I want to focus here on the validationist approach. Before doing so, however, it is necessary to add two qualifications: First, the Buddhist account of rebirth consists of two related claims: 1. The claim that the human mind can continue to exist after the destruction of our body, and 2. The claim that this mind is subsequently attached to another body, and that its relation to the new body is (at least in the case of rebirth as a human) of the same kind as the relation of our present body to our present mind. The first claim concerns mental continuity and is weaker than second, which concerns rebirth. Our minds might continue after death in some disembodied state yet never re-attach to a body, in which case 1. is true and 2. is false. Our discussion will be restricted to the first claim about mental continuity after death. Second, it is useful to divide arguments for mental continuity into general and particular arguments. General arguments are based on (presumed) facts about the world that hold at a fairly high level of abstraction, while particular arguments concentrate on specific cases. A familiar example of a general argument is Dharmakīrti’s argument for mental continuity, which relies on general premiss specifying what kinds of causes can have what kinds of effects.8 Examples of particular arguments include case-based research into phenomena like near-death experiences, past-life regressions, children’s memory of past lives, and xenoglossy.9 My discussion will deal exclusively with general arguments.

2 Goodman 2.1 Constructions and Anti-foundationalism The first argument I want to consider is that the constructivism of the kind defended by Nelson Goodman entails mental continuity. If this is right, and if we find reason to adopt Goodmanian constructivism (a theory that has considerably affinities with certain Madhyamaka ideas), mental continuity appears to be supported. Amongst the different forms of constructivism the Goodmanian variety is one of the most extreme, and one of the most interesting. Unlike various localized

 See, for example, Batchelor (2011, 2015), Flanagan (2013), Westerhoff (2017).  Arnold (2008), Jackson (1993). 9  For a discussion of all of these and of their place in the Buddhist religious tradition see Anālayo (2018), ch. 3. Amongst these Ian Stevenson’s rearch on children’s memory of past lives is probably best know. On this see Stevenson (1974), as well as the bibliography in Anālayo (2018). 7 8

144

J. Westerhoff

constructivist theories it does not just claim that scientific theories or social institutions are constructs but that everything is a construct.10 Note that the relation ‘is constructed from’ is an existential dependence relation, like ‘is fathered by’, ‘is caused by’, or ‘is a conglomeration of parts of’. This means that if the antecedent of the relation did not exist, the consequent would not exist either: if there is no father there is no son, if there is no fuel there is no fire, if there are no lego blocks there is no lego structure. An intriguing question concerning dependence-relations is whether they have a foundation, that is whether there is something which is only the antecedent of the relation but does not have an antecedent itself, or at least no antecedent different from it. The discussion concerning the foundation of causation and parthood stretches back to the early days of the philosophy. Both the existence of a first cause and the existence of partless atoms have been debated ever since. For Goodman at least the question of foundations is easily settled. Because constructions ‘go all the way down’ there is no unconstructed construct which could begin the chain. This, Goodman claims, is no problem; worrying about the beginning of the chain of constructions he regards as both misguided and best left to theology.11 Whether a downwards infinite constructive chain is philosophically problematic is not a question I set out to answer here. The main point I want to discuss is of a different but nevertheless related nature. It is this: either we assume that constructions take place in time, or they do not. If we accept the first horn of this dilemma, we get a chain of temporal processes stretching infinitely back in time. Yet constructions are supposed to be something humans do, but humans have not always been around. So either non-human minds existed before the existence of humans or there must have been constructions without constructors. If we accept the second horn of the dilemma instead, however, and assume that constructions do not take place in time, all constructions must already exist in some non- temporal realm, independent of human constructions. This does not sit well with the claim that it is humans who do the constructing.  ‘As nothing is at rest or is in motion apart from a frame of reference, so nothing is primitive or derivationally prior to anything apart from a constructional system.’ (Goodman 1978: 12), ‘And this, as I have mentioned earlier, goes all the way down. Not all differences between true versions can be thought of as differences in grouping or marking off within something common to all. For there are no absolute elements, no space-time or other stuff common to all, no entity that is under all guises or under none.’ (Goodman 1983: 107, note 6), ‘We cannot find any world-feature independent of all versions. […] No firm line can be drawn between world-features that are discourse-­ dependent and those that are not.’ (Goodman 1980: 212), ‘The line between convention and content is arbitrary and variable’ (Goodman 1980: 214). 11  ‘We might take construction of a history of successive development of worlds to involve application of something like a Kantian regulative principle, and the search for a first world thus to be as misguided as the search for a first moment of time.’ (Goodman 1978: 7, note 8), ‘The many stuffs – matter, energy, waves, phenomena – that worlds are made of are made along with the worlds. But made from what? Not from nothing, after all, but from other worlds. Worldmaking as we know it always starts from worlds already on hand; the making is remaking. Anthropology and developmental psychology may study social and individual histories of such world building, but the search for a universal or necessary beginning is best left to theology.’ (Goodman 1978: 6–7). 10

How Can Buddhists Account for the Continuity of Mind After Death?

145

2.2 Three Premises and One Dilemma Let us take matters a bit more slowly. There are three premisses in play: 1. No construct is constructed from something which is not a construct itself. 2.a Every construct must be constructed from something (its basis of construction). 2.b Every construct must be constructed by something (its constructor). 3. Bases of construction temporally precede the constructs. The first premiss just captures the fact that for Goodman there cannot be any nonconstructed objects which serve as the foundation for the constructional hierarchy. I am not going to argue for this premiss here, apart from noting that once we transpose it from the idiom of construction to that of dependence we get the claim that “no dependent object depends on something that does not itself depend on anything”, which follows directly from the Madhyamaka claim of universal emptiness. Premiss 2.a, a modern incarnation of the Parmenideian maxim ex nihilo nihil may appear to be controversial. Does not the empty set constitute a counterexample? Is not pure set theory a clear case of an intricate structure built literally from nothing at all? I think not. Set formation should not be thought of as a composition operation which makes one thing out of many things. After all it can also make a new thing out of a single thing by forming its singleton. The composition of a thing with itself appears to be either ill-defined or just identical with that very thing. It is more plausible to think of it as a collecting operation (David Lewis calls this the ‘lasso hypothesis’),12 collecting together things in a container (in the widest possible sense of the word). On this understanding we can make sense of the idea that a thing in a container is different from the thing itself, and that we can have containers containing nothing at all. Since we can also put empty containers into other containers we have the resources of building very complex structures from nothing but containers. But in this case the containers cannot be conceived of as mere operations, so that we could proceed from an empty world to a world with an object in it just by applying the set-forming operation to that world. Rather, taking set theory ontologically seriously commits us to an infinity of containers, and it is from these containers that the structures of set theory are built. They are not created from nothing at all. Could we deny premiss 2.b, and claim that some constructs are not constructed by anything and have no constructor? After all, if there can be design without a designer,13 could there not also be constructs without a constructor? Note, however, that we could be very permissive about what we allow to play the role of the constructor. In fact anything that brings about a construct could be so labelled, from the termite colony that constructs a mound, to the evolutionary process that brings a well-adapted organism into existence. But could such constructors be called minds? When talking about Goodmanian constructivism we do not speak about any constructions whatsoever, but specifically about the constructions of 12 13

 Lewis (1991: 42–45).  Dennett (1995: 68–73).

146

J. Westerhoff

versions or “world-making”. Each version is constructed together with the “many stuffs” it contains, and is the appearance of a specific world. But once we have the appearance of a world, it seems, we have a consciousness to which the world appears.14 Constructed version and constructing consciousness to which the version appears may arise in mutual dependence, but it seems difficult to introduce the idea of a constructed version without a constructor that constitutes an intentional agent.15 If we accept premises 1 and 2a/b, accepting or denying premiss 3 leads to the bifurcation of the dilemma into “minds all the way back” or “bases do not temporally precede the constructs”. If we look at the examples Goodman himself gave to illustrate his notion of construction it is obvious that they all take place in time, thereby suggesting to accept premiss 3 and embrace the first horn of the dilemma. He gives examples of making chairs, books, planes,16 and assembling stereo-­ systems.17 Making a stereo-system takes time, and before we have assembled its components we have the parts of a stereo-system, but nothing we can use to play our music with. All of this does not contradict the valid observation that Goodmanian making is not just the assembly of material components.18 But worldmaking is what humans do, and all things humans do take time, whether it is the making of a physical object, such as a cake or a castle, or a non-physical one, like a game or a theory. Consider the case of a theory, such as a theory of the physical world. Such a theory contains concepts, which are constructs. These concepts have been constructed from other concepts which have been around before the theory was around. By the first premiss these latter concepts have been constructed too, and so on. If we assume that these constructions take even the smallest amount of time the history of constructions will go back indefinitely.19 The history of constructions may go back infinitely, but human beings do not. Humans split from apes some time between five and seven million years ago, while biological life has not been around for much more than 3.5 billion years. But accepting premiss 3 together with the other two entails that nine billion years ago, before there even was an earth some constructions were around.  “Consciousness is the appearance of a world. The essence of the phenomenon of conscious experience is that a single and unified reality becomes present: If you are conscious, a world appears to you.” (Metzinger 2009: 15), “The activation of a unified, coherent model of reality within an internally generated window of presence, when neither can be recognized as a model, is the appearance of a world. In sum, the appearance of a world is consciousness” (Metzinger 2009: 192). 15  Nothing we have said here rules out the possibility that such intentional agents may be artificial intelligences constructing their own world-version. 16  Goodman (1980: 213). 17  Goodman (1983: 103–104). 18  ‘The worldmaking mainly in question here is making not with hands, but with minds, or rather with languages or other symbol-systems’ (Goodman 1980: 213). 19  Note that we do not have to assume that a construct is always more complex than what it is constructed from. A construct does not have to be more complex than each member of its basis, as it is if e.g. we merge two libraries. We can equally construct a new library by splitting one into two, or create a new one by splitting off parts from two old libraries and merging these. 14

How Can Buddhists Account for the Continuity of Mind After Death?

147

Now if constructions need a constructor, and since the constructions we have in mind in the Goodmanian framework are instances of worldmaking, and since the constructors of world-making are minds, we need to assume that there were non-­ human minds around at that time who carried out the constructions. Assuming that worldmaking is an activity only carried out by fairly complex minds this would also entail that complex minds could exist (or at least could have existed) independent of human bodies and stretch back in time as far as the constructions do, namely indefinitely far back. We thus end up with a somewhat unexpected proof that minds always have existed.20 If we don’t want to go down that route we need to deny premiss 3 and embrace the second horn of the dilemma, claiming that at least in some situations bases of construction do not temporally precede the constructs. It is also evident that the constructs cannot temporally precede the bases, otherwise there would have been nothing to construct the constructs from, since the bases did not exist at that time. But if there are two objects such that neither temporally precedes the other there are only two options: either the objects are simultaneous, or they do not stand in temporal relations relative to each other (according to the Platonist understanding mathematical objects exist in this way). The first option seems to be preferable, since the examples of constructs Goodman gives (the Big Dipper, chairs, books, planes, stereo-systems), as well as their bases (“matter, energy, waves, phenomena”) are located in time. In the case of the stereo-­ system embracing the second horn of the dilemma either implies the false claim that the stereo-system exists while its components exist or the obscure claim that it exists in some form in which it cannot fulfil its functional role (such as playing music). Considering non-material constructs such as physical theories we would be forced to assume that the conceptual construction of quantum mechanics was already present at the time of Newton. Moreover, it is not clear for what reason certain bases do not precede the corresponding constructs while others do. In the absence of such a reason it is preferable to assume that the precedence relation between basis and construct is the same for all basis-construct pairs. In this case it would follow that since new bases cannot come into existence apart from being constructed themselves21 all constructs must exist simultaneously in the present together with their bases. All of this is not very satisfactory, but matters do not improve if we assume that constructs and their bases do not exist in time. For now we have to assume some atemporal simulacra of the stereo-system and its components which stand in a relation of temporal dependence, and which are the proper subject-matter of Goodman’s  Note that because of our very permissive conception of what counts as a constructor in the discussion of premiss 2.b this does not entail the falsity of physicalism, at least as long we do not assume that matter has not always existed, and was preceded by some non-material stuff. 21  ‘We start, on any occasion, with some old version or world that we have on hand and that we are stuck with until we have the determination and skill to remake it into a new one. […] Worldmaking begins with one version and ends with another.’ (Goodman 1978: 97); ‘All we have available is scrap material recycled from old and stubborn worlds’ (Goodman 1980: 213). 20

148

J. Westerhoff

theorizing. If we go down that road this form of constructivism is transformed beyond recognition and loses any attractive features it might have had.

2.3 Further Worries Nevertheless, we might still have further worries about the argument presented so far. One is the suspicion that Goodman’s use of the expression “all the way down” is just a metaphor for the claim that what holds for one concept holds for all. The claim “All A are B” does not entail an infinite regress or an infinite number of As. When one says that a company is corrupt all the way down what we mean is that what holds for one employee holds for all, i.e. that all employees are corrupt. We are not committed to saying that the company has infinitely many employees. The reason why we do not get a regress in the case of the company is that the mereological relation ‘is a part of the company’ is well founded, that is, the company has atomic parts. It has parts which do not have parts which are parts of the company too. But we cannot assume the same for the existential dependence relation between a construct and what it is constructed from without falling into the kind of substantivalism Goodman rejects. Note that it would not help to relativize atomicity to an area of discourse. Someone might say that the people in the company are atomic only insofar as we speak of ‘employee-parts of the company’ and not, for example, when we speak of ‘material parts’. Similarly, one could argue, the basis of some constructs is constructed only relative to one perspective, not relative to another one. But then we still have to ask ‘is there something which has no parts under any understanding of parthood’ (an absolute atom)? This would be like asking ‘is there something not constructed no matter how we construe the notion of construction (an absolute constructional atom)?’ Goodman cannot allow these, but if there are no absolute constructional atoms the mereological case generates a regress as well. A second worry concentrates on the notion of time involved. Surely time is a construct too, and in fact Goodman pointed this out when faced with the criticism that things existing before human beings existed (such as the Big Dipper) could not possibly be human constructs. According to Goodman we construct time, and part of our construction of time is that there are certain things which are placed temporally prior to us.22 This is true, but does not resolve the dilemma, which arises when we try to assess the consequences of the ‘constructions all the way down’ approach relative to the notion of time we have constructed. As matters stand now we have the conception of time we have, and relative to this construction we are faced with the choice of either accepting that minds always have existed, or embracing the  ‘Does he ask how we can have made anything older than we are? Plainly, by making a space and a time that contains those stars.’ (Goodman 1980: 213). Note that this also resolves the problem of the ‘ancestral world’ raised by philosophers like Quentin Meillassoux against various forms of constructivism (see James 2017). 22

How Can Buddhists Account for the Continuity of Mind After Death?

149

unsatisfactory consequences of assuming that constructs and their bases are either simultaneous, or not located in time at all. This choice does not disappear if we agree that time is a human construction too. Where does all of this leave us? The easiest way of avoiding the dilemma is surely by rejecting the first premiss, that is by affirming that there is, after all, “something stolid underneath”,23 thereby rejecting universal constructivism. But this is of course no option open to the Goodmanian, nor would it be open to the Mādhyamika. And as neither premises 2.a and 2.b, nor premiss 3 seem easily assailable, the suggestion made here is to regard the preceding as an argument for the claim that minds (in the suitably generalized sense described above) always have existed. In fact it should not be too surprising that there is a fairly direct connection between the extreme constructivism we find defended in theories like Goodman’s “Ways of Worldmaking” and the Buddhist notion of beginningless ignorance that extends infinitely backwards, and the necessity of having sufficiently many minds present at any prior temporal stage of the infinitely regressive process of construction.

3 Interface Theory/Irrealism The second kind of argument I want to consider is formulated against the background of a theory of perception that is located centrally within the naturalist paradigm. It is a view of perception that sees human perception as analogous to the way an automatized cleaning robot perceives the flat it cleans. The robot first will first of all use information from its various sensors to construct an internal model of the contours of the flat. In the same way, it is argued (in a manner that is identical in principle, though not in complexity) we use information coming in through our senses in order to construct an inner model of the world.24 According to the view of perception we have in mind here this model constitutes an interface through which we interact with the reality modelled, much like a graphical user interface of a computer lets us interact with its underlying hardware. As in the case of the user interface, practical utility, rather than veridicality is the most important feature. The interface may present a group of ten files to us as located in a folder on our desktop, even though there is no contiguous physical location in the computer’s memory where all ten files are located. A question that immediately arises for this view of perception is the relation of the interface to the world behind the interface. Even though some approaches assume a rather naïve correspondence between items in the interface and items in

 Goodman (1978: 96).  Authors defending this picture include Hoffman et al. (2015), Dawkins (1999), Metzinger (2003, 2009), and Minsky (1988). See also Westerhoff (2016). 23 24

150

J. Westerhoff

the world,25 further reflection shows that this cannot be maintained. In the same way as the properties and individuals that populate our graphical user interface do not have corresponding properties or individuals corresponding to them in the computer’s innards (there is no little hourglass or any instance of its specific colour to be found anywhere in the hardware), there will also be no straightforward correspondence between the constituents of our perceptual interface and those of the world that caused it. There seem to be three possibilities of the possible relation between the representational interface and the represented: (a) a very general correspondence between some of the structural features of each, (b) the postulation of the mere existence of the represented, without the possibility of saying anything more about it, and (c) the subsumption of the represented under the representation. I have argued elsewhere26 that the prospects of filling (a) with any interesting contents are exceedingly slim, and that considerations of parsimony make (b) inferior to (c). What precisely any structural correspondence between our representation of the world and the world behind the veil of representation might be is very hard to say, and the mere postulation of an entity ‘out there’, beyond all representations, about which we cannot say anything, apart from the fact that it is there does not seem to explain philosophically much more than the simple assumption that the represented is part of the representational interface. My aim here is not to repeat this discussion, but to assess what consequences the view of perception just sketched, and the possibilities (a) to (c) available for spelling out its contents have for the question of mental continuity. One immediate consequence seems to be that the interface model of perception undermines the strongest reason we have for believing that the mind’s existence is limited in time. This is because on the assumption that the world that appears to surround us is in fact only the equivalent of a graphical user interface, a merely convenient device for facilitating our interactions, our brain, which is part of the world that appears to surround us, is part of the user interface as well. Of course if the process that is the mind is generated by the physical processes that take place in the brain, and if these processes cease once the brain is destroyed, the existence of the mind ceases at the same time. But according to the picture we have just discussed, the brain that generates the mind is in turn generated by the mind, insofar as it is part of the user interface in which we live our lives. Consider the following example. Assume your computer had a function that allowed you to look ‘under the hood’, making it possible to see the working of its hardware in action. In essence, the computer would produce a form of virtual reality in which you could walk around in the computer’s inside.27 You would be able to see how the computer’s logic gates operate, with their operation suitably slowed down and spatially enlarged so that you could comfortably perceive them. Also assume

 Dawkins (1999: 284).  Westerhoff (2016). 27  Hoffmann (1998: 195–196) describes a virtual tour of the underlying hardware of a computer simulating this very virtual tour. 25 26

How Can Buddhists Account for the Continuity of Mind After Death?

151

that the computer represented the working of the logic gates either as mechaninal devices regulating the flow of little balls28 running through a set of tubes, or as a water flowing through a set of pipes, regulated by suitable valves.29 Once you have acquainted yourself with the inner workings of the computer on the basis of this virtual reality you might infer that either the force of gravity, or the flow of water are essential for the working of the computer, and that if you suspended the computer in space or (in the case of the hydraulic logic gates) cooled it down below zero degrees Celsius all its operations would cease. You would, of course, be wrong about this, and this is because you made the faulty assumption that tubes and the balls, or the water and the pipes provided the fundamental level on which the computer’s activity is to be understood. Yet these are only representations themselves, and therefore unable to support any inferences on what the computer could or could not do, at least as long as these inferences are based on features of the computer only present in this specific representation. In the same way, if the brain does not exist at the bottom level of reality, facts about the brain cannot allow us to reach conclusions about something that exists at an even lower level, namely the perceptual interface and the mind that contains it. The question we would want to ask is whether the represented world behind the interface is such that the entire structure of the mind generating the brain generating the mind can be shown to be temporally limited, in a way a similar to that of referring to the impermanence of the brain in order to argue for the impermanence of the mind. It should be obvious that the prospects of succeeding in this demonstration are dim. Even for the strongest understanding of the represented/representation relation given by option (a), according to which there is some structural correspondence between each it is hard to make a strong case for the temporal limitation of the existence of the mental. If we cannot even be sure if the represented behind the interface has temporal features, how could we use it in order to argue that specific parts of the representation must have particular temporal features? So it appears that agnosticism regarding the mental continuity is the most plausible option: we simply cannot tell whether our perceptual interface is temporally limited, and considerations relating to the impermanence of the brain are strictly irrelevant to this. At least this is the case if we accept one of the first two ways of spelling out the possible relation between the representation and the represented, as either sharing some general structural features, or as a mere something about which we cannot know anything whatsoever. But things may appear somewhat differently if we assume that the represented needs to be subsumed as part of the representation, rather than being considered as an entity standing behind it.30 The idea behind option (c) is that we can reasonably refer to “the reality behind the representation”,  Fredkin/Toffoli (1982).  Taberlet et al. (2018). 30  Westerhoff (2016). I take this to be analogous to the denial of the ultimate reality of ultimate truth implied by the Madhyamaka’s idea of the ‘emptiness of emptiness,’ and the consequent subsumption of ultimate under conventional truth. 28 29

152

J. Westerhoff

but have to regard this as itself part of the representational interface.31 The representational interface presents matters to us as if there was something beyond the representation, and if we refer to “the reality behind the representation” we exclusively refer to the representation of the reality behind the representation, and not to any entity that is ontologically independent of the representation. To put matters in Carnapian terms, asking about the “reality behind the representation” only makes sense as an internal question, not as an external question. As an internal question, the answer is obviously affirmative (the way we see the world there are individuals, properties, and relations that correspond to our representations of such entities), but it only tells us what holds given a specific conceptual framework which we happen to operate in. The external question whether independent of our representational framework there is a reality behind the representation is not one we can coherently answer. But this entails that agnosticism about mental continuity is not a position we can adopt, for it would entail that there are some facts about mental continuity, but we can never know what these facts are. Yet if the level at which these facts would obtain (i.e. the level of “reality behind the representation”) cannot even be coherently formulated outside of a given representative framework it does not make sense to assume that these facts are there, but are somehow unknown. One fact that is important when assessing the status of mental continuity on the basis of this understanding of the relation between the representation and the represented is the idea that we cannot imagine our own non-existence.32 Stated like this, the claim seems patently false. I can, for example, imagine the day of my own funeral, in as much detail as necessary, and this is presumably a situation where my body (still) exists, though I myself and my mind no longer do. But if we visualize the details of the funeral it must be done from some perspective: from high above in the sky, from below in the coffin, or from the perspective of someone attending the funeral. But what I am imagining here is a future version of me having certain visual experiences: this is very different from imagining me not existing. We might argue that all this means is that we can only imagine our own non-­ existence from a third person, but not from a first person perspective.33 We can imagine what it means for us not to exist from the outside, but not from the inside. However, this does not seem quite right either. For suppose that I imagine the proceedings through the eyes of someone else, say, my spouse. I would then not have to presuppose that somehow my mind still existed, but that somebody else’s mind did. However, imagining some event “through the eyes” of some other person implies that we are somehow privy to that person’s inside view, a perspective that is ordinarily hidden from us. In this case, even though it is another person observing  See Button (2013: 80–81) for an inconsistent way of describing a similar position. The inconsistency results from the fact that this view (which Button calls ‘nonrealism’) cannot be combined with belief in the possibility of a god’s eye point of view. 32  This idea has already been discussed by Freud (1959: 304–305). See Nichols (2007) for a more recent account. 33  Nichols (2007: 218–219). 31

How Can Buddhists Account for the Continuity of Mind After Death?

153

the event (my spouse observing my funeral, say), it is still me observing the observation of the event.34 An omniscient narrator lets us get into the characters’ heads and see things as they appear them, yet it is still us reading the novel and perceiving things as the characters perceive them. So no matter how we set up the imaginary scenario, with us as a disembodied spirit hovering above it all, or lodged in the mind of some embodied perceiver separate from us, our role as an observer never goes away. And in fact it seems like an impossible task to ask us to imagine something, but then stipulate that we cannot play any part in the imaginative act. It is obvious that we may fail to feature in the content of the imaginative act (as when we simply imagine the world before we were born), but that does not rule out our involvement as the imagining agent. We cannot carry out an action and at the same time not be involved in some form of carrying in out. There appears to be some evidence that the inability of imagining our own non-­ existence is not simply a failure of our imaginative capacities, but might be biologically hardwired, and as such “an inevitable by-product of self-consciousness”.35 In particular the fact that younger children are more likely to express a belief in the continuity of mental life after death than older children appears to undermine the assumptions that such beliefs primarily have a cultural, rather than a natural origin.36 Now what are the implications of the inability to imagine our own non-existence? If we assume the standard view of the relation between mind and brain it does not imply terribly much,37 apart from the fact that we are prone to a specific,  This seems to entail the peculiar consequence that we can never have a third person perspective on anything. This is not entirely correct. There is still an enormous difference between imagining, say, that I come to the parking lot and see my car is missing (a first person imagining), and me imagining that particular parking lot with a particular car missing (which happens to be identical to my car) (this is usually regarded as a third person imagining). In the second case I abstract from various features of my perspective on the world (such that I own this car) and hence the second scenario produces a different emotional response from the first (“Who stole my car?”). Nevertheless, even the second perspective presupposes a view of the world which is in some important way our view. A more differentiated way of seeing the situation would therefore be to say that we can abstract in various ways from features of our peculiar perspective on the world, but we cannot imagine anything without an imaginer that is somehow identified with us. We can, therefore, imagine a universe without humans (Weisman 2007), but not one without observers. Suppose we imagine a universe with only two rocks in it. According to what we have just said, we can in fact only imagine a universe with two rocks and one observer in it. Can’t we then simply subtract the observer and leave the two rocks in place? In this case it is unclear what justification we still have to speak about a universe with two, rather than one rock in it. For it is not anything out there that guarantees the existence of a particular number of rocks (these are, after all, imagined universes). So it is the presence of the observer that allows us to distinguish between the two universes. Without the observer the two universes are simply two black boxes into which we cannot peek. And as such we cannot coherently say which is which (or whether there is even a difference between the two). 35  See Bering (2002). 36  Bering/Bjorklund (2004), Bering/Hernández-Blasi/Bjorklund (2005). 37  See Nichols (2007: 229). 34

154

J. Westerhoff

possibly hardwired cognitive illusion and that, as with other such illusions like the continuity of the visual field, choice blindness, or change blindness, our awareness that it is an illusion does not make the illusion go away. Indeed, what we believe about mental continuity, even if this belief is strongly supported by some neurobiological facts should be strictly irrelevant for whether our minds really are continuous. A sentence may say about itself that it contains exactly five words, but that does not tell us much when trying to determine whether it does so in fact. For this we have to inspect the sentence’s syntactic form, not what it says. In the same way we have to inspect the underlying physical form of the mind in order to determine is temporal continuity or limitation, not what it says about its own temporal continuity or limitation. But according to the picture under which we are currently operating there is precisely no such underlying physical form of the mind that could settle this issue by dint of its ontological fundamentality. Of course the familiar beliefs about how the brain relates to the mind are still part of our representational framework, but they have lost all authority to settle the status of the representation qua the brain’s role as basis of representation, since the brain is a representation itself. It makes no sense to ask whether absolutely the continuity of the mental obtains, if by absolutely we mean “outside of any system of representation”. This is because on the present understanding there is simply no perspective outside of any system of representation, since any perspective we come up with will inevitably be part of our representative framework.38 We can only ask the question concerning the continuity or discontinuity of the mental from within our representative framework. Consider such cases as bubble in a liquid, a pause in a piece of music, or a piece of text missing from a novel. It is straightforward to determine the volume of the bubble, the duration of the pause, or the number of missing characters. However, it would be impossible to do so if we wanted to determine volume, duration, or length of the bubble, pause, or missing piece of text as it is independent of the liquid, music, or novel in which it exists.39 All three depend on the entities in which they exist, and ascribing properties to them in an independent manner does not appear to make any sense. In the same way, as far as our mind has properties like temporal continuity or discontinuity it does so within the context of the representational interface, but not apart from it. And once we accept the position that we can only make sense of the “reality behind the representation” from within the representation, reference to the brain will no longer support the case for discontinuity. As soon as we think about the continuity or discontinuity of the mental from within our representational framework, the fact that we cannot imagine our own non-existence becomes relevant. For if we cannot coherently imagine that our mental existence ceases at some time, we could argue that the only way we can conceive  This would only be regarded as a kind of epistemological pessimism if we mistakenly regard the view from the outside of any representational framework as a more objective perspective we can never occupy. However, according to the present understanding the view from the outside of any representational framework is precisely not a more objective perspective, but no perspective at all. 39  For further discussion on the metaphysical status of holes see Casati/Varzi (1994). 38

How Can Buddhists Account for the Continuity of Mind After Death?

155

of us and our mental life at all is in terms of mental continuity.40 But if the assumption of mental continuity is a precondition of the possibility of conceiving of our mental life, and as we unproblematically assume that we have mental lives we seem to have the beginning of a transcendental argument for mental continuity.41 An obvious reply to the position that it is somehow constitutive of how we think about ourselves that we project our existence indefinitely into the future would be to say that thinking that x is so will not in all cases imply that x is so, even if we could not even think about x unless we thought of it in these terms. If some benighted creature could not conceptualize the universe as having more than five entities in it, or could not think about the universe unless it was thinking of it as containing less than five entities this does not mean that there are only five things in the universe. The reasons for this is that we, who have a more comprehensive perspective on the world than the creature can see that there are more than five things in the universe, and that its attempt to reason from its limited epistemic perspective to ontological claims is mistaken. In the same way, it would seem, there is a more comprehensive perspective on continuity according to which we can see that we will not continue to exist indefinitely into the future. This perspective is, of course, the view that our minds are supported by our physical bodies, and that once these bodies cease to exist, whatever they supported will cease to exist too. So even if we can argue that there is some (perhaps evolutionarily grounded) necessity to believe in our continued mental existence, this does not have any consequences for whether we really continue to exist. At this point it might be useful to introduce a parallel with the Kantian idea of time as a form of intuition. For the Kantian the existence of time is established by its role of a precondition of the possibility of perceptual experience. If we have experiences at all, they have to be structured in time, and as such we are justified in referring to the existence of time relative to our perceptual experience, without any necessity (or indeed any license) to ascribe temporal structure to the things in themselves. We might now similarly regard the conception of our own mental existence as open-ended as form of introspection and as such transcendentally real. There are, of course, positions where the Kantian view of time is ruled out as untenable. These are all the cases that include the possibility of somehow jumping out of the perceptual situation to something more objective, so that we can assess the existence of time independent of the role it plays in framing our perceptions from a perspective

 As Nichols (2007: 229, note 18) points out, bringing in the impossibility of imagining our own non-existence to explain a belief in mental continuity after death also makes it the source of a belief in retrograde continuity, i.e. a belief in mental continuity before birth. 41  Edward Craig (1985: 105) notes about the nature of imaginative limits that “[w]hat makes the limits of our imagination so important to us is the fact that they are also the limits within which reality must lie if it is to be intelligible.” If it is true that mental discontinuity lies beyond our imaginative limits this would then imply that the only way we can make sense of the mental is by conceiving of it in terms of mental continuity. 40

156

J. Westerhoff

outside of our perceptions. In these cases it may be true that even though our experience is temporally structured, there is still no time. But this presupposes our ability to have insight into the structure of things in themselves, something the Kantian position does not allow. In the same way, if it is possible for us to acquire a third-person perspective on the first-person perspective, jumping outside of our first-person experiential frame in order to assess it from outside of this frame it might turn out that although our mental stream looks as if it was open-ended from the inside it is in fact limited when examined from the outside. Yet again, this response presupposes that we are able to have experiences outside of our experiential frame. Once we accept the position we can only make sense of the “reality behind the representation” from within the representation, it seems that we would need a specific argument for the claim that sequence of representations will simply stop at a given time. All our reasons for this claim have been based on the fact that (within the series of representations that our experience constitutes) specific material processes stop at some time, and that with their cessation the mental processes that depend on them will also cease. Yet if these processes are representations themselves, this argument will no longer hold. As far as our first-person evidence goes, every mental moment is followed by another mental moment,42 and if we want to argue that this sequence does not continue indefinitely we have to provide some reason for why this should be so, rather than take this to be a default assumption. On the basis of the assumption that we can only make sense of the “reality behind the representation” from within the representation, the ‘mortality paradox’43 is dissolved in a way opposite to its normal resolution. The normal way of resolving the paradox is to propose that any sense of mental continuity based on the difficulty of imagining our own non-existence is simply an illusion.44 If there is a tension between what follows from our belief that the mind is brought about by the operation of the brain, and what follows from the way our mental life appears to us from the inside (even from the way our mental life has to appear to us, compelled by the

 This does not rule out the possibility of unconscious states during deep sleep or coma. The point is that even if we have a sequence of conscious (c) and unconscious (u) mind-moments c1, u1, …, un, c2, the first conscious moment after a period of unconsciousness (c2) is identified by us as belonging to the same person to which the last moment before the period of unconsciousness (c1) belonged. We thus do not have to presuppose that our mental existence is continuous. Even if we assume that the stream of consciousness is essentially discontinuous (such as in Galen Strawson’s ‘pearl’ model (1997: 424)), we can conceive of consciousness as a constantly restarting process, but as one where the successions of restartings extends indefinitely into the future. 43  “On the one hand, our powerful intellects come inexorably to the conclusion that we, like all other living beings around us, must one day die. Yet on the other, the one thing that these minds cannot imaging is that very state of nonexistence; it is literally inconceivable. Death therefore presents itself as both inevitable and impossible” (Cave 2012: 17). 44  Bering (2012: 130): “The mind is what the brain does; the brain stops working at death, therefore the subjective feeling that the mind survives death is a psychological illusion operating in the brains of the living.” 42

How Can Buddhists Account for the Continuity of Mind After Death?

157

psychological necessity described above) the former wins, simply because facts about the material world (which includes brains) constitutes the ontological bottom level, while mental appearances only sit on top of the material world in a dependent and derivative manner. But on the present understanding we simply have two appearances of our mental life, one based on our inability of imagining our own non-existence, and one based on our beliefs of how the brain brings about the mind. Neither corresponds to “reality outside of any system of representation”. And understood in this way, the implication of the absence of mental continuity implied by our beliefs about how the brain brings about the mind has dissolved. Yes, we may still accept all the connections between mental and neuronal activity that we have accepted previously but we will no longer be able to draw any conclusions about the presence or absence of mental continuity from them. All the conceptual resources we have for thinking about our mental lives involve a framework of representation that includes the representation of something that is not itself a representation. We are thus faced with representations wherever we turn in our analysis of reality. Any analysis of parts of the representational scheme will end up with more representations, and past experience involved each individual representation being followed by a temporally subsequent one. Thus, based on the evidence we have available to us it is not unreasonable to suppose that the temporal sequence of representations will continue indefinitely into the future. One might worry that one difficulty with this approach, especially when put forward as a defence of the Buddhist theory of mental continuity is that it commits us to accepting specific, sufficiently fundamental beliefs about ourselves as true. This might appear to conflict with the Buddhist theory of non-self. For one might argue that it appears to most people as if they existed in a substantial manner, i.e. as a self-­ sufficient, unified, temporally extended entity that functions as a locus of control and moral agency. Would we then not be able to argue that, as believing in our substantial existence is as much part and parcel of our first-person conception of our experiential frame as the belief in the open-endedness of our existence that if we want to support one, we also need to support the other? However, the difference between the two cases is that the insubstantiality of a person is something that can — according to the Buddhist perspective — be realized in an introspective manner. By analyzing the impermanent and unsatisfactory nature of the constituents that make up the person we are able to understand intellectually and realize experientially that we cannot exist in the substantial manner we intuitively thought to exist in. As such we can understand that, even though the belief in our own substantiality is a key part of how we believe ourselves to exist, it is nevertheless mistaken. On the other hand all the introspective evidence we can find seems to point at the open-endedness of our existence. Certainly by generalizing on the basis of past experience every moment of consciousness appeared to have been followed by another moment of consciousness. To find an argument for the limited nature of the succession of our experiences we need to appeal to third-person facts, to facts about the material basis of what produces the first-person perspective, together with

158

J. Westerhoff

considerations about their impermanent nature. Questioning the open-endedness of our conscious experiences appears to involve resources that go considerably beyond what we need to question our substantial nature. If this account is successful, what conception of mental continuity does it leave us with? Not one in which mental continuity is objectively vindicated, just as the Kantian account would not provide us with a conception of time and space as objective structural features of time and space. Rather, the conception is intersubjectively vindicated: we (and beings like us) think about our minds as continuing indefinitely into the future. But it is not just that we all happen to agree on this. The only way beings like us could conceive of mental continuity is by projecting it indefinitely far into the future, as the only way beings like us could conceptualize our experience is in spatial and temporal terms. And this, to an extent, is all the insight into the matter that could possibly be gained. We cannot peek beyond the forms of intuition to see if there really is time and space corresponding to them, nor could we step outside of our experience of a connected series of past, present and future events to see whether it terminates somewhere. All we can hope to achieve is a reflection-based view of the transcendental features that characterize our perspective from the inside.

4 Boltzmann Brains The third argument for mental continuity I want to explore is interesting because it does not require the adoption of the kind of large-scale philosophical assumptions that the previous two arguments presupposed. We do not have to accept that it is constructs all the way down, or that the only way we can make sense of reality behind the representational interface is by regarding it as part of that very interface. Rather, it is based on the idea that it is highly likely that, given enough time, there will be more or less exact duplicates of our brains, our bodies, or even of the entire environment in which we live.45 (I will refer to such a future duplicate that does not only include my brain, but also a sufficient amount of the perceptual environment in which this duplicate is embedded my Boltzmann successor). The idea of such future duplicates does not seem to cohere with the second law of thermodynamics, and as a matter of fact we never see the puddle of water that is a melted ice-cube to be replaced by a duplicate of the original ice-cube, nor do we find that heaps of shards suddenly re-assemble into a pot, or that ink dispersed in a glass of water re-groups into a blob of ink. However, as the second law of thermodynamics is a probabilistic law, all it can really be taken to say is that such occurrences are very unlikely, and if the sequence of trials is long enough even very unlikely events will occur. If we throw a coin a sufficient number of times we will get a sequence of ten consecutive heads, and if we keep on calculating the decimal expansion of π long enough, we will find a succession a hundred zeros  in a row.

45

 For the background of this idea see Carroll (2017).

How Can Buddhists Account for the Continuity of Mind After Death?

159

If we assume for the sake of argument that it is highly likely that our Boltzmann successor will come into existence at some time in the future, would that be of any interests for attempts of spelling out the Buddhist theory of mental continuity? Note that the Buddhist notion of mental continuity and rebirth does not require the transmission of any mental substance, or of any non-material ‘soul pellet’ that is passed on after the demise of our present body. One example used to illustrate the process of rebirth we find in the Buddhist literature is that of the impression of a seal.46 Clearly, when we impress a seal on a piece of clay, no matter has been transmitted from the seal to its impression (and if it has, this is entirely accidental). What has been transmitted is a piece of information going from the seal to its impression by the causal process of the action of impressing the seal. In the same way, what is required for the connection between a given person and its rebirth is a causal process that connects two persons, one living now and another, future one that is considered to be its rebirth. Could the relation between me and my Boltzmann successor qualify as a suitable connection? The answer to this question depends on what kind of causal relations you consider to be necessary for one person to count as a successor of another person. If we require that only physical and psychological processes that are very similar to processes that are ordinarily taken to connect different stages of a person are admissible47 it is evident that no personal continuity beyond the present body would be possible. But if we have this restrictive view of personal continuity we are presumably not concerned with the possibility of mental continuity and rebirth in the first place. It is therefore sensible to explore a somewhat more inclusive view. Assume that all we require is some causal continuity, and make the following three background assumptions. First, we spell out causality in terms of counterfactual dependence and understand the counterfactual that “if c had not occurred, e would not have occurred” in the following manner: take a possible world that is just like ours, apart from minimal changes ensuring c does not hold in it. Then run the history of this world according to the laws of nature in our world and check whether e occurs. If there is no e, e is counterfactually dependent on c. Second, assume that the mental supervenes on the physical. If we change the world such that a specific mental event no longer occurs, we will have changed the physical properties of the world. Third, assume determinism: a complete description of the physical state of the world fixes a unique present and a unique past.48 Now consider a specific mental event e of mine. If e had not been there, by the supervenience of the mental on the physical the physical constitution of the world would have been different. And given determinism, the state of the world a long time into the future when my Boltzmann successor forms will be different from the  Skilling (1997: 254).  Unger (1990: 70) proposes that survival of a person requires psychological processes to be “carried forward in ways that are, on the whole, not terribly different from the ways that psychological continuity is achieved in ordinary cases”. 48  See Loew (2017: 765–766) for a fuller account of these assumptions and of the argument based on them. The third (and most controversial) assumption might be dispensable. 46 47

160

J. Westerhoff

state that would have resulted at the same time if e had not been there. So there are specific events at the time of my Boltzmann successor that will not have obtained in a world in which e had not occurred earlier, and so these events will be counterfactually dependent on, and hence caused by my mental event e.49 If we can accept the causal connection between us and our Boltzmann successor, it can be the case that a certain human h dies at some time, and that at a much later time a Boltzmann successor h′ arises that is causally connected with h, and that this can be regarded as the rebirth of h in h′. That an enormous amount of time might have elapsed between h and h′ is not necessarily problematic, at least as long as h′ regards itself as the successor of h, in the same way as we might regard ourselves as the successor of a person that previously entered deep sleep or a coma. A major advantage of this approach would be that it allows to account for the large-scale continuity of mental causal connectivity and mental development without having to move out of the naturalist paradigm according to which the mind is what the brain does, and without having to make the large-scale philosophical assumptions discussed in the previous two sections. Accounting for the overall continuity of mental causal connectivity and mental development is essential in order to make sense of the Buddhist concepts of karma and soteriology. Making sense of karma in this life in terms of a connection between intentions on the one hand, and hedonic states and experiences on the other seems to be possible50 without having to postulate some kind of hitherto undiscovered force that ensures good intentions are rewarded and bad intentions are sanctioned. However, we don’t get the full picture of karmic conditionality as conceived by the Buddhists if we simply consider the scope of this life. In order to demonstrate its full complexity, and in order to fulfill its explanatory role karma has to be conceived of as stretching indefinitely into the past and into the future, without being restricted to this time of our lives. Similarly, the soteriological urgency of suffering explained in the first noble truth, and the complexity of the path leading to the cessation of suffering explained in the second do not become apparent if we simply focus on this present life. To put matters very simply, what is so problematic about our suffering if it can only last for the limited duration of this life, and how could we justify the extremely detailed and demanding path to the full cessation of suffering if all our suffering will automatically end with the dissolution of this body? In order to appreciate the existential horror of the first noble truth we have to understand that, unless we do something about it, the cycle of suffering repeats itself again and again into the future, and in order to find the motivation to follow the path described in the second noble truth we need to be assured that no automatic release from suffering will ensue after our present body dies.

 Loew (2017: 768–772) also points out that the causal connection between us and our Boltzmann successor could also be established if we are somewhat less promiscuous regarding the causal relations we admit. The argument also goes through if we demand that not just any causal relations, but only law-like causal relations are admissible. 50  Siderits (2007: 158). 49

How Can Buddhists Account for the Continuity of Mind After Death?

161

If we assume that in the future some Boltzmann successor of ours will have the same psychological attributes and behavioural tendencies we have, and will experience suffering in a way very similar to how we experience it we might be more motivated to change these attributes and tendencies for good, in a way that makes our mental lives less conducive to the production of suffering, and more conducive to the production of happiness.

5 Conclusion All the three accounts we have discussed above, universal constructivism, the subsumption of reality behind the appearances amongst the appearances, and the assumption of Boltzman successors face significant philosophical objections. This should come as no surprise, as any philosophical theory of any interest faces significant objections. The point of the preceding discussion was not to provide a proof for the continuity of mind after the death of our body in order to convince somebody who assumes that the end of our bodily existence is the end of our existence tout court. I see the goal of philosophical discussion in any case not so much as trying to come up with arguments that leave the opponent with no rational choice but to agree with our conclusion,51 but as measuring the philosophical price of a theory.52 If we consider large-scale philosophical disagreements in philosophy, like those between realism and anti-realism, materialism and idealism, or consequentialism and deontology it would be naïve to assume that any of them will be finally resolved by a master argument, some conceptual razor-blade that cuts through the Gordian knot of centuries of philosophical debate and forces all anti-realists to convert to realism (or the other way round). Instead, if we measure the price of a philosophical theory we focus on the fact that every philosophical theory has to make some assumptions (its price), and will as a result be able to explain specific features of the world (its payoff). Now given the different views of which features need explaining, and different views of what constitutes a good explanation we may have different view on which price (in the form of philosophical assumptions) we are willing to pay. This is not anything that can be settled by rational argument itself, but what such arguments can settle is how the different assumptions a philosophical theory makes can generate philosophical explanations, and how these explanations relate to other philosophical theories in the vicinity. It has been a common assumption so far that the theoretical price to pay for taking on board key parts of the Buddhist philosophical system is a rejection of the

51 52

 See Nozick (1981: 4).  See Lewis (1983: x).

162

J. Westerhoff

naturalist conception of how the brain relates to the mind.53 I believe that this assumption is mistaken. None of the three approaches I have described above is likely to be regarded as an excursion into ‘hocus pocus’ territory,54 yet all of them provide a basis for making sense of mental continuity after death along broadly Buddhist lines, as well as of the accompanying beliefs about karmic conditionality in an intellectually responsible way. Whether any of them is fully satisfactory, both in terms of its intrinsic plausibility and in terms of its ability to spell out the relevant Buddhist concepts is something for further discussion to decide. What is clear, however, is that taking Buddhist ideas about mental continuity and karma seriously does not entail the rejection of key naturalist ideas about how the body relates to the mind. In fact there are various approaches with firm naturalist grounding that can be explored in order to attempt to explicate certain key ideas of Buddhism,55 and there is very little justification for producing an amputated version of essential Buddhist ideas out of the desire to cohere with a naturalist view of the world.

References Anālayo, Bhikkhu. 2018. Rebirth in Early Buddhism & Current Research. Boston: Wisdom Publications. Arnold, Dan. 2008. Dharmakīrti’s dualism. Critical reflections on a Buddhist proof of rebirth. Philosophy Compass 3 (5): 1079–1096. ———. 2012. Brains, Buddhas, and Believing: The Problem of Intentionality in Classical Buddhist and Cognitive-Scientific Philosophy of Mind. New York: Columbia University Press. Batchelor, Stephen. 2011. Confessions of a Buddhist Atheist. New York: Spiegel & Grau. ———. 2015. After Buddhism. Rethinking the Dharma for a Secular Age. New Haven, CT: Yale University Press. Bering, Jesse. 2002. Intuitive conceptions of dead agents’ minds: The natural foundations of afterlife beliefs as phenomenological boundary. Journal of Cognition and Culture 2: 263–308. ———. 2012. The God Instinct. The Psychology of Souls, Destiny, and the Meaning of Life. New York: W.W. Norton & Company. Bering, Jesse, and David Bjorklund. 2004. The natural emergence of afterlife reasoning as a developmental regularity. Developmental Psychology 40: 217–233. Bering, Jesse, Carlos Hernández-Blasi, and David Bjorklund. 2005. The development of ‘afterlife’ beliefs in religiously and secularly schooled children. British Journal of Developmental Psychology 23 (2005): 587–607. Bitbol, Michel. 2017a. Why the moon follows me: observation and relationality in phenomena. In The Monastery and the Microscope. Conversations with the Dalai Lama on Mind, Mindfulness, and the Nature of Reality, ed. Wendy Hasenkamp and Janna R. White, 51–84. New Haven/ London: Yale University Press.

 Unless, that this, one accepts one of several proposal for watering down the Buddhist theories of karma and rebirth in such a way that the remainder does not generate any conflicts with naturalist assumptions. As I explained in greater details elsewhere (Westerhoff 2017), I am highly skeptical of these proposals. 54  As Flanagan (2013: 4) describes Buddhist beliefs about rebirth and karma. 55  For an approach not discussed here see Bitbol (2017a, b). 53

How Can Buddhists Account for the Continuity of Mind After Death?

163

———. 2017b. A strange loop of relations: phenomenology and experience. In The Monastery and the Microscope. Conversations with the Dalai Lama on Mind, Mindfulness, and the Nature of Reality, ed. Wendy Hasenkamp and Janna R. White, 171–191. New Haven/London: Yale University Press. Buddhaghosa, Bhadantācariya. 1976. The Path of Purification. Berkeley/London: Shambala. Button, Tim. 2013. The Limits of Realism. Oxford: Oxford University Press. Carroll, Sean. 2017. Why Boltzmann brains are bad. arXiv:1702.00850. Casati, Roberto, and Achille Varzi. 1994. Holes and Other Superficialities. Cambridge, MA: MIT Press. Cave, Stephen. 2012. Immortality. The Quest to Live Forever and How it Drives Civilization. London: Biteback Publishing. Craig, Edward. 1985. Arithmetic and fact. In Exercises in Analysis, ed. Ian Hacking, 89–112. Cambridge: Cambridge University Press. Dawkins, Richard. 1999. Unweaving the Rainbow: Science, Delusion, and the Appetite for Wonder. London: Penguin. Dennett, Daniel. 1995. Darwin’s Dangerous Idea. Evolution and the Meanings of Life. London: Penguin. Edwards, Paul. 1996. Reincarnation. A Critical Examination. Amherst: Prometheus Books. Flanagan, Owen. 2013. The Bodhisattva’s Brain. Buddhism Naturalized. Cambridge, MA: MIT Press. Franco, Eli. 1997. Dharmakīrti on Compassion and Rebirth. Vienna: Arbeitskreis für Tibetische und Buddhistische Studien. Fredkin, Edward, and Tommaso Toffoli. 1982. Conservative logic. International Journal of Theoretical Physics 21 (3–4): 219–253. Freud, Sigmund. 1959. Thoughts for the times on war and death. In Collected Papers, vol. 4, 288–317. New York: Basic Books. Goodman, Nelson. 1978. Ways of Worldmaking. Indianapolis: Hackett. ———. 1980. On starmaking. Synthese 45: 211–216. ———. 1983. Notes on the well-made world. Erkenntnis 19: 99–108. Hoffman, Donald D. 1998. Visual Intelligence, How We Create What We See. New York/London: W.W. Norton & Company. Hoffman, Donald D., Manish Singh, and Chetan Prakash. 2015. The interface theory of perception. Psychonomic Bulletin & Review 22 (6): 1480–1506. Jackson, Roger. 1993. Is Enlightenment Possible? Dharmakīrti and rGyal tshab rje on Knowledge, Rebirth, No-Self and Liberation. Ithaca: Snow Lion. James, Simon. 2017. Madhyamaka, metaphysical realism and the possibility of an ancestral world. Philosophy East and West 68: 4, 2019, 1116–1133. https://doi.org/10.1353/pew.0.0132. Lewis, David. 1983. Philosophical Papers 1. New York: Oxford University Press. ———. 1991. Parts of Classes. Oxford: Blackwell. Loew, Christian. 2017. Boltzmannian immortality. Erkenntnis 82: 761–776. Metzinger, Thomas. 2003. Being No One: The Self-model Theory of Subjectivity. Cambridge, MA: MIT Press. ———. 2009. The Ego Tunnel. New York: Basic Books. Minsky, Marvin. 1988. The Society of Mind. New York: Simon & Schuster. Nichols, Shaun. 2007. Imagination and immortality: thinking of me. Synthese 159 (4): 2015–2233. Nozick, Robert. 1981. Philosophical Explanations. Cambridge, MA: Harvard University Press. Siderits, Mark. 2007. Buddhist as Philosophy. Aldershot: Ashgate. Skilling, Peter. 1997. “Eight appropriate similes” (‘Thun pa’i dpe brgyad): verse 5 of Nāgārjuna’s Pratītyasamutpādahṛdayakārikā and the Sūtra on the questions on how transmigration occurs. In Dharmadūta — Mélanges offerts au Vénérable Thích Huyên-Vi à l’occasion de son soixantedixième anniversaire, ed. Bhikkhu Tampalawela Dhammaratana and Bhikkhu Pāsādika, 251–260. Paris.

164

J. Westerhoff

Stevenson, Ian. 1974. Twenty Cases Suggestive of Reincarnation. 2nd revised and enlarged edition. Charlottesville: University of Virginia Press. Strawson, Galen. 1997. The self. Journal of Consciousness Studies 5–6: 405–428. Taberlet, Nicolas, Quentin Marsal, Jérémy Ferrand, and Nicolas Plihon. 2018. Hydraulic logic gates: building a digital water computer. European Journal of Physics 39: 025801. Unger, Peter. 1990. Identity, Consciousness, and Value. Oxford: Oxford University Press. Weisman, Alan. 2007. The World Without Us. London: Virgin. Westerhoff, Jan. 2016. What it means to live in a virtual world generated by our brain. Erkenntnis 81 (3): 507–528. ———. 2017. Buddhism without reincarnation? Examining the prospects of a “naturalized” Buddhism. In A Mirror is for Reflection. Understanding Buddhist Ethics, ed. Jake Davis, 146–165. New York: Oxford University Press.

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy and Cognitive Science Evan Thompson

[O]ne and the same philosophical issue can be a matter of contention in philosophical centers entirely distinct from each other, widely separated in time and space, and belonging to disparate cultures. And this in its turn provides a compelling argument for two things: first, for the genuine universality of some major philosophical problems; and, second, for the desirability of further comparative study of the respects in which the two philosophical traditions in question may illustrate this universality…1

A recurrent problem in the philosophical debates over whether there is or can be nonconceptual experience or whether all experience is conceptually structured, mediated, or dependent is the lack of a generally accepted account of what concepts are. Without a precise specification of what a concept is, the notion of nonconceptuality is equally ill defined. This problem cuts across contemporary philosophy and cognitive science as well as classical Indian philosophy,2 and it affects how we go about philosophically “engaging Buddhism”3 in particular. Buddhist philosophers generally argue that our everyday experience of the world is conceptually constructed. For example, they argue that what appear to us to be stable, enduring entities possessing properties and belonging to kinds are fictions created by the imposition of concepts onto the incessant flux of momentary events. At the same time, “nonconceptual cognition” (nirvikalpa jñāna) marks the limits of 1  P. F. Strawson, “Reply to Arindam Chakrabarti,” in Lewis Edwin Hahn, ed., The Philosophy of P.F. Strawson (Chicago and LaSalle, IL: Open Court Publishing, 1999), p. 327. 2  See Mark Siderits, Tom Tillemans, and Arindam Chakrabarti, eds., Apoha: Buddhist Nominalism and Human Cognition (New York: Columbia University Press, 2011), especially the Introduction by Chakrabarti and Siderits. 3  Jay Garfield, Engaging Buddhism: Why It Matters to Philosophy (New York: Oxford University Press, 2015).

E. Thompson (*) University of British Columbia, Department of Philosophy, Vancouver, BC, Canada e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_9

165

166

E. Thompson

conceptuality. But what precisely do “conceptual” and “nonconceptual” mean? Consider that “concept” is routinely used to translate the Sanskrit term, vikalpa; nirvikalpa is accordingly rendered as “nonconceptual.” But vikalpa has also been rendered as “imagination,”4 “discriminative construction,”5 “discursive thought,”6 and “discrimination.”7 Related terms, such as kalpanā (conceptualization/mental construction) and kalpanāpoḍha (devoid of conceptualization/mental construction), have also been rendered in various ways. Besides the question of how to translate these terms in any given Buddhist philosophical text, how should we relate them to current philosophical or cognitive scientific uses of the term “concept”? More generally, given that the relationship between the conceptual and the nonconceptual has been one of the central and recurring issues for the Buddhist philosophical tradition altogether, can Buddhist philosophy bring fresh insights to our contemporary debates about whether experience has nonconceptual content? And can contemporary philosophy and cognitive science help to illuminate or even resolve some older Buddhist philosophical controversies? A comprehensive treatment of these questions across the full range of Buddhist philosophy and contemporary philosophy of mind and cognitive science would be impossible. I will restrict my focus to certain core ideas from three strands of Indian Buddhist philosophy—the Abhidharma scholastic philosophers (roughly 150 B.C.E. to 450 C.E.), the epistemological tradition (pramāṇavāda) deriving from Dignāga (ca. 480–540 C.E.) and Dharmakīrti (ca. sixth-seventh centuries C.E.),8 and the Yogācāra tradition (one of the dominant Indian Buddhist philosophical movements from roughly the fifth century C.E. onwards). I will juxtapose the Buddhist ideas with both cognitive science theories of concepts and the current philosophical debate between conceptualist versus nonconceptualist views of experience.

 Bimal Krishna Matilal, Perception: An Essay on Classical Indian Theories of Knowledge (Oxford: Oxford University Press, 1986), pp. 312–315. 5  Hugh B. Urban and Paul J. Griffiths, “What Else Remains in Śūnyatā?” An Investigation of Terms for Mental Imagery in the Madhyāntavibhāga-Corpus, Journal of the International Association of Buddhist Studies 17(1) (1994): 1–25. 6  Alex Wayman, “Buddhist Terms for Recollection and Other Types of Memory,” in Janet Gyatso, ed., In the Mirror of Memory: Reflections on Mindfulness and Remembrance in Indian and Tibetan Buddhism (Albany, NY: State University of New York Press), pp. 133–147, at p. 137. 7  Robert Sharf, “Knowing Blue: Early Buddhist Accounts of Non-Conceptual Sense Perception,” Philosophy East and West 68(3) (2018): 826–870. 8  The term “pramāṇavāda” means “pramāṇa theory.” A pramāṇa is an “instrument” or reliable means for obtaining knowledge in the sense of reliable cognition. Pramāṇavāda is concerned with what constitutes a pramāṇa. This type of philosophical discourse originally derived from the Brahmanical Nyāya school. See Matthew Dasti and Stephen Phillips, The Nyāya-sūtra. Selections with Early Commentaries (Indianapolis, IN: Hackett Publishing Company, 2017). The Buddhist pramāṇavāda philosophers (pramāṇavādins) can also be described as belonging to the Yogācāra-­ Sautrāntrika movement. For a useful introduction to pramāṇavāda, see John D. Dunne, Foundations of Dharmakīrti’s Philosophy (Somerville, MA: Wisdom Publications, 2004), chapter 1. 4

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

167

My approach will be an example of what Mark Siderits calls “fusion philosophy,” which he describes as the “counterpoising of distinct traditions” in the effort to solve common philosophical problems.9 In the case at hand, the common problem is to specify what a concept is and accordingly what nonconceptual experience could be. My presentation has six sections and will proceed as follows. First, I analyze some Abhidharma issues about the relation between nonconceptual sense perception and conceptual cognition. The impetus for my analysis is a recent paper by Robert Sharf.10 He argues that the structure of the Abhidharma theories of perception cannot help but raise the problem of how to relate nonconceptual sense perception to conceptual cognition and that the way that the Abhidharma philosophers (Ābhidharmikas) frame the problem makes it insoluble. Second, I use cognitive science to clarify and sharpen the problem. Third, I argue that although the problem is intractable within the Abhidharma framework, it may be resolvable using a model of concept formation based on reading Buddhist epistemology (pramāṇavāda) through cognitive science glasses. Fourth, I use this model to present a Buddhist epistemological perspective on the contemporary conceptualist versus nonconceptualist debate. Fifth, I argue that Yogācāra offers a conception of nonconceptual experience that is absent from this debate. In many Yogācāra texts, awareness that is said to be free from the duality of “grasper” (grāhaka) and “grasped” (grāhya) is nonconceptual. The grasper-grasped duality is the subject-object structure of ­experience. According to the Yogācāra philosophers (Yogācārins), conceptual experience necessarily has a subject-object structure; hence, any experience lacking the subject-object structure must be nonconceptual. None of the contemporary philosophical arguments for nonconceptualism is adequate or suitable for explicating this unique understanding of conceptuality and nonconceptuality. Thus, Yogācāra provides an original perspective on what has been called the problem of the “scope of the conceptual,” that is, “how the conceptual/nonconceptual distinction should be drawn.”11 For this reason, among others, Yogācāra has something to offer philosophy of mind.12 Moreover, using cognitive science, we may be able to render some of the Yogācāra ideas in a new way, while in turn recasting ideas from cognitive science. I illustrate these points using recent behavioural and neuroscientific studies of the effects of Buddhist mindfulness meditation practices on the perception and

 See Mark Siderits, Personal Identity and Buddhist Philosophy: Empty Persons (Ashgate Publishing, 2015), p. xi. 10  Sharf, “Knowing Blue,” op. cit. 11  Stephen Laurence and Eric Margolis, “The Scope of the Conceptual,” in Eric Margolis, Richard Samuels, and Stephen P. Stich, eds., The Oxford Handbook of Philosophy of Cognitive Science (Oxford: Oxford University Press, 2012), pp. 291–317, at p. 291. 12  I am inspired here by Jay Garfield’s approach to relating Yogācāra to philosophy of mind. See his Engaging Buddhism, op. cit., and Jay Garfield, Empty Words: Buddhist Philosophy and Cross-­ Cultural Interpretation (New York: Oxford University Press, 2011), Part II: Yogācāra. See also Christian Coseru, Perceiving Reality: Consciousness, Intentionality, and Cognition in Buddhist Philosophy (Oxford: Oxford University Press, 2012). 9

168

E. Thompson

experience of pain. Sixth, I end by combining Buddhist philosophical ideas with the enactive approach in cognitive science to sketch an enactive perspective on concepts. From this perspective, a concept is not a mental entity by which an independent subject grasps or represents independent objects, but rather one aspect of a complex dynamic process in which the mind and the world are interdependent and co-emergent poles.13 In pursuing these aims, I hope to illustrate Siderits’s idea of fusion philosophy and to show the value of thinking about the mind from a cross-cultural philosophical perspective. In Jonardon Ganeri’s words, “Philosophy of mind is indeed a transcultural undertaking: the search for a fundamental theory of mind must never limit itself to the intuitions and linguistic practices of any one community of thinkers but should be ready to learn from diverse cultures of investigation into the nature of mind and mind’s involvement in world.”14

1 Seeing Blue In the Abhidharma, the problem of conceptuality versus nonconceptuality centres on how to understand the relation between the sensory grasping of a quality-­ particular (e.g., a blue quality-particular) and the mental grasping of an object via a name or mental label. The former (sensory) part of the cognitive process is supposed to be the nonconceptual part; the latter (mental) part of the cognitive process is supposed to be the conceptual part. The Abhidharma project is to defend the truth of the Buddha’s word, as preserved in the Buddhist scriptures, especially the Buddha’s teaching that all phenomena arise dependent on causes and conditions, and that no phenomena, singly or collectively, constitute a self, in the sense of an enduring subject of experience or an internal controller of the body and the mind. In the Buddhist scriptures known as the Nikāyas (the early Pāli language texts containing the teachings attributed to the Buddha), the functioning of the sense organs, plus the kinds of sensory awareness based on them and the associated kinds of feeling or hedonic tone, are distinguished from the perceptual recognition of a qualitative characteristic (e.g., a shape or a colour), but whether these elements are all separate mental events or inseparable parts of one mental event is not clear. It is also not clear whether perception, properly speaking, is the functioning of the sense

 For the enactive approach in cognitive science, with connections to Buddhist philosophy, see Francisco J. Varela, Evan Thompson, and Eleanor Rosch, The Embodied Mind: Cognitive Science and Human Experience (Cambridge, MA: MIT Press, 1991, revised edition 2017). See also Evan Thompson, Mind in Life: Biology, Phenomenology, and the Sciences of Mind (Cambridge, MA: Harvard University Press, 2007). 14  Jonardon Ganeri, Attention, Not Self (Oxford: Oxford University Press, 2017), p. 5. 13

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

169

organs and the associated sensory awareness, or the recognition of a characteristic feature.15 Consider this passage: Dependent on the eye and forms, eye-consciousness arises. The meeting of the three is contact. With contact as condition there is feeling. What one feels, that one perceives. What one perceives, that one thinks about. What one thinks about, that one mentally proliferates. With what one has mentally proliferated as the source, perceptions and notions [born of] mental proliferation beset a man with respect to past, future, and present forms cognizable through the eye.16

The Pāli term translated as “perceives” in “what one perceives” is saññā (saṃjñā in Sanskrit); it has the sense of recognizing something based on a mark or characteristic.17 Recognition enables perceptual judgement and thought (“what one perceives, one thinks about”). It is not possible to tell from this passage, however, whether a given sensory awareness (e.g., an “eye-consciousness” or moment of visual awareness), the associated “feeling” (vedanā, a hedonic sensation that is either pleasant, unpleasant, or neutral), and the associated recognition (saññā/saṃjñā) arise sequentially or simultaneously. It is also not possible to tell whether these elements are separate “processing stages” or inseparable aspects of one perceptual event. These issues bedevil the Ābhidharmikas and figure in their controversies about perception.18 To carry out their project of systematizing and defending the Buddha’s teaching, the Ābhidharmikas elaborate a relentless form of reductionism, analyzing the entire universe, including all experienced phenomena, into a set of elementary existents called dharmas. Many scholars now agree that dharmas (according to the Ābhidharmikas) are elementary tropes or quality-particulars, existing at space-time points.19 For example, according to the Ābhidharmikas, a particular pot is blue in virtue of the presence of many miniscule and elementary blue dharmas or blue quality-particulars. (They are elementary in the sense that they have no parts that are

 See John Taber, A Hindu Critique of Buddhist Epistemology: Kumārila on Perception. The “Determination of Perception” chapter of Kumārila Bhaṭṭa’s Ślokavārttika (London and New York: Routledge Press, 2005), p. 171, n. 78. 16  Madhupiṇḍika Sutta 16. Bhikkhu Ñāṇamoli and Bhikkhu Bodhi, trans., The Middle Length Discourses of the Buddha. A Translation of the Majjhima Nikāya (Somerville, MA: Wisdom Publications, 1995), p. 203. The text repeats these statements for the remaining four sense organs and objects (ear and sounds, nose and odours, tongue and flavours, body and tangibles) and the mind and mind-objects. 17  See R.  L. M.  Gethin, The Buddhist Path to Awakening (Oxford: Oneworld Publications, 2001), p. 41. 18  See Bhikkhu KL Dhammajoti, Abhidharma Doctrines and Controversies on Perception (Hong Kong: Centre of Buddhist Studies, University of Hong Kong, 2007), and Sharf, “Knowing Blue,” op. cit. 19   See Jonardon Ganeri, Philosophy in Classical India (London: Routledge Press, 2001), pp.  101–102; Charles Goodman, “The Treasury of Metaphysics and the Physical World,” The Philosophical Quarterly 54 (2004): 389–401; and Mark Siderits, Buddhism as Philosophy (Surrey, UK: Ashgate Publishing Company and Indianapolis, IN: Hackett Publishing Company, 2007), pp. 113–118. 15

170

E. Thompson

themselves blue quality-particulars.) The pot itself is nothing but a collection or bundle of dharmas. Similarly, a mind is nothing but a bundle of mental dharmas. Indeed, from the ultimate analytical standpoint, there are no pots or minds; there are just bundles of dharmas. It will be important to keep this metaphysical framework in mind as we consider the Abhidharma controversies about perception. On the one hand, the Sarvāstivāda-Vaibhāṣika (henceforth “Vaibhāṣika”) school and the Sautrāntrika school agree that the object is first grasped nonconceptually by the senses and then grasped conceptually by the mind.20 The Sarvāstivāda treatise, Collection on Consciousness (Vijñanakaya-śāstra) expresses this view in its statement that the eye-consciousness (visual awareness) grasps only blue, not “this is blue,” whereas the mental consciousness can grasp both blue and “this is blue,” the latter by way of grasping its name.21 This statement that visual awareness grasps blue, whereas the mind can also grasp “this is blue,” becomes a pericope for Indian Buddhist philosophy.22 Vasubandhu (ca. fourth to fifth centuries C.E.) repeats it in his masterwork of Abhidharma thought, Commentary on the Treasury of Abhidharma (Abhidharmakośabhāṣya), which criticizes the Sarvāstivāda-Vaibhāṣika view from the Sautrāntrika perspective.23 Dignāga, the originator of the Buddhist logical and

 The Sarvāstivāda, “Teaching That All Exists,” was one of the most influential Abhidharma schools. Its name derives from its thesis that all “conditioned factors” (dharmas) continue to exist (sarvam asti) throughout all three time periods of past, present, and future. The Vaibhāṣika (“Followers of the Vibhāṣā”) was a Sarvāstivāda subschool active in northwestern India (Kashmir-­ Gandhāra). It claimed that its principal text (the Abhidharmamahāvibhāsa) was originally spoken by the Buddha. The Sautrāntrika (“Followers of the Sūtras”), who were likely a dissenting Sarvāstivāda offshoot, rejected this claim. They upheld the doctrine of “momentariness” (kṣaṇikavāda), according to which only instantaneous, present-moment activity exists. See Alexander von Rospatt, The Buddhist Doctrine of Momentariness: A Survey of the Origins and Early Phase of this Doctrine Up to Vasubandhu (Stuttgart: Steiner, 1995). 21  The Collection on Consciousness (Vijñanakaya) is thought to have been composed in the first century C.E. but exists only in Xuanzang’s Chinese translation from 649. Dhammajoti, Abhidharma Doctrines and Controversies on Perception, op. cit. p. 108, translates the passage as follows: “The visual consciousness can apprehend only a blue colour (nīlam), but not ‘it is blue’ (no tu nīlam iti). Mental consciousness can also apprehend a blue colour. [But] so long as it is not yet able to apprehend its name, it cannot apprehend ‘it is blue’. When it can apprehend its name, then it can also apprehend ‘it is blue’.” Sharf, “Knowing Blue,” op. cit., also translates this passage (and includes the Chinese): “Eye-cognition is only able to discern blue; it is unable to discern ‘this is blue.’ Mind-cognition is also able to discern blue. As long as [mind-cognition] is unable to discern [the color’s] name, it is unable to discern ‘this is blue.’ But should it be able to discern its name, it discerns, at the same time, both ‘blue’ and ‘this is blue.’ The colors yellow, red, white, and so on are [analyzed] in the same way as the color blue.” Sharf’s translation includes the continuation of the passage with the same formula for “sound” and “this is a sound,” and he notes that the passage goes on to repeat the formula for nose and odour, tongue and taste, body and touch, and that it concludes by saying that “mind-cognition is also able to discern all dharmas.” 22  See Sharf, “Knowing Blue,” op. cit. 23  Leo Pruden, Abhidharmakośa Bhāṣyam by Louis de la Vallée Poussin. Volumes 1-IV (Berkeley, CA: Asian Humanities Press, 1991). These four volumes provide an English translation of Louis de la Vallée Poussin’s French translation of Vasubandhu’s text. See Volume 2, p. 425. 20

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

171

epistemological tradition, also repeats it, in his Commentary on the Compendium of Epistemology (Pramāṇasamuccayavṛtti).24 The basic idea is that each of the five kinds of sensory awareness has access only to its proper sense object (visual awareness to visual objects, hearing to sounds, and so on), whereas the mental awareness has access not only to mental objects but also to the five senses and their objects. In addition, whereas the visual awareness can grasp only blue, the subsequent mental awareness can also grasp “this is blue”—or, as we might say, “that something is blue”25— by associating the name “blue” with the sensory content. On the other hand, the Ābhidharmikas disagree about exactly how the perceptual process works. To bring out their disagreements, recall the passage from the Nikāyas quoted above, which begins, “Dependent on the eye and forms, eye-consciousness arises.” This might mean that the visual sense faculty and the material visual object precede the arising of visual awareness, or it might mean that the three elements are strictly contemporaneous, such that they arise simultaneously. In addition, if they are contemporaneous, is this a case of simultaneous causation or of some kind of metaphysical dependence relation (perhaps the eye-consciousness supervenes on the eye and forms)? Furthermore, what exactly does the seeing? Is it the eye or the eye-consciousness? And when the passage states, “The meeting of the three is contact. With contact as condition there is feeling,” is “contact” another real element (dharma) or just a way of designating the coming together of sense organ, object, and awareness? Finally, even if the sensory awareness grasps merely blue, and not “this is blue,” how exactly is this kind of discrimination supposed to be nonconceptual, given that it requires grasping blue versus yellow or white or red, and so on? The Vaibhāṣikas maintain that “the eye sees,” by which they mean that it is the visual organ that sees, but that it can do so only when associated with visual awareness, which depends on the eye as its base.26 They also maintain that the visual organ, the material object, and the visual awareness—all understood as continuants made up of a series of bundles of momentary quality-particulars—arise simultaneously and are “mutually conjoined causes (saṃprayuktaka-hetu) to one another; the absence of any one renders their very co-nascence impossible; either they arise all together, or none of them can come into existence at all.”27 According to this view, an episode of sensory awareness registers its object directly, that is, without the mediation of concepts and with no time-lag between its occurrence and the occurrence of the object. The Sautrāntrikas, however, deny that “the eye sees.” They maintain that the seeing is nothing other than the momentary event of visual awareness itself.28 According  Masaaki Hattori, Dignāga on Perception. Being the Pratyakṣapariccheda of Dignāga’s Pramāṇasamuccaya. (Harvard Oriental Series 47) (Cambridge, MA: Harvard University Press, 1968). See p. 26 (Daa-2). 25  See Taber, A Hindu Critique of Buddhist Epistemology, op. cit., p. 30. 26  Dhammajoti, Abhidharma Doctrines and Controversies on Perception, op. cit., pp. 51–52. 27  Ibid., p. 96. 28  Ibid., p. 55. 24

172

E. Thompson

to Vasubandhu, there is no agent of seeing; there is just the occurrence of visual awareness as a momentary event in a causal series. We can say that “the visual awareness sees,” but we should not be misled by this form of expression into thinking that the grammatical subject of the statement refers to an agent.29 The Sautrāntrikas also reject the Vaibhāṣika doctrine of simultaneous causation. Rather, the sense faculty and the sense object are the cause of the subsequently arising visual awareness. The momentary awareness occurs after the momentary object; in other words, there is a time-lag between the two, such that the object is no longer present when the visual awareness occurs.30 The subsequent awareness is said to resemble or conform to the previously occurring object; specifically, the awareness is said to bear the “form” or “aspect” (ākāra) of the object.31 Thus, for the Sautrāntrikas, the awareness of the external object is not direct in the way that it is for the Vaibhāṣikas; rather, the awareness registers the object only after it has disappeared and by way of bearing the object’s “form.” In addition, the Vaibhāṣikas maintain that “contact” is a real element (dharma) that occurs simultaneously with the arising of the sense organ, the object, and the sensory awareness. They reason that it must be a real element for it to serve as the cause for the arising of feeling (only real entities as opposed to nominal ones have causal efficacy), and that, given simultaneous causation, it can co-exist in the same moment with the sense organ, the object, and the sensory awareness, while being the effect of their convergence. The Sautrāntrikas reject all of this, arguing that all causal relations require the effect to occur after the cause, that “contact” is just a nominal entity (prajñapti), namely, a way of designating the combination of the sense organ, the object, and the (subsequent) awareness, and that it is this combination that gives rise to feeling. Despite these disagreements, the Vaibhāṣikas and the Sautrāntrikas agree that the full recognition of the object happens only when the mental awareness, which arises subsequent to the sensory awareness and is conditioned by various “mental factors” (caitasika), such as memory and understanding, grasps the content of the sensory awareness. Nevertheless, they disagree about the exact sequence of events and about what is directly versus indirectly grasped. For the Vaibhāṣikas, the sequence has two stages: first, the sensory awareness arises along with the sense organ and the object; and second, the mental awareness occurs. Whereas the sense organ serves as the sense faculty for the sensory awareness, that initial moment of sensory awareness, which is a mental event, is said to serve as the mental faculty for the subsequently arising mental awareness. For the Sautrāntrikas, the sequence has three stages: first, the sense faculty and object arise;  See Birgit Kellner, “Changing Frames in Buddhist Thought: The Concept of Ākāra in Abhidharma and in Buddhist Epistemological Analysis,” Journal of Indian Philosophy 42 (2014): 275–295, at pp. 278–280. 30  For a helpful discussion of the Sautrāntrika “time-lag argument” in relation to the metaphysics of momentariness, see Siderits, Buddhism as Philosophy, op. cit., pp. 130–137. 31  See K.  L. Dhammajoti, “Ākāra and Direct Perception (Pratyakṣa),” Pacific World Journal 3 (2007): 245–272, and Kellner, “Changing Frames in Buddhist Thought,” op. cit. 29

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

173

second, the sensory awareness occurs; and third, the mental awareness occurs. In addition, whereas they both agree that the mental awareness does not directly grasp the sense object (as a momentary occurrent, it is gone by the time the mental awareness arises), the Sautrāntrikas maintain that the sensory awareness, too, does not directly grasp the object; rather, it grasps it only indirectly by way of directly bearing the object-form (ākāra) that is caused to arise at stage two of the sequence. We are now ready to take up the question of what is supposed to be nonconceptual in either version of this sequence. Recall the formula that the visual awareness grasps only blue and not “this is blue.” According to the Ābhidharmikas, the visual awareness per se is the awareness of the momentary quality-particular blue. For there to be the perception this is blue, the subsequent moment of mental awareness must take the name (nāma) “blue” as its object along with the sensory content, namely, the blue quality-particular that is the object of the antecedent visual awareness. Whereas this kind of object (called the viṣaya) lasts only a moment and is gone by the time the mental awareness arises, the name (concept) can persist from one moment to the next. It thereby enables the mental awareness to have same “perceptual object” (called the ālambana) from moment to moment, and thus to support the perceptual judgement, “this is blue.”32 More generally, since the sensory awareness per se does not require or involve the application of a name (concept), it is ­nonconceptual, whereas once the mental awareness takes a name as its object along with the sensory content, the awareness is conceptual. The problem, however, is that the visual awareness must be able to distinguish between blue, green, yellow, red, white, black, and so on, before the application of the corresponding names (concepts) by the mental awareness. What accounts for this ability? More generally, how is sensory awareness able to distinguish between quality-particulars without the application of concepts corresponding to them?33 A generous reading of the Abhidharma answer to this question would see it as trying to differentiate between what we would describe as the sensory discrimination of particulars (the sensory detection of differences between stimuli), the perceptual recognition of kinds, and the mental conceptualization of kinds in linguistic thought; a critical reading would see it as waffling or equivocating. The problem is tied to the semantic range of the Sanskrit word, vikalpa, which is variously translated as “discrimination,” “discriminative construction,” “concept,” “conception,” and “imagination.” Generally speaking, in later Buddhist philosophy, the term vikalpa refers to the conceptual activities of the mind (manovijñāna, the mental consciousness), which is said to operate with mental images (ākāra) whose contents are “generic characteristics” (sāmānyalakṣana), in contrast to direct perception (pratyakṣa), which grasps the particular “inherent characteristic” (svalakṣana) and is free from thought (concept use).34 Given this understanding, it would seem to

 See Dhammajoti, Abhidharma Doctrines and Controversies on Perception, op. cit., pp. 107–109.  See Sharf, “Knowing Blue,” op. cit. 34  See, for example, the entry for vikalpa in Robert E. Buswell and Donald S. Lopez, eds., The Princeton Dictionary of Buddhism (Princeton, NJ: Princeton University Press, 2014), p. 970. 32 33

174

E. Thompson

follow that sensory awareness must be free of vikalpa, and indeed the Abhidharma texts state that this is the case: such awareness is said to be avikalpaka (“devoid of conceptualization”).35 Nevertheless, as we have seen, sensory awareness must be able to distinguish between quality-particulars. The Ābhidharmikas are thus forced to concede that there is a kind of rudimentary vikalpa at work in sensory awareness, which they call “intrinsic vikalpa” or “vikalpa per se” (svabhāva-vikalpa).36 One way to understand this kind of vikalpa is that it is a mere discrimination that simply registers or detects the presence of a given quality-particular, such as the presence of a blue quality-particular, either at exactly the same moment as that particular’s occurrence (according to the Vaibhāṣikas) or immediately afterwards (according to the time-lag reasoning of the Sautrāntrikas). The Ābhidharmikas distinguish between this kind of bare discrimination and two other types of progressively more complex vikalpa, one using “recollection” (anusmarana-vikalpa) and one using “examination” (abhinirūpaṇā-vikalpa). The “vikalpa through recollection” requires being able to hold the object in mind for more than one moment, and the “vikalpa through examination” applies discursive thought to what is thus held in mind. With this threefold distinction in hand, the Ābhidharmikas are able to explain that when sensory awareness is said to be devoid of conceptualization (avikalpaka), what is meant is that it is devoid of the second and third kinds of vikalpa, but not the first kind, which simply registers quality-particulars in a coarse way—as in the case of grasping blue versus yellow, and so on—with no mental interpretation or conceptualization.37 Nevertheless, these distinctions do not so much solve the problem as restate it.38 On the one hand, no “discriminative construction” (vikalpa) is supposed to be present in sensory awareness; on the other hand, “discrimination per se” (svabhāva-­ vikalpa) must be present in sensory awareness. On the one hand, every sensory awareness involves not only “feeling” (vedanā) but also an element of “recognition” (saṃjñā); on the other hand, this kind of recognition is said to be too weak or too coarse to be the same as the kind of recognition that operates in the mental awareness through contact with a name. In other words, one kind of recognition lies on the nonconceptual side of the divide; the other kind of recognition lies on the conceptual side. These manoeuvres do not bridge the divide; they simply restate it. Clearly, the shuffling of terms and proliferation of scholastic distinctions testify to

 See Dhammajoti, Abhidharma Doctrines and Controversies on Perception, op. cit., pp.  104, 135, n. 58. 36  Ibid. 37  Vasubandhu, in his Commentary on the Treasury of Abhidharma (Abhidharmakośabhāṣya), building on the Vaibhāṣika analyses, identifies “vikalpa per se” (svabhāva-vikalpa) with the mental factor called vitarka, which is a rudimentary advertence to an object before more refined discrimination and naming or conceptualization (1.33a-b, 2.33a-b). See Pruden, Abhidharmakośa Bhāṣyam, op. cit., Volume 1, p. 97, 202–204. See also Dhammajoti, Abhidharma Doctrines and Controversies on Perception, op. cit., pp. 97, 105–107; and Sharf, “Knowing Blue,” op. cit. 38  See Sharf, “Knowing Blue,” op. cit. 35

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

175

the recognition of a problem—how to bridge the gap between nonconceptual content and conceptual content—but not to any stable solution. Vasubandhu seems to recognize the problem, for he rationalizes the terminological manoeuvres this way: One says that sensory awareness is free from discriminative construction (vikalpa) “in the same way that when a horse has only one foot, one says that it does not have any feet.”39 The analogy is telling. A horse with one foot does have a foot, but it does not have any footing and cannot run. The problem is to explain how the horse runs— how cognition works—and for that we need an account that puts together nonconceptual sense perception and conceptualized perception. In short, we need a theory of concept formation—what concepts are and how they get formed. The Ābhidharmikas do not provide such an account. For an account of concept formation, we must turn to Dharmakīrti and his commentators. But first we will take a detour through cognitive science.

2 A View from Cognitive Science Looked at from the perspective of cognitive science, the Ābhidharmikas appear to be grappling with the problem of how to understand the relationship between sensation and thought, including how to distinguish among sensory discrimination, perceptual recognition (which requires type-identification or categorization), and linguistic thought. A cognitive science perspective can help us to sharpen our assessment of vikalpa in Abhidharma. It will also prepare the way for showing how Dharmakīrti’s later model of concept formation not only constitutes a decisive advance beyond Abhidharma but also casts new light on familiar issues in cognitive science. First, a comment on the term “consciousness” in these discussions is necessary. The Ābhidharmikas do not make any systematic distinction between what we would call “conscious” experience and “unconscious” sensory or cognitive processing. From their perspective, every moment of sensory or mental awareness has a very brief or instantaneous duration, rapidly follows its predecessor, and immediately conditions the occurrence of its successor. Although most of this discrete mental sequence lies below the threshold of ordinary cognition, and so in that sense is unconscious, it is thought to be observable by a highly refined, meditative mental awareness, and so is not unconscious in the sense of being inaccessible in principle to any kind of mental awareness.40

 Commentary on the Treasury of Abhidharma (Abhidharmakośabhāṣya), 1:33a-b. Pruden, Abhidharmakośa Bhāṣyam, op. cit., Volume 1, p. 97. 40  For further discussion, see Evan Thompson, Waking, Dreaming, Being: Self and Consciousness in Neuroscience, Meditation, and Philosophy (New York: Columbia University Press, 2015), chapter 2. 39

176

E. Thompson

Cognitive scientists, in contrast, generally distinguish between conscious experience, on the one hand, and unconscious processing, on the other. Here “conscious” means either useable in intentionally guided thought and behaviour (including verbal report), or having a qualitative or phenomenal character.41 Thus, “unconscious” can mean either inaccessible to intentionally guided thought and behaviour (including verbal report), or lacking a qualitative or phenomenal character.42 We can now turn to sensory processing. Whether such processing is conscious or unconscious (in one or another sense), we need to distinguish between sensory discrimination and categorical perception. The former tracks differences in stimulus magnitudes (e.g., this amount of brightness versus that amount of brightness); the latter tracks differences between kinds (e.g., blue versus yellow, or turquoise versus aqua). Sensory discrimination is not sufficient for categorical perception. For example, being able to tell the difference (discriminate) between two shades of blue that differ in hue, saturation, or brightness in a pairwise comparison is not sufficient for being able to say which shade is which when either one is presented in isolation. Discrimination is psychophysically defined in terms of the just-noticeable difference (JND) or least perceptible difference, which is the amount a stimulus magnitude must be changed for one to notice a difference at least 50% of the time. (The JND is a statistical measure, because the difference that one notices varies from trial to trial, and hence many trials are needed to determine the JND for a particular subject.) Discrimination, so defined, does not imply that one can repeatedly identify the discriminated stimuli. For example, one may be able to discriminate a particular shade of blue without being able to recognize it immediately afterwards as the shade just presented. Such recognition requires having a perceptual category for the shade and (according to the standard view in perceptual psychology) being able to repeatedly apply a mental representation of that category to what one sees. Thus, recognition, unlike mere sensory discrimination, requires categorical perception. Categorical perception consists in making the same response not only to repeated instances of qualitatively identical presentations (e.g., repeated colour presentations having the same hue, saturation, and brightness) but also to repeated instances of what one takes to be presentations of the same kind (e.g., two presentations of blue), and in

 Being conscious in the first sense is being “access conscious;” being conscious in the second sense is being “phenomenally conscious.” For the “access conscious” versus “phenomenally conscious” distinction, see Ned Block, “Concepts of Consciousness,” in David Chalmers, ed., Philosophy of Mind: Classical and Contemporary Readings (New York: Oxford University Press, 2002), pp. 206–218. 42  Whether there can be qualitative or phenomenal states whose contents “overflow” what is cognitively accessed and available for use in intentionally guided thought and behaviour is a controversial issue. For the “overflow” thesis, see Ned Block, “Perceptual Consciousness Overflows Cognitive Access,” Trends in Cognitive Sciences 12 (2011): 567–575. For a critical rejection of the thesis, see Michael Cohen and Daniel Dennett, “Consciousness Cannot Be Separated from Function,” Trends in Cognitive Sciences 15 (2011): 358–364. 41

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

177

making a different response to what one takes to be a different kind of presentation (e.g., yellow).43 With this distinction between sensory discrimination and perceptual recognition/ categorical perception in hand, we can raise a question about the meaning of vikalpa in Abhidharma. Does discriminating between two shades of blue without being able to recognize them individually count as a case of vikalpa, or does vikalpa require recognition and categorical perception? Admittedly, the question is anachronistic. Nevertheless, it is unavoidable, given our aim of “engaging Buddhism” in contemporary philosophy of mind and given that vikalpa is sometimes translated as “discrimination” and sometimes as “concept.” So, let’s consider the options. On the one hand, if we render vikalpa as “concept,” then we make our Ābhidharmikas skirt incoherence, because the minimal case of vikalpa, namely, svabhāva-vikalpa (“intrinsic vikalpa,” “vikalpa per se”), simply registers the presence of a quality-particular. Concepts, however, have generality; they reach beyond particulars, enabling reidentification and grasping something as an instance of a kind. The Ābhidharmikas explicitly distinguish the minimal case of vikalpa per se from the kind of vikalpa that relies on memory (anusmarana-vikalpa). Memory is required for perceptual recognition and concept application, so if vikalpa per se does not involve memory, then it cannot be conceptual. On the other hand, if we render vikalpa as “discrimination,” then we run the risk of losing sight of the fact that the word vikalpa paradigmatically designates some kind of mental construction. For the Ābhidharmikas, this construction is the mental superimposition of kinds and wholes onto a world consisting ultimately only of impartite quality-particulars. Although the Ābhidharmikas do not exactly distinguish between discrimination and perceptual recognition/categorical perception in the way that we do in cognitive science, they appear to be grappling with something very close to this distinction. Indeed, a charitable way to interpret their terminology is to read it as implicitly making a parallel distinction. Consider again that svabhāva-vikalpa (“vikalpa per se”), which belongs to visual awareness per se, is said to register only blue, and not “this is blue.” It is reasonable to take this statement as meaning that sensory awareness per se only discriminates its object—where this means registering or detecting it—without being able to categorize and recognize it. According to this reading, when the Ābhidharmikas assert that the “eye-consciousness” (visual awareness) grasps only blue and not “this is blue,” they mean that the visual awareness only discriminates (detects) blue but does not recognize “this is blue.” And when they assert that mentally grasping “this is blue” requires that the mental awareness grasp the name “blue,” they mean that recognizing “this is blue” requires applying a mental representation of the hue category blue. From a cognitive science perspective, this mental representation does not have to be linguistic (many nonlinguistic

 See Stephen Harnad, “Categorical Perception,” in Encyclopedia of Cognitive Science (Nature Publishing Group/Macmillan, 2003), pp. 67–74. 43

178

E. Thompson

animals have categorical colour perception44). Similarly, in later Buddhist epistemology (pramāṇavāda), as we will see below, the object of the mental awareness that is needed to grasp “this is blue,” is not a linguistic entity (a word) but is rather a nonlinguistic mental representation (a mental image) that is said to be suited to be associated with words. Recall also that the Ābhidharmikas adhere to the Buddha’s word as recorded in the Nikāyas that the bundle of psychophysical events that make up what we ordinarily think of as a person always contains saṃjñā (recognition). This word gets variously translated as “perception,” “cognition, “recognition,” and “ideation.” Its basic sense is that of noting something based on a distinguishing mark or characteristic.45 In cognitive science terms, this way of grasping an object—via a mark or characteristic that other objects can also be perceived to have—goes beyond mere sensory discrimination and requires categorical perception and the ability to type-identify stimuli. The Ābhidharmikas, however, maintain that the kind of saṃjñā present in mere sensory awareness is too weak or coarse to be the same kind of recognition that operates in the mental awareness through contact with a “name.” Mental contact with the name “blue” enables the mental awareness to grasp that this characteristic is blue, whereas sensory awareness by itself can grasp only the bare particular blue. Here, too, the Ābhidharmikas seem to have arrived at something like our distinction between mere sensory discrimination and repeatable identification through perceptual recognition/categorical perception. But is categorical perception a conceptual process? By definition, it involves the recognition of kinds. So, the question is whether kind recognition suffices for conceptuality. This question cannot be answered apart from an account of what concepts are. In cognitive psychology, a concept is generally described as an organized body of knowledge about a category (a recognizable kind).46 Here “knowledge” is not used in the factive sense (in which to know something implies that it is the case and one cannot know something that is not the case). Rather, it is used to mean content or information retained in memory and useable in perception, cognition, and action.47 Beyond this consensus, however, cognitive psychologists diverge considerably in their use of the term “concept.”  See Evan Thompson, Colour Vision: A Study in Cognitive Science and the Philosophy of Perception (London: Routledge Press, 1995), chapter 4. 45  The formal Theravāda Abhidhamma definition of saññā from Buddhaghosa’s (ca. fifth century C.E.) text, Path of Purification (Visuddhimagga XIV:130) is as follows: “Its function is to make a sign as a condition for perceiving again that ‘this is the same,’ as carpenters, etc., do in the case of timber, and so on. It is manifested as the action of interpreting by means of the sign as apprehended, like the blind who ‘see’ an elephant… Its proximate cause is an objective field in whatever way that appears, like the perception that arises in fawns that see scarecrows as men.” Bhikkhu Ñāṇamoli, trans., Visuddhimagga: The Path of Purification (Kandy, Sri Lanka: Buddhist Publication Society, 1975), p. 464. See also Gethin, The Buddhist Path to Awakening, op. cit., p. 41. 46  See Edouard Machery, Doing Without Concepts (New York: Oxford University Press, 2009), chapter 1. 47  Ibid., p. 8. 44

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

179

Some psychologists and philosophers consider any process of categorization as equivalent to the employment of a concept.48 To categorize is to apply a mental representation of a kind, and such a representation, according to this minimalist viewpoint, meets the minimally sufficient criteria for being a concept, because its content has generality beyond the particular and is reidentifiable. For example, categorizing a visual stimulus as blue suffices for possessing a concept of blue, and the perceptual recognition of blue is accordingly a perception with conceptual content. (As we will see, Dharmakīrti holds a version of minimalism, because he holds that sense perception is always nonconceptual and simply discriminates or registers its object without categorizing it; categorization is the work of conceptualization and requires “mental construction” added to or superimposed onto nonconceptual sense perception.)49 A more stringent understanding of concepts is that categorization is necessary but not sufficient for concept possession. The principal additional requirements are (i) that the body of knowledge or information about a category be retained by long-­ term memory; (ii) that the knowledge or information be available for use by higherorder cognitive processes (thinking, reasoning, planning, judging, etc.);50 and (iii) that concepts have a compositional structure, so that more complex concepts can be formed by putting together their simpler constituents. For my purposes here, it will be useful to split the difference, as it were, between the minimalist and the stringent accounts, by describing categorical perception as “proto-conceptual.” A sensory process is proto-conceptual if it categorizes stimuli into kinds (equivalence classes) that are repeatedly recognizable and can be used or relied on in perception and action. Proto-conceptuality is necessary for

 See Jesse Prinz, Furnishing the Mind: Concepts and Their Perceptual Basis (Cambridge, MA: The MIT Press, 2002), and Eric Mandelbaum, “Seeing and Conceptualizing: Modularity and the Shallow Contents of Perception,” Philosophy and Phenomenological Research 97(2) (2017): 267–283. 49  See John D. Dunne, “Key Features of Dharmakīrti’s Apoha Theory,” in Siderits et al., Apoha, op. cit., pp. 84–108, and Georges Dreyfus, “Apoha as a Naturalized Account of Concept Formation,” in Siderits et al., Apoha, op. cit., pp. 207–227. 50  Thus, Machery, Doing Without Concepts, op. cit., p. 12, describes how most psychologists understand what a concept is as follows: “A concept of x is a body of knowledge about x that is stored in long-term memory and that is used by default in the processes underlying most, if not all, higher cognitive competencies when those processes result in judgments about x.” Machery makes the important point that concepts, so defined, are heterogeneous, because the relevant bodies of knowledge can be organized in different ways—around prototypes (typical examples), exemplars (remembered instances of a category), theories (miniature theories about a given domain), or perceptual schemata. An individual typically has several heterogeneous kinds of concepts for each category or kind it recognizes. On this basis, Machery argues that the very notion of concept should be eliminated from the theoretical vocabulary of cognitive science, because the heterogeneity of these differently organized bodies of knowledge implies that the class of concepts is not a natural kind. A more moderate view is “concept pluralism:” A given concept can have multiple types of structure or each type of structure can be a concept on its own. See Stephen Laurence and Eric Margolis, “Concepts and Cognitive Science,” in Eric Margolis and Stephen Laurence, eds., Concepts: Core Readings (Cambridge, MA: MIT Press, 1999), pp. 3–81. 48

180

E. Thompson

conceptuality, but I allow that it may not be sufficient, especially for the philosophical meanings of “concept” I discuss later. I wish to call attention to three points about proto-conceptuality as a way of bridging back to Abhidharma and forward to the pramāṇavāda (Buddhist epistemology) and Yogācāra views still to be discussed. First, proto-conceptual processes structure our sensorimotor engagement with the world at a basic and prelinguistic level. We and other animals come equipped with sensorimotor systems that sort the flux of the world into stable, recognizable kinds. For example, our visual system, as well as the visual systems of many other animals, sorts the region of the electromagnetic spectrum to which it is visually sensitive into stable and discrete hue categories. Second, experimental studies indicate that the perceptual categorization of objects can happen very quickly (13–18  ms),51 and that it usually happens at the basic level (e.g., blue, banana) rather than at the superordinate level (e.g., colour, fruit).52 This evidence supports a view of perception in which categorization is a very early part of the perceptual process, rather than being a later cognitive operation performed on “raw,” uncategorized (and hence non-proto-conceptual) perceptual content.53 Finally, our recognitional abilities are affectively biased. We orient toward and selectively attend to perceptual kinds or categories (e.g., happy faces, angry faces) that are affectively salient to us as a result of our evolutionary and developmental histories, a phenomenon known as “affect-biased attention.”54 Affect-biased attention is inseparable from approach-versus-avoid action tendencies.55 It is also possible that the perceptual response to an object’s affective salience does not occur after the object has been identified, but rather that affective responses support perception from the moment that stimulation begins.56 These points are readily stateable in the Abhidharma terminology. When sensory “contact” (sparśa) occurs, so too do “feeling” (vedanā) and “perceptual recognition” (saṃjñā), along with “attentional orienting” (manaskāra), “intention” (cetanā, an intentional or volitional mental or bodily action tendency), and “awareness”

 Mary C. Potter, Brad Wyble, Carl Erik Hagmann, and Emily S. McCourt, “Detecting Meaning in RSVP at 13 ms per Picture,” Attention, Perception and Psychophysics 76 (2014): 270–279. 52  Mary C. Potter and Carl Erik Hagmann, “Banana or Fruit? Detection and Recognition Across Categorical Levels in RSVP,” Psychonomic Bulletin Review 22 (2015): 578–585. 53  See Mandelbaum, “Seeing and Conceptualizing,” op. cit. 54  Rebecca M. Todd, William A. Cunningham, Adam K. Anderson, and Evan Thompson, “Affect-­ Biased Attention as Emotion Regulation,” Trends in Cognitive Sciences 7 (2012): 365–372. 55  Jelena Markovic, Adam K. Anderson, and Rebecca M. Todd, “Tuning to the Significant: Neural and Genetic Processes Underlying Affective Enhancement of Visual Perception and Memory,” Behavioural Brain Research 259 (2013): 229–241. 56  Lisa Feldman-Barrett and Moshe Barr, “Seeing It with Feeling: Affective Predictions During Object Perception,” Philosophical Transactions of the Royal Society of London B Biological Sciences 364 (2009): 1325–1334. 51

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

181

(vijñāna). Not every element in this package considered on its own is (in my terms) proto-conceptual. “Awareness” (vijñāna) per se and “feeling” (vedanā) per se are nonconceptual; they discriminate but do not categorize (“awareness” simply registers the bare presence of something, and “feeling” is just a sensation with a hedonic tone of pleasant, unpleasant, or neutral). Nevertheless, the whole package is proto-­ conceptual, due to its ineliminable recognitional/categorical element (saṃjñā). Although we have now sharpened the terms of the Abhidharma framework with the help of cognitive science, we have not yet addressed the Abhidharma problem of how to bridge the gap between nonconceptual content and conceptual content. Given the Abhidharma metaphysics, this problem is especially acute, because the gap to be bridged is between the discrete and momentary awareness of evanescent quality-particulars and the apparently persisting cognition of apparently persisting and reidentifiable perceptual objects belonging to apparently persisting recognizable kinds. Addressing this problem requires not just the distinction between sensory discrimination and perceptual recognition/categorical perception, but also an account of how discriminatory and recognitional/categorical activities contribute to concept formation. What are the generative principles or mechanisms required for producing conceptual cognitions in a world of particulars? The Ābhidharmikas repeatedly run up against this problem but do not have the means to answer it.57 To make progress we need to turn to later Buddhist epistemology (pramāṇavāda), specifically the apoha or “exclusion” theory of the Yogācāra-Sautrāntrika philosophical movement.58

3 The Apoha Theory The Buddhist philosopher Dignāga (ca. 480–540 C.E.) introduced the Sanskrit term apoha, which means “exclusion,” in the context of “Buddhist nominalism” and the problem of universals.59 According to Buddhist nominalism, reality consists of nothing but unique and momentary quality-particulars. The human mind, however, has a deeply entrenched tendency to carve up the world into the fictional mental constructs of enduring substantial things belonging to distinct kinds. This happens through a kind of “lumping” and “splitting.” We lump some particulars together

 See Sharf, “Knowing Blue,” op. cit.  See Siderits, et al., Apoha, op. cit. 59  See Hattori, Dignāga on Perception, op. cit. See also Siderits, et al., Apoha, op. cit., especially the “Introduction” by Chakrabarti and Siderits. “Buddhist nominalism” is a term used by historical and textual scholars; it is not a term native to the Indian and Tibetan Buddhist philosophical tradition. See also Mark Siderits, Studies in Buddhist Philosophy (New York: Oxford University Press, 2016), pp. 138–160. 57 58

182

E. Thompson

while splitting them off from everything else. The core idea of the apoha theory is that we cluster some particulars together while overlooking their mutual differences and excluding them from everything else. For example, we call some quality-­ particulars “blue” by overlooking their mutual differences and excluding them from everything else. According to the apoha theory, what unites them is not any real universal “blueness,” but rather that they are all not nonblue. The collection of nonblue things is large and heterogeneous, and there is no reason to suppose that there is any universal “nonblueness,” which the things lumped into the collection share. Blue, however, is simply the exclusion of nonblue, so similarly there is no reason to suppose that the things lumped together as “blue” all share a universal; on the contrary, these things are all different from each other in numerous ways, but we ignore these differences by uniting them under the concept “blue.” More generally, the basic idea of the apoha theory is that the extension of a concept consists in what is excluded from that which is other, where the operation of “exclusion” (apoha) is understood in psychological terms to be an ignoring of differences and in logical terms to be a kind of negation. An example from Dharmakīrti can help to illustrate the basic idea.60 Several different medicinal plants can be used to reduce fever, and they work differently to produce this effect. They have no single common intrinsic property; what they have in common is just that they cause fever to be lowered in us. We take an interest in this effect, so we ignore all the ways that they differ and use the single word “antipyretic” to refer to them. This word designates a class formed by exclusion, namely, those things that are not non-fever-reducing. To think that the members of this class must all share some intrinsic property in common, namely, “antipyreticness,” is to project our cognitive interests onto the world and erroneously to reify an exclusion. It has become standard to distinguish between “top-down” versus “bottom-up” versions of the apoha theory.61 A top-down approach uses the tools of logic, such as negation operators, to work from linguistic meaning and the generality of concepts down to real particulars, which lack any generality or shared natures, while avoiding any commitment to real universals. A bottom-up approach works from pure particulars up to concepts via causal chains from particulars to the generation of sensed resemblances and to judgements of sameness and difference as the core of

60   The example comes from Dharmakīrti’s major work, Commentary on Valid Cognition (Pramāṇavārttika), I. 74. See Vincent Eltschinger, John Taber, Michael Torsten Much, and Isabelle Ratié, Dharmakīrti’s Theory of Exclusion (apoha). Part 1. On Concealing. An Annotated Translation of Pramāṇavārttikasvavṛtti 24, 16–45, 20 (Pramāṇavārttika 1.40–91). Studia Philologica Buddhica. Mongraph Series XXXVI (Tokyo: The International Institute for Buddhist Studies of the International College for Postgraduate Buddhist Studies, 2018), p.  85. See also Siderits, Buddhism as Philosophy, op. cit., pp. 221–222. 61  Tom Tillemans, “How to Talk About Ineffable Things: Dignāga and Dharmakīrti on Apoha,” in Siderits et al., Apoha, op. cit., pp. 50–63.

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

183

conceptual cognition. Dignāga’s original version of the apoha theory takes a topdown approach, but Dharmakīrti transforms the theory into a bottom-up account.62 The bottom-up account is the one that concerns us here. It has been interpreted as offering a “naturalized” account of concept formation,63 and, as such, it helps us to tighten the links between Buddhist theories of concepts and cognitive science. For this purpose, the following elements of Dharmakīrti’s account deserve emphasis. First, because of our bodily constitution and the historical formation of our dispositional tendencies,64 we spontaneously experience sensory similarities and dissimilarities in the incessant flux of events. These resemblances and differences are not mind-independently real, but rather are a function of how we spontaneously evaluate things, given our constitution and how things affect us.65 Second, such resemblances provide a basis for the cognitive construction of equivalence classes through the ignoring of differences and the exclusion from that which is other (anyāpoha).66 The most primitive form of an equivalence class is what cognitive scientists would call a “perceptual category” (an object category that can be perceptually recognized, such as “cow,” or a quality category, such as “blue”), which enables type-identification, recognition, and perceptual judgement (e.g., “that’s a cow,” “this is blue”). (Here it is important to avoid terminological confusion. For Dharmakīrti, sense perception—pratyakṣa—is nonconceptual and what we ordinarily think of perception is thoroughly conceptual, because the categories we recognize in perception are constructed via exclusion. Cognitive scientists use the term “sensation” to refer to what Dharmakīrti means by pratyakṣa.)

 “The major change that happens between Dignāga and Dharmakīrti is the use of a causal approach to link language to the world. This causal chain from particulars to perception and finally thought and language is entirely absent in Dignāga and so constitutes a substantial evolution in the theory: indeed, it is arguably a new Apohavāda [apoha theory] that Dharmakīrti has devised,” ibid., pp. 54–55. Dignāga’s theory was criticized by the Brahmanical philosophers Kumārila (seventh century C.E.) and Uddyotakara (sixth century C.E.). The basic objection is that the theory is circular: to form the idea “non-cow,” one must already have the idea “cow,” so explaining “cow” as “not non-cow” presupposes knowing the meaning of “cow.” For these arguments and the Buddhist responses, see Pascale Hugon, “Dharmakīrti’s Discussion of Circularity,” Masaaki Hattori, “The Apoha Theory as Referred to in the Nyāyamañjarī,” and Prabal Kumar Sen, “The Apoha Theory of Meaning: A Critical Account,” in Siderits et al., Apoha, op. cit., pp. 109–124, 134–148, and 170–206, respectively. Dharmakīrti’s bottom-up approach was extensively elaborated in Tibet. For the Tibetan reception of Dharmakīrti, see Georges Dreyfus, Recognizing Reality: Dharmakīrti’s Philosophy and Its Tibetan Interpretations (Albany, NY: State University of New York Press, 1997). 63  See Dreyfus, “Apoha as a Naturalized Account of Concept Formation,” op. cit. 64  For the role that karmic history plays in Dharmakīrti’s account, see Catherine Prueitt, “Karmic Imprints, Exclusion, and the Creation of Worlds of Conventional Awareness in Dharmakīrti’s Thought,” Sophia 57 (2018): 313–335. 65  See Dreyfus, “Apoha as a Naturalized Account of Concept Formation,” op. cit. 66  See Jonardon Ganeri, “Apoha, Feature-Placing, and Sensory Content,” in Siderits et al., Apoha, op. cit., pp. 228–246. Ganeri relates apoha to the concept of a quality space as discussed by Austen Clark in his Sensory Qualities (New York: Oxford University Press, 1993) and A Theory of Sentience (New York: Oxford University Press, 2000). 62

184

E. Thompson

Third, Dharmakīrti provides a principled answer to the problem of the “scope of the conceptual” or “how the conceptual/nonconceptual distinction should be drawn.”67 The answer consists in providing a way to determine whether a cognitive process is conceptual (involves concept application) and whether mental content is conceptual or nonconceptual: a cognitive process is conceptual if and only if it involves the exclusion operation (apoha), and mental content is conceptual if and only if its content is constructed via exclusion. Since perceptual categories (in the cognitive-science sense) are constructed by exclusion from that which is other, they are conceptual. Thus, Dharmakīrti holds a minimalist view of concepts, according to which object and quality recognition suffice for conceptual cognition. Fourth, a concept is a mental particular (a representation) whose content can be described in cognitive science terms as prototypical (structured via a prototype) in the sense that it presents its object as belonging to a kind according to the object’s having a range of expected qualities and characteristics (rather than satisfying a definition in terms of necessary and sufficient conditions). Fifth, conceptual cognition is motivationally and affectively biased. Concept formation—the construction of equivalence classes via exclusion—always happens under the influence of past experience and in the approach-versus-avoid context of obtaining the desirable and avoiding the undesirable.68 We are motivated both to focus on certain causal effects as being relevant to our goals and to overlook others as being irrelevant. As a result, conceptual cognition is cognitively and affectively interest-relative and context-dependent. Sixth, it follows that conceptual cognition necessarily has a subject-object structure. It presents the world as consisting of objects belonging to determinate kinds that can be cognitively grasped by the subject. Seventh, concepts in the minimalist sense—mental representations of object and quality categories constructed via exclusion—are (as Dharmakīrti puts it) “fit to be associated with words” or “capable of being conjoined with discourse.”69 The mental representation of object and quality categories does not have to be verbally expressed or designated, or linguistic in form, but language in general and words for kinds or sortals in particular require such mental representations (minimalist concepts). Thus, Dharmakīrti defines conceptual thought (kalpanā) as “a cognition in

 Laurence and Margolis, “The Scope of the Conceptual,” op. cit., p. 291.  See Dunne, “Key Features of Dharmakīrti’s Apoha Theory,” op. cit., p. 93, and Prueitt, “Karmic Imprints, Exclusion, and the Creation of World’s in Dharmakīrti’s Thought,” op. cit. 69  See Tom J. F. Tillemans, “Dharmakīrti on Prasiddha and Yogyatā,” in Agata Bareja-Starzyńska and Marek Mejor, eds., Aspects of Buddhism: Proceedings of the International Seminar on Buddhist Studies (Warszawa, Poland: Instytut Orientalistyczny, 1997), pp. 161–176.

67

68

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

185

which a representation is fit to be associated with a verbal designation.”70 His commentators take this definition as implying that there are concepts not actually associated with words and that very young infants and nonhuman animals have such concepts, because they have recognitional abilities but lack language. Thus, minimal conceptuality does not require language; it requires only type-identification and recognition of something as an instance of a kind. Finally, strictly speaking, according to this Dharmakīrtian view, conceptual cognition is erroneous, because there is no sameness outside of our conceptual constructions via exclusion. Each conceptual cognition, considered as a mental event, is a unique, nonrepeatable particular, but considered in terms of its content, involves a representation of sameness. But there is no real sameness “out there” or “in here,” so taking sameness to be real is mistaken. It also follows, according to this view, that the subject-object structure of conceptual cognition is erroneous, for there are no enduring subjects and objects, but only evanescent particulars connected via cause-­ effect relations. Nevertheless, conceptual cognitions can guide effective action if they successfully track causal regularities. What follows is a bare-bones outline of Dharmakīrti’s bottom-up apoha theory.71 Our example will be the cognition, “this is blue,” which we can imagine thinking or uttering whenever we are presented with a blue patch, such as a blue paint chip or a blue light stimulus in a colour discrimination test. Three streams of discrete and causally efficacious momentary events interact to produce this cognitive episode: the causal stream of material particulars constituting the perceived object; the causal stream of material particulars constituting the sense faculty; and the causal stream of mental particulars constituting the mind.72 The theory combines a causal account with an error theory: First, a particular object arises in contact with the visual sense faculty at a given moment.

 Dharmakīrti, Drop of Reasoning (Nyāyabindu), I. 5 and Ascertainment of Epistemology (Pramāṇaviniścaya), I. Ibid., p. 166. Dunne translates this sentence as “a concept is a cognition with a phenomenal appearance that is capable of being conjoined with a linguistic expression” (“Key Features of Dharmakīrti’s Apoha Theory,” op. cit., p. 87); Dreyfus translates it as “conceptualization is that consciousness in which a representation [lit., appearance] is fit to be associated with words” (“Apoha as a Naturalized Account of Concept Formation,” op. cit., p. 219). See also Taber, A Hindu Critique of Buddhist Epistemology, op. cit., pp. 207–208. 71  I rely here on Dreyfus, Recognizing Reality, op. cit., pp.  217–232; Dreyfus, “Apoha as a Naturalized Account of Concept Formation,” op. cit.; Dunne, Foundations of Dharmakīrti’s Philosophy, op.  cit., pp.  116–144; and Dunne, “Key Features of Dharmakīrti’s Apoha Theory,” op. cit. 72  Dunne, “Key Features of Dharmakīrti’s Apoha Theory,” op. cit., p.  87. Strictly speaking, the three causal series of particulars are all that ultimately exists; they constitute the perceived object, the sense faculty, and the mind, respectively, only because we conceptualize and conventionally refer to them that way. 70

186

E. Thompson

Second, this sensory contact causes an image having the “phenomenal form” or “phenomenal aspect” (ākāra) of the object to arise in the mind in the immediately subsequent moment.73 Third, this “phenomenal representation,”74 which is a mental particular and is the immediate object of sensory awareness, in conjunction with other mental factors— namely, our interests, expectations, and “traces,” “imprints” or “residual impressions” (vāsanās) from previous experience—causes us to judge in the next moment, “this is blue.” The third step is conceptual because it regards things that are in fact unique as being the same via exclusion; the first two steps are free of the exclusion operation and the type-identifying of particulars, and so are nonconceptual. Moreover, at the third step, error occurs, because the perceiver mistakenly thinks that the content of the judgement is not just a mentally created exclusion but also reflects how the world is in itself. Thus, conceptual cognition, despite being pragmatically efficacious in serving our interests, is inherently erroneous. We overlook the differences that exist between all particulars, think in terms of common properties by mentally constructing kinds through exclusion, and naively take these kinds to be mind-­ independently real rather than recognizing them as our own inventions. Let us now put some metaphysical and epistemological flesh on this bare-bones account. For something to be real, it must be a causally efficacious particular. Causal efficacy (arthakriyā) is the criterion for being real and only particulars have causal efficacy.75 A particular’s causal efficacy is not a property it possesses, for particulars are unities in which there is no difference between property and property-possessor. What a particular does is no different from what it is; a particular’s causal efficacy is itself particular and is the same as that particular.76 Every causally efficacious

73  For discussion of the differences between the concept of ākāra in Abhidharma and in Buddhist epistemology (pramāṇavāda), see Kellner, “Changing Frames in Buddhist Thought,” op. cit. 74  I use this term to render “ākāra,” which is variously translated as “image,” “perceptual image,” “cognitive image,” “representation,” “phenomenal form,” and “phenomenal aspect.” 75  Dharmakīrti, Commentary on Valid Cognition (Pramāṇavārttika), III.3 (the additions in brackets are from Manorathanandin’s twelfth century commentary): “Whatever has causal powers (arthakriyāsamartha), that really exists (paramārthasat) in this context [i.e., when we examine reality]. Anything else is declared to be [just] customarily existent (saṃvṛtisat) [because it is practically accepted through mere conceptual fictions]. These two [i.e., the real and the customary] are [respectively] particulars and universals.” Translated by Tom Tillemans, “Dharmakīrti,” The Stanford Encyclopedia of Philosophy (Spring 2017 Edition), Edward N. Zalta (ed.), URL = https:// plato.stanford.edu/archives/spr2017/entries/dharmakiirti/, Section 1.2. Dunne, Foundations of Dharmakīrti’s Philosophy, op. cit., p. 392, translates this passage as follows: “In this context, that which is capable of telic function is said to be ultimately real. The other one is said to be conventionally real. They are, respectively, the particular and the universal.” 76  Thus, Dharmakīrti and Buddhist epistemologists generally, like their Abhidharma predecessors, can be understood as trope theorists. See the references in note 19 above and Tillemans, “Dharmakīrti,” op. cit., Section 1.4.

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

187

particular is unique, momentary, and nonrepeatable.77 As a result of our history, bodily constitution (which itself is nothing other than a collection of causally related particulars), and how particulars affect us, we spontaneously experience sensory similarities and dissimilarities in the incessant flux of events. These experienced similarities and dissimilarities are not objective, that is, they do not exist “out there” apart from us. Rather, they are products of our interaction with the world. Put another way, there are no real similarities and dissimilarities; there are only evaluations as of similarity and dissimilarity. Such evaluations, however, are not arbitrary, for they are naturally induced as a result of how causally efficacious particulars affect us, given our constitution and history. To use an example familiar to cognitive scientists, numerous physically distinct combinations of light wavelengths can be perceived as blue, and the same physical stimulus magnitudes can be perceived as dissimilar colours depending on the context. These experienced similarities and dissimilarities are a function of how physical stimuli, given their causal efficacy, affect us, given our neurophysiological constitution and evolutionary and developmental history. Such naturally experienced similarities and dissimilarities provide the nonarbitrary basis on which object and quality categories and linguistic concepts are elaborated. To cognize “this is blue” requires more than just being able to discriminate blue in the sense of being differentially sensitive to it among other visual qualitative similarities and dissimilarities. As the Ābhidharmikas say, seeing blue is not the same as discerning “this is blue.” The latter requires being able to cognize multiple instances as being the same in respect of blue, which is to say that it requires type-­ identification, categorical perception, and recognition. So there needs to be a way to apply the unique content of a particular phenomenal representation (ākāra) across multiple instances. In other words, there needs to be a way for the content to be “repeatable” or to have “continuity” or “distribution” (anvaya), so that it can be used to construe two or more things as the same and type-identify them. Cognizing “this is blue” is a case of recognition. It requires construing the present visual content as the same as a previously experienced visual content, so that they count as two tokens of the same type. There are two sides to this construal, according to the bottom-­up apoha theory.78 On the one hand, any given phenomenal representation is a unique particular that arises from unique causes and has unique effects, and thus is excluded from all other particulars, which have their own unique causes and effects. It thereby serves as the basis for excluding the phenomenal representations produced by other particulars. On the other hand, its content must somehow be generalized and distributed across a range of particulars to type-identify them. This generalization happens because the content is taken to resemble the contents of other previous representations, so that their objects are taken to be of the same kind.  For Dharmakīrti’s arguments for the momentariness of all that exists—that particulars cannot endure longer than one instant—see Dreyfus, Recognizing Reality, op. cit., pp.  60–65; Dunne, Foundations of Dharmakīrti’s Philosophy, op. cit., pp. 91–97; and Tillemans, “Dharmakīrti,” op. cit., Section 1.3. 78  See Dunne, “Key Features of Dharmakīrti’s Apoha Theory,” op. cit., pp. 92–93. 77

188

E. Thompson

For Dharmakīrti, this construal of items that are in fact unique as the same because they are taken to resemble each other is what conceptuality is all about. The third stage of his account tells us how us how this conceptual cognition comes about. Recall that particulars are causally efficacious and we take an interest in certain effects. Some physical particulars cause us to see blue—to have a phenomenal representation of blue—and we take an interest in this effect, whether explicitly (say, in painting a room) or implicitly by selectively orienting and attending to blue under certain conditions. Since the physical stimuli are unique, it is not the case that they share the common nature of causing us to see blue. Each stimulus brings about this effect in its own specific way, but we overlook these differences given our interest in the effect and our conditioning from previous experience. We are sensitive to the effect but not to the differences in the objects that cause it. As a result of our conditioning, when a particular phenomenal representation occurs, it activates a “trace,” “imprint,” or “residual impression” (vāsanā) from a previous experience, so that the object causing the present experience (the present phenomenal representation) is taken to be the same as the object that caused the previous experience (the previous phenomenal representation). More precisely, the content of the present phenomenal representation is construed to be the same as the content of the previous phenomenal representation, and hence their objects are taken to be the same. All that these objects actually share, however, is their difference from those things that do not bring about this effect, and this difference is just a negation. More precisely, all that these phenomenal representations actually share is their difference from those representations that do not activate this imprint, and this difference is just a negation. Thus, the sense of sameness across multiple instances is based on applying a particular type of negation, namely, the exclusion (negation) of those representations that do not have the expected effect. It is this exclusion (negation) that applies across multiple instances, despite every phenomenal representation being a unique particular. Thus, construing things as the same—type-identifying them—comes from combining that which is not distributed (lacks repeatability/anvaya) with that which is distributed. The phenomenal representation as a unique particular lacks repeatability or distribution. The exclusion as a negation is repeatable and has distribution across many phenomenal representations. Neither the phenomenal representation alone nor the negation alone is sufficient for type-identification. Rather, it is the phenomenal representation construed through the exclusion of that which does not have the expected effects that provides the basis for type-identification.79 To say that certain things that we call “blue” are excluded from other things because those other things do not have the expected effects—for example, they do not produce a phenomenal representation with a “cool” feeling but rather with a “warm” one—implies that all those things that are excluded have the same effects (the ones we expect). But what exactly are these same effects? We cannot just say that they consist in the production of the same kind of phenomenal representation, because that would imply that there is some pre-existing type-identification of

79

 Ibid., p. 94.

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

189

phenomenal kinds, and we are assuming that phenomenal representations are unique mental particulars and have no common natures. In other words, if we take the objects that we call “blue” to be the same because we take their effects—the phenomenal representations they bring about—to be the same, what is the basis for our taking these phenomenal representations to be the same, given that they are supposed to be unique particulars? Dharmakīrti’s answer is that we take the phenomenal representations to be the same because they all cause us to judge perceptually “this is blue,” where this judgement is conditioned by antecedent and historically formed dispositions.80 In other words, we take the representations to be the same because they all produce the same effect, namely, the same perceptual judgement. The perceptual judgement takes as its object the phenomenal content of the representation and construes it as being the same in each case. The construal of sameness depends on the imprints left by previous experiences. Perceptually judging “this is blue” is a case of recognition, because the perceptual judgement takes as its object the content of a present phenomenal representation and construes it as the same as the content of a previous representation whose imprint the present representation activates. As John Dunne explains: “It is the placement of this imprint that crucially allows for the fundamental ‘unification’ or ‘construal as the same’ (ekīkaraṇa) which is the principal marker of conceptual cognition in Dharmakīrti’s system.”81 Thus, according to Dharmakīrti’s bottom-up apoha theory, the causal chain runs from unique particulars (grouped together because they are excluded from other things that do not have the expected effects) to unique phenomenal representations (grouped together because they are excluded from other representations that do not have the expected effects) to the making of the same perceptual judgement. Going in the reverse top-down direction, the resulting sameness of perceptual judgement warrants taking the phenomenal representations to be the same (type-identifying them), and taking the phenomenal representations to be the same warrants taking their objects to be the same (type-­ identifying them). This answer to the question of how we take the phenomenal representations to be the same, however, raises another question: what makes the perceptual judgements the same? If objects are taken to be the same only because they produce the same effects, namely, phenomenal representations that are taken to be the same, and phenomenal representations are taken to be the same only because they produce the same effects, namely, the same perceptual judgement, then what warrants taking the perceptual judgements to be the same? If we say that the perceptual judgements are taken to be the same only because they in turn produce the same effects, then we look to be facing a vicious infinite regress. Dharmakīrti stops the regress because he maintains that the perceptual judgements are taken to be the same not because they have the same effects, but simply  Ibid., pp. 94–99; Dunne, Foundations of Dharmakīrti’s Philosophy, op. cit., pp. 109–126; and Prueitt, “Karmic Imprints, Exclusion, and the Creation of Worlds of Conventional Awareness in Dharmakīrti’s Thought,” op. cit. 81  Dunne, “Key Features of Dharmakīrti’s Apoha Theory,” op. cit., p. 100. 80

190

E. Thompson

because they present their contents as the same; in his terms, they “cover over” each phenomenal content with a “nondifference.”82 Once we arrive at perceptual judgement, and thus conceptual cognition, the differences between the phenomenal representations that cause the perceptual judgements are simply overlooked. The content of the phenomenal representation, as an object of perceptual judgement/ conceptual cognition, is said to lack the vividness or clarity it has as a sensory representation of its object, and so it is easily conflated with the contents of other representations, especially given the priming or biasing that occurs as a result of the activation of the imprints of previous experiences.83 As philosophers often say, “explanations must come to an end somewhere,”84 and here the end is to assert that it is just the nature of particulars to cause us to make perceptual judgements of sameness and difference as a result of the causal chain described. Śākyabuddhi (ca. 660–720 C.E.), an early commentator on Dharmakīrti, gives an analysis of the three meanings of “exclusion of other” (anyāpoha), which we can use to summarize the bottom-up apoha theory.85 The first meaning is the excluded particular. It is excluded from that which is other in the sense that it is unique and completely distinct from every other particular. The second meaning is the exclusion itself—the “mere exclusion” or “mere negation” that is distributed or continuous across all its instances. It is a purely mental construction, a selective differentiation, based on taking one of the differentiations afforded by the unique particular and biased by an imprint, and making it into a kind or type. The third meaning is the excluded phenomenal representation (ākāra), which serves as the basis for the overlooking of differences, the construction of sameness through the selective negation or exclusion from that which is other, and the object of perceptual judgement/recognition/conceptual cognition. The bottom-up apoha theory thus purports to give a comprehensive account of conceptual cognition, beginning with the object of sense perception (the particular), which gives rise to a mental representation, which in turn gives rise to a perceptual judgement (a conceptual cognition), which in turn supports action.86 At the heart of this account is what John Dunne calls a “Janus-faced” theory of concepts.87 A concept is a combination of a mental representation—the phenomenal representation— and a mental operation—the exclusion or negation. Considered as a mental event, a concept is a mental particular. Considered as a mental operation, a concept is a  Dharmakīrti, Commentary on Valid Cognition (Pramāṇavārttika), I.109 and his auto-­commentary Pramāṇavārttikasvopajñavṛtti (also called Svavṛtti), translated by Dunne, ibid., p. 97, and Dunne, Foundations of Dharmakīrti’s Philosophy, op. cit., pp. 123–124. 83  See Dunne, “Key Features of Dharmakīrti’s Apoha Theory,” op. cit., pp. 87, 101. 84  Ludwig Wittgenstein, Philosophical Investigations, trans. G.  E. M.  Anscombe (Oxford: Basil Blackwell, 1958), §1, p. 3. 85  See Shōryū Katsura, “Apoha Theory as an Approach to Understanding Human Cognition,” in Siderits et al., Apoha, op. cit., pp. 125–133, and Dunne, Foundations of Dharmakīrti’s Philosophy, op. cit., pp. 131–144. 86  Dunne, Foundations of Dharmakīrti’s Philosophy, op. cit., pp. 127–128, 131–132. 87  Dunne, “Key Features of Dharmakīrti’s Apoha Theory,” op. cit., p. 104. 82

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

191

negation or exclusion that is distributed over and applies to multiple instances. Considered phenomenologically, a concept is a mental image, that is, conceptual cognition as a subjective experience is an experience mediated by a mental image.88 Cognizing “this is blue” involves recalling a particular mental image (recallable thanks to an imprint) that serves as an exemplar from memory and a prototype of what one expects, and on the basis of which one forms the appropriate exclusion (non-blue) and its complement (the class of things like the paradigm), thereby enabling recognition (matching the perceptual object to the remembered image/ paradigm).89 It is important to note that the subjective experience of conceptual cognition is not of the present image as such, that is, of the present image as a mental particular. Indeed, were one to have a direct experience of the present phenomenal representation as a unique mental particular without selectively excluding it from that which is other, then the experience would be nonconceptual. Rather, in the subjective experience of conceptual cognition, one experiences the content of the phenomenal representation as indistinct from the recalled paradigm and one erroneously takes that generic content to be a real object belonging to a real kind. Our initial reason for examining Dharmakīrti’s account was to see how it can be read as a solution to the Abhidharma problem about the relation between nonconceptual awareness and conceptual cognition. The bottom-up apoha theory constitutes a decisive advance beyond Abhidharma because it gives an account of concept formation that includes a principled criterion for specifying which experiences are conceptual and which are nonconceptual. If it is possible and sometimes legitimate to speak of philosophical progress, relative to a tradition of inquiry and its concerns, as I believe it is, then the bottom-up apoha theory of concept formation qualifies as a case of philosophical progress. Moreover, the theory continues to have importance for us today. It is possible, as we have seen, to describe the theory in cognitive science terms and to present it as providing an answer to the question of the scope of the conceptual. Conceptual mental states are ones that result from exclusion; nonconceptual mental states are ones that do not undergo exclusion. Recognizing “this is blue” is conceptual; merely seeing (detecting or discriminating) blue is nonconceptual. The difference between these two kinds of experience is not explained using fractionations of the one theoretical term vikalpa (svabhāva-vikalpa/vikalpa per se versus anusmarana-­vikalpa/ vikalpa-cum-recollection and abhinirūpaṇā-vikalpa/vikalpa-cum-­examination). Rather, it is explained in terms of the presence versus absence of causal chains involving exclusion (apoha).

 “Likewise, the sameness established by exclusions actually applies to the cognitive images; hence, our conceptual cognitions are actually presenting an image as the same as other images, and not a particular as the same as other particulars.” Dunne, Foundations of Dharmakīrti’s Philosophy, op. cit., p. 139. 89  See Chakrabarti and Siderits, “Introduction,” op. cit., p. 36 and Parimal G. Patil, “Constructing the Content of Awareness Events,” in Siderits et al., Apoha, op. cit., pp. 149–169. 88

192

E. Thompson

Unlike the Abhidharma viewpoint, according to the bottom-up apoha account, grasping “this is blue” does not, strictly speaking, require grasping the name “blue,” but it does require having a mental representation (ākāra) that is suited to be associated with the name “blue,” though this association is entirely conventional (a different word could have been used in setting up our colour-naming conventions). The bottom-up apoha theory thus purports to bridge the gap between nonconceptual content and conceptual content by providing a naturalistically interpretable account of concept formation, according to which causal chains from particulars to mental events and residual impressions that function as exemplars and prototypes lead to the formation of perceptual categories that are fit to be linked to words through linguistic conventions. Of course, Dharmakīrti’s theory raises a whole host of new questions, which were the focus of intense analysis and debate over many centuries in India and Tibet, and have sparked new discussions using the tools of modern logic and formal semantics.90 This attention testifies to the theory’s continuing importance, far beyond its being merely an advance on Abhidharma in the Buddhist philosophical tradition, and thus is another indication of the philosophical progress the theory achieved. My intention is not to defend Dharmakīrti’s extreme nominalism—his refusal to accept that there are real (mind-independent) resemblances. Rather, my interest is in his idea that exclusion—the selective ignoring of differences—undergirds conceptual cognition. My aim is to bring this idea, and the apoha theory of concept formation more generally, to bear on philosophy of mind and cognitive science. The rest of this paper will pursue this aim, beginning with the conceptualist versus nonconceptualist debate in contemporary philosophy.

4 Conceptualism Versus Nonconceptualism In the simplest terms, conceptualism is the thesis that all mental states are conceptual states, and nonconceptualism is the contradictory thesis that not all mental states are conceptual states. Stated this way, it is easy to see that the debate is hostage to how we specify what concepts are. Although I will focus on experiential mental states, we can allow for unconscious mental states (according to the two senses of “unconscious” mentioned above, namely, inaccessible to intentionally guided thought and behaviour, and lacking a qualitative or phenomenal character). Whether mental content must be conceptual or can be nonconceptual is the usual  For coverage of these questions and approaches, see Siderits et  al., Apoha, op.  cit. See also Dreyfus, Recognizing Reality, op. cit., and Dan Arnold, Brains, Buddhas, and Believing: The Problem of Intentionality in Classical Buddhist and Cognitive-Scientific Philosophy of Mind (New York, NY: Columbia University Press, 2012). 90

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

193

way of putting the issue between conceptualism and nonconceptualism. To say that it must be conceptual means that the subject cannot have mental contents without having the concepts required to specify them; to say that it can be nonconceptual means that the subject can have mental contents while lacking the concepts required to specify them. Nevertheless, the conceptual-versus-nonconceptual distinction can be drawn in another way, as between kinds of mental states rather than kinds of mental contents.91 According to this way of drawing the distinction, the debate is about whether there are mental states that do not depend on conceptual states or that have different causal or functional roles from those of conceptual states (even if they do not have a different type of content from conceptual states). My concern, however, is with mental content. Moreover, given that content differences between mental states can be understood as a special kind of state difference, if there are mental states with nonconceptual content, it follows that there are nonconceptual mental states.92 Hence, I will understand the two positions as follows: according to conceptualism, a subject’s conceptual capacities determine the contents of all its mental states (all its mental contents are a function of the concepts it possesses); according to nonconceptualism, a subject’s conceptual capacities do not determine the contents of all its mental states (some of its mental contents are not a function of the concepts it possesses). The philosophical notion of a concept at play in this debate differs from the cognitive science notion of a concept as an organized body of knowledge. Philosophers describe concepts as the constituents of thought. Although there is no generally accepted philosophical account of what concepts are, we can point to some generally agreed upon requirements. First, concepts have generality, that is, they reach beyond particulars by enabling one to recognize a particular as an instance of a general kind of thing. Second, concepts are systematically related, that is, being able to use concepts implies being able to apply the same concept to different things and different concepts to the same thing. Third, concepts are freely combinable to form new thoughts. Finally, these requirements, taken together, imply that concepts are compositional, that is, that their formal structure and semantic content are a function of the formal structure and semantic content of their parts. Gareth Evans’s “generality constraint” is often cited in this context.93 His idea is that to possess a given concept one must be able to use it in thought in all the meaningful combinations into which it could enter with the other concepts one

 See Richard Heck, “Nonconceptual Content and the Space of Reasons,” Philosophical Review 109 (2000): 483–523. 92  See Laurence and Margolis, “The Scope of the Conceptual,” op. cit., pp. 292–293. 93  Gareth Evans, Varieties of Reference (Oxford: Oxford University Press, 1992), pp. 100–105. 91

194

E. Thompson

possesses.94 Thus, if one can think that a is F and b is G, one must be able to think that a is G and b is F. The generality constraint has been taken to imply that conceptual content presents the world as being made up of states of affairs, that is, as being made up of individual things belonging to kinds and standing in various relations to each other. It follows that if experience can present the world as not carved up in this way, that is, as not carved up into states of affairs that serve as the truth-conditions for propositional judgements, then experience can be said to have nonconceptual content.95 Nonconceptualists accordingly point to various ways that our experience of the world outstrips or does not depend on our ability to specify what we experience (the contents of our experience) in terms of propositional judgements. The following kinds of experience are the ones usually mentioned: (1) As discussed above, we can discriminate many particular determinate qualities, such as particular shades of colour, that we cannot hold in memory and re-identify (type-identify and recognize).96 (2) Our experience of being located in a global and orientable space, whose directions are intrinsic and body-centred, is not a function of our conceptual capacities, but rather is a form of intuition (in the Kantian and Husserlian sense of an experience that presents the object itself without the mediation of thought), and is a condition of possibility for the perceptual recognition and conceptualization of things in space.97 (3) Conscious experience is for a subject; it belongs to the content of a conscious experience that it is given to the subject. This kind of subjectivity or experiential self-specification is independent of having a self-concept and being able to use the first-person pronoun.98 (4) In absorbed, skillful pursuits and “flow

 Atomists about concepts, such as Jerry Fodor, reject this idea and maintain that whether one has a particular concept is independent of the other concepts one may have. See Jerry Fodor, Concepts: Where Cognitive Science Went Wrong (Oxford: Oxford University Press, 1998). See also Jerry Fodor, “The Revenge of the ‘Given’,” in Brian P.  McLaughlin and Jonathan Cohen, eds., Contemporary Debates in Philosophy of Mind (Oxford: Blackwell Publishing, 2007), pp. 105–116. For discussion, see Katalin Balog, “Jerry Fodor on Non-conceptual Content,” Synthese 170 (2009): 311–320. 95  See Adrian Cussins, “Content, Conceptual Content, and Nonconceptual Content,” in York H.  Guenther, ed., Essays on Nonconceptual Content (Cambridge, MA: MIT Press, 2003), pp. 133–164. 96  See Diana Raffman, “On the Persistence of Phenomenology,” in Thomas Metzinger, ed., Conscious Experience (Thorverton, UK: Imprint Academic, 1995), pp.  293–308. Imprint Academic; Sean Kelly, “The Non-conceptual Content of Perceptual Experience: Situation Dependence and Fineness of Grain,” Philosophy and Phenomenological Research 62 (2001): 601–608. 97  See Robert Hanna, “Kant and Nonconceptual Content,” European Journal of Philosophy 13 (2005): 247–290; Robert Hanna, “Kantian Non-conceptualism,” Philosophical Studies 137 (2008): 41–64; and Lucy Allias, “Kant, Nonconceptual Content, and the Representation of Space,” Journal of the History of Philosophy 47 (2009): 383–413. 98  See José Luis Bermudez, The Paradox of Self-Consciousness (Cambridge, MA: MIT Press, 1998); Dan Zahavi, Subjectivity and Selfhood: Investigating the First-Person Perspective (Cambridge, MA: MIT Press, 2005); and Kristina Musholt, Thinking About Oneself: From Nonconceptual Content to the Concept of a Self (Cambridge, MA: MIT Press, 2015). 94

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

195

states,” we are acutely sensitive to the world and have an experiential, activity-based knowledge, but one that presents the world as a milieu of attractions and repulsions, not as a domain of objects and properties that would stand as the truth-conditions of judgements.99 (5) The background affect of existential moods, such as anxiety, despair, and joy, affect how the whole world seems or feels, but not in a way that seems adequately expressible by any propositional judgement.100 It is important to note that the nonconceptualist motivation here is to call attention to forms of experience that do not show the kind of generality, systematicity, and freely recombinable structure of conceptual thought and its expression in linguistic judgement. The motivation need not be to provide an immediate, nonconceptual, ground of justification for knowledge, so there need be no allegiance to the so-called epistemological “myth of the given.”101 In response to these kinds of nonconceptualist considerations, John McDowell has revised the conceptualist position.102 He proposes that for an experience to be the exercising of a conceptual capacity, it is not necessary that the experience have a propositional content. Rather, its content can be a structured unity while being nondiscursive. Nevertheless, he maintains that such an experience is conceptual in its being conceptualizable. Although much of our lives may proceed without conceptualized thoughts, we can nonetheless “embrace” all of it in thought. McDowell earlier argued that we can form “demonstrative concepts,” such as “that blue,” to pick out particular qualities and that such concepts are as fine-grained as the perceptual experiences are.103 The conceptualist might also observe that we can conceptualize and describe in detail the experiential contours of body-anchored, egocentric space versus allocentric space; that we have a rich conceptual repertoire for describing and teaching skillful activities; and that we have various concepts of emotions and moods that we bring to bear in describing how we experience the world when we are anxious, depressed, or elated. Nonconceptualists, however, argue that this line of thought is beside the point. On the one hand, if to “embrace” such experiences in thought means just that we can think about them, such embrace does not imply that the experiences themselves  See Cussins, “Content, Conceptual Content, and Nonconceptual Content,” op. cit.; Hubert Dreyfus, “Overcoming the Myth of the Mental: How Philosophers Can Profit from the Phenomenology of Everyday Expertise,” Proceedings and Addresses of the American Philosophical Association 79 (2005): 47–65; Hubert Dreyfus, “The Myth of the Pervasiveness of the Mental,” in Joseph Shear, ed., Mind, Reason, and Being-in-the-World: The McDowell-Dreyfus Debate (London: Routledge Press, 2013), pp. 15–39. A locus classicus for this idea is Maurice Merleau-­ Ponty, The Structure of Behavior, trans. A.  Fisher (Pittsburgh, PA: Duquesne University Press, 1963), pp. 168–169. 100  See Matthew Ratcliffe, Feelings of Being: Phenomenology, Psychiatry, and the Sense of Reality (Oxford: Oxford University Press, 2008). 101  See Susan Hurley, “Non-conceptual Self-consciousness and Agency: Perspective and Access,” Communication and Cognition 30 (1997): 207–248. 102  John McDowell, “Avoiding the Myth of the Given,” in his Having the World in View: Essays on Kant, Hegel, and Sellars (Cambridge, MA: Harvard University Press, 2009), pp. 256–272. 103  John McDowell, Mind and World (Cambridge, MA: Harvard University Press, 1994). 99

196

E. Thompson

are conceptual or concept-dependent. Nonconceptualists maintain that not everything presented in an experience is conceptualized; to accept this thesis is tantamount to accepting that there is nonconceptual content.104 Whether we are always in a position to go on to conceptualize such content, where that means to make use of it in discursive thought and judgement, is a further question. On the other hand, if to “embrace” in thought means to present such contents themselves to thought, then concepts are insufficient in the absence of having the relevant nonconceptual experiences. For example, according to the nonconceptualist reading of Kant, the nonconceptual intuition of space is a necessary condition of possibility for being able to cognize objects in space.105 Furthermore, in the case of fine-grained perceptual discriminations, the nonconceptualist argues that even if we can use a demonstrative to indicate “that blue” in the presence of an instance of a particular shade of blue, we cannot deploy that demonstrative content recognitionally. In other words, we cannot use it to reidentify that particular blue no matter how brief the time interval between the first and second presentations. This limitation on the content calls into question its credentials as a concept. In Dharmakīrti’s terms, the mental content lacks distribution (anvaya); it cannot be distributed or repeated over its instances, so it fails the crucial test for being a concept. To put the point another way, our ability to discriminate that particular shade of blue is independent of our ability to remember it, so it cannot be the case that our ability to experience it depends on our having a recognitional demonstrative concept for it; therefore, the content is nonconceptual. The conceptualist may reply that this line of reasoning is faulty because it commits the “conceptualization fallacy,” namely, “supposing that when conceptualization occurs given a prior representational state that the prior state isn’t itself conceptual.”106 The thought here is that “the conceptualization involved needn’t be based on an unconceptualized state, but might itself be a matter of reconceptualization (i.e., a move from one type of conceptualized representation to a different conceptualization).”107 Although we should guard against this faulty reasoning, it does not apply in the present context. We are supposing that concepts are either systematic and freely combinable constituents of thought, or mental representations that can be held in memory to enable perceptual recognition. Conceptualists bear the burden of showing that the kinds of experiential content at issue are conceptualized in one of these ways and then reconceptualized. At a minimum, they must show that the experiential content is generalizable or has repeatability (anvaya) across a range of instances, so that it can be used to type-identify particulars. The bottom-up apoha theory provides a useful perspective on this debate because of the way that it delineates the scope of the conceptual. Type-identification or  See Tim Crane, “The Given,” in Joseph Shear, ed., Mind, Reason, and Being-in-the-World, op. cit., pp. 229–249. 105  See Hanna, “Kant and Nonconceptual Content,” op. cit. and “Kantian Non-conceptualism,” op. cit.; and Allias, ““Kant, Nonconceptual Content, and the Representation of Space,” op. cit. 106  Laurence and Margolis, “The Scope of the Conceptual,” op. cit., p. 295. 107  Ibid. 104

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

197

recognition of something as an instance of a kind is necessary and sufficient for conceptual cognition. Merely seeing (discriminating) this blue is not a conceptual cognition and does not require possessing the concept “blue.” Recognizing this blue is a conceptual cognition and requires possessing a concept of blue. Such a concept is constructed via exclusion and is repeatable or has distribution across a range of instances. Mere demonstrative content—detecting this blue—is not sufficient for possessing a recognitional demonstrative concept that enables one to re-identify the shade in a subsequent moment. Possessing such a concept requires “mental construction” (kalpanā) through the bottom-up exclusion (apoha) process. Buddhist epistemologists (Yogācāra-Sautrāntrikas) are nonconceptualists, both according to the above definition of nonconceptualism (mental contents are not entirely determined by the concepts one possesses) and according to their own account of conceptual cognition. They hold that perception (pratyakṣa) is representational and that not all representational content is conceptual content. Perception is of its object as that object is phenomenally presented in a mental image, but the mental content is not conceptualized until the image is conflated with the content of a residual impression and undergoes exclusion from that which is other. This Buddhist version of nonconceptualism is unique and offers a perspective not found in our contemporary debate. On the one hand, experience contains and causally depends on nonconceptual content. On the other hand, ordinary experience of the world—the content of everyday experience taken at face value—is thoroughly conceptualized, because it is constructed by the exclusion process. Hence, we are hardly ever aware of nonconceptual content as such. Indeed, such awareness would require either taking the exclusion process offline, as it were, which requires inhibiting or overcoming the motivational and affective biases that drive exclusion, or it would require allowing the exclusion process to operate but without getting caught up in its falsification of reality, that is, without mistaking its inherently erroneous representations for how things really are.108 We can put these ideas in Kantian terms. Ordinary perceptual experience results from what Kant calls “synthesis,” the putting together of representations to produce a single mental content. For the Buddhist epistemologists, synthesis or “mental construction” (kalpanā) takes the form of the moment-to-moment conflation of the content of a particular mental image with the contents of residual impressions functioning as expectation-recognition exemplars and prototypes. For Kant, synthesis in general is the result of the imagination and is not the same as conceptualizing (only the “synthesis of recognition in a concept” involves concept application, and whether the two other forms of synthesis—“apprehension in intuition” and “reproduction in imagination”—are concept-dependent is a contentious interpretative

 For further discussion, see Paul J. Griffiths, “Omniscience in the Mahāyānasūtrālaṅkāra and Its Commentaries,” Indo-Iranian Journal 33 (1990): 85–120; John D. Dunne, “Thoughtless Buddha, Passionate Buddha,” Journal of the American Academy of Religion LXIV (1996): 525–556; and C.  Ram-Prasad, “Conceptuality in Question: Teaching and Pure Cognition in Yogācāra-­ Madhyamaka,” Religious Studies 36 (2000): 277–291. 108

198

E. Thompson

issue).109 For the Buddhist epistemologists, however, all aggregation involves mental construction. Only momentary particulars ultimately exist but the mind (the stream of momentary mental particulars) creates the impression of there being enduring substances that come sorted into determinate kinds. This kind of “mental construction” (kalpanā) is inextricably both imaginational (it conflates mental images) and conceptual (it deploys the logical operation of exclusion/negation and iterates it across multiple instances).110 It is also fundamentally erroneous, because it presents a false representation of the world. It follows that a nonerroneous or undistorted cognition would have to be nonconceptual (though, as we will see, it does not follow that a cognition is undistorted simply in virtue of being nonconceptual or that there cannot be distorted nonconceptual cognitions). But what exactly would such a cognition be like for the subject? I now turn to this question.

5 Nonconceptual Awareness The Ornament of the Great Vehicle Sutras (Mahāyānasūtrālamkāra), a major work of Yogācāra philosophy traditionally attributed to Maitreyanātha (“Protector Maitreya,” an epithet of the future Buddha, Maitreya) and said to be received by the philosopher Asaṅga (ca. fourth century C.E.), Chapter XI: 31, states: False imagination, neither correct Nor false, not conceptual, Not conceptual and not nonconceptual— This characterizes all objects of cognition.111

The principal Indian and Tibetan philosophical commentators—Vasubandhu, Sthiramati (475–555 C.E.), and Mipham (1846–1912)—explain this verse using a threefold distinction between (i) ordinary dualistic awareness, (ii) nonconceptual and nondualistic awareness, and (iii) awareness subsequent to nonconceptual and nondualistic awareness. “False imagination” refers to ordinary cognition, which is dualistic. Everything in ordinary awareness is falsely imagined to be really divided into a cognitive subject—the “grasper” (grāhaka)—and an object of cognition— that which is “grasped” (grāhya). Liberation or awakening requires overcoming this fundamental delusion. “Neither correct nor false” refers to the period when one hears and reflects on the Buddha’s teaching. One’s cognition is incorrect, because it is still dualistic, but it is not entirely ignorant, because it is informed by the Buddha’s   See Allias, Kant, Nonconceptual Content, and the Representation of Space,” op. cit., pp. 396–397. 110  Thus, it is not surprising that translators of Buddhist philosophical texts sometimes translate vikalpa and kalpanā as “imagination.” For further discussion, see Matilal, Perception, op. cit., pp. 312–313, and Urban and Griffiths, “What Else Remains in Śūnyatā?”, op. cit. 111  Ornament of the Great Vehicle Sūtras. Maitreya’s Mahāyānasūtrālamkāra with Commentaries by Khenpo Shenga and Ju Mipham, trans. Dharmachakra Translation Committee (Snow Lion, 2014), p. 327. 109

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

199

teaching that the subject-object dualism is a false mental imposition. “Not conceptual” refers to the fully awakened, direct, nonconceptual realization of all phenomena (including that very realization) being “empty” (śūnya) of subject and object. Thus, “nonconceptual” here means “nondualistic.” Finally, “not conceptual and not nonconceptual” refers to the so-called “mundane wakefulness” that is attained in the wake of the nonconceptual, supreme awakening. This condition is called the “subsequent awareness” or “subsequently obtained worldly awareness.” It enables the awakened one (the bodhisattva) to act and teach in the everyday world. It is not conceptual, because it does not engage in any dualistic reification of subjects and objects. Nevertheless, since it completely understands and directly cognizes all phenomena as empty of subject and object, it is not merely nonconceptual; it is also nondualistic. Exactly what this means, however, especially in the case of the Buddha, who is said to have attained a state free of all conceptualization and desire, is far from clear and proved to be a very thorny philosophical and religious issue for the tradition.112 What could motivate the Buddha to teach, if he lacked all desire, and how could he teach anything without employing conceptual cognition? My purpose here is not to delve into those difficulties. Rather, I am going to assume that Yogācārins are committed to nondualistic awareness being an attainable experience, and not just a soteriological ideal. Nevertheless, I would like to distinguish between this kind of experience in its normatively and soteriologically ideal type—the sustained nondualistic awareness of which the Buddha and advanced bodhisattva are supposed to be capable—and the possibility of shorter, maybe even rather brief, episodes of nonconceptual awareness with an attenuated subject-object duality, of which we ordinary beings may be capable.113 Similarly, I would like to distinguish between the ideal form of the “subsequent awareness” and the possibility of more ordinary forms, which would ensue after brief moments of nonconceptual awareness with an attenuated subject-object duality. I propose that we may be able to use the philosophical and cognitive science ideas already presented to make sense of such moments of nonconceptual awareness and subsequent awareness, where by “make sense,” I mean in a broadly naturalistic and psychologically plausible way. Vasubandhu is the philosopher responsible for using the phrase, “grasped and grasper” (grāhyagrāhaka), specifically as a gloss on the term, “duality.”114 Yogācārins assert that the nature of reality (dharmatā) is nondual and that dualities are unreal mental fabrications. Vasubandhu takes the grasper-grasped duality as the fundamental unreal and mentally fabricated duality. “Grasper” refers to  See the references in note 108, especially Dunne, “Thoughtless Buddha, Passionate Buddha,” op. cit. See also Roy Tzohar, “Enjoy the Silence: The Relation Between Non-conceptual Awareness and Inexpressibility According to Some Yogācāra Sources,” paper presented at the workshop, “Conceptuality and Nonconceptuality in Buddhist Thought,” University of California, Berkeley, Center for Buddhist Studies, November 4–6, 2016. 113  I am indebted to Jowita Kramer for discussion of this idea. 114  See Jonathan Gold, Paving the Great Way: Vasubandhu’s Unifying Buddhist Philosophy (New York: Columbia University Press, 2015), pp. 158–169. 112

200

E. Thompson

­consciousness (vijñāna), understood as a causally produced moment of awareness, and “grasped” refers to the object that appears to consciousness. Described phenomenologically, the grasper-grasped structure is the apparent subject-object structure of ordinary awareness; more precisely, it is the apparent subject-object structure of intentionality, especially of what philosophers today call “phenomenal intentionality,” the mental directedness upon an object that is taken to be characteristic of (and, according to some philosophers, constituted by) consciousness.115 Vasubandhu rejects one of the key, agreed-upon assumptions of the contemporary conceptualist versus nonconceptualist debate, namely, that the subject-object form of intentionality or mental-directedness is a real, intrinsic feature of episodes of conscious awareness or cognition and that it accurately reflects reality. Rather, for Vasubandhu and the Yogācārins in general, subject-object intentionality is an illusory distortion of and imaginary superimposition onto phenomena that lack this structure. For the Yogācārins, ordinary intentional experience is delusional; it not only experiences things in terms of the subject-object structure but also takes this structure to be an accurate representation of reality.116 My concern here is not to examine the Yogācāra arguments for this viewpoint, but rather to see what sense can be made of the possibility of a less than fully enlightened form of experience with an attenuated grasper-grasped structure. As we saw in our discussion of the bottom-up apoha theory, conceptual cognition, according to this theory, necessarily has a subject-object structure, because it presents the world as consisting of objects belonging to determinate kinds that are cognitively grasped by the subject. So, one might be led to think that the subject-object dualism results from conceptualization, specifically that it is a function of “exclusion” (apoha). The thought would be that the mental operation of “exclusion from that which is other” is responsible for the subject-object structure. Although the (non-Buddhist) Pratyabhijñā Śaiva philosophers, Utpaladeva (ca. 925–975 C.E.) and Abhinavagupta (ca. 975–1025 C.E.), who build on Dharmakīrti, take this route and consider that the mere exclusion of subject from object, and vice-­ versa, makes the subject-object dualism a conceptual construction, this is not Dharmakīrti’s view.117 On the contrary, he maintains that the grasper-grasped duality is nonconceptual. In other words, he maintains that the subject-object structure is an unreal appearance or illusion that already distorts nonconceptual awareness before conceptualization happens. In Dharmakīrti’s terms, the grasper-grasped duality arises from an “internal distortion” (antarupaplava) of the mind, not from misconstruing the object (the momentary particular) through a perceptual

 For a study of intentionality and consciousness in Buddhist philosophy, see Coseru, Perceiving Reality, op. cit. 116  For an important discussion of this theme, see Jay Garfield, “Vasubandhu’s Treatise on the Three Natures: A Translation and Commentary,” in Garfield, Empty Words, op. cit., pp. 128–151. 117  See Catherine Prueitt, “Shifting Concepts: The Realignment of Dharmakīrti on Concepts and the Error of Subject/Object Duality in Pratyabhijñā Śaiva Thought,” Journal of Indian Philosophy 45 (2017): 21–47. 115

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

201

judgement of sameness.118 So, the absence of the grasper-grasped duality entails the absence of conceptuality, but the absence of conceptuality does not entail the absence of the grasper-grasped duality. Conceptual cognition requires the subject-­ object structure, because conceptuality requires the mental construction of equivalence classes “out there” versus equivalence classes “in here.” For example, recognizing (type-identifying) this blue requires both that it is seen as being qualitatively the same as that (previously seen) blue, and that the subject of this seeing is taken to be the same as the subject of that (previous) seeing, or to use a more minimalist Buddhist formulation, that this seeing and that seeing are taken at least to belong to the same psychophysical causal series. Nevertheless, according to Dharmakīrti and the Yogācārins in general, the presence of the subject-object structure does not by itself entail the presence of conceptuality, because the subject-­ object structure is already present as an “internal distortion” of nonconceptual awareness. Dharmakīrti, when writing from a Yogācāra viewpoint, likens the presence of the grasper-grasped duality to the apparent presence of floaters in the visual field.119 The error or distortion is nonconceptual, because it is already given in the particular phenomenal image, prior to combining the image with an exclusion (vyāvṛtti, apoha) and making a perceptual judgement of sameness based on the supposed similarity of the content of the image to that of previous experiences. Whereas ocular floaters are caused by a defect in the eye, the apparent presence of the grasper-­ grasped duality is caused by a defect in the mind (the mental causal series), namely, our fundamental ignorance (avidyā). Someone who lacks the concepts to think about the distortion caused by having ocular floaters can still suffer from their presence and the visual impairment that they cause. Similarly, the grasper-grasped duality already distorts ordinary awareness before conceptual cognition ensues. Thus, nonduality, the absence of the dualistic appearance of subject and object, which occurs when ignorance is eliminated, implies nonconceptuality, but mere nonconceptuality, the absence of the construction of sameness/difference classes through exclusion—does not imply nonduality. Nonduality requires transcending ignorance and the dualistic appearance of subject and object (the grasper-grasped duality), which are already present in ordinary nonconceptual awareness prior to conceptualization. This discussion suggests that one way to interpret the “subsequent awareness” would be to understand it as somehow allowing for awareness of and action in the world but without the presence of the dualistic illusion of subject and object. A further question would be whether the “subsequent awareness” could involve a kind of recognitional cognition and use of language that somehow do not presuppose or rest on this dualistic illusion. Perhaps there is a way to allow conceptualization to  See Dunne, Foundations of Dharmakīrti’s Philosophy, op. cit., pp. 88–89, 315–317.  Commentary on Valid Cognition (Pramāṇavārttika), III. 217. See Dunne, Foundations of Dharmakīrti’s Philosophy, op. cit., p. 409. For further discussion, see Prueitt, “Shifting Concepts,” op. cit., pp. 23–34. For a discussion of Buddhist philosophical thinking about perceptual illusion versus cognitive errors, see Coseru, Perceiving Reality, op. cit., pp. 182–191. 118 119

202

E. Thompson

return—or, as we might also say, to come back online—but without instinctively, as it were, grasping onto it as real, that is, without immediately and spontaneously reifying it. In any case, my concern is less with the nondual awareness and subsequent awareness as soteriological ideals and more with whether we can make sense of brief and limited or attenuated versions of these types of awareness. I propose the following hypothesis. There can be moments of alert and acute awareness with highly attenuated affective and motivational biases of attention, highly attenuated attentional selection and attentional inhibition, and highly attenuated approach-versus-avoid action tendencies. To the extent that such attenuation occurs, the habitual grasper-grasped duality of experience will be weakened. In other words, there will be less habitual interpretation of mental content as presenting an independent subject encountering an independent object, and more experience of mental events as transitory mental occurrences situated within a field of embodied awareness. In this way, the dualistic subject-object structure of phenomenal intentionality will be reduced and there will be less reification of mental content. Such moments of experience may arise spontaneously or be induced through various practices, can be of variable duration, and can have effects on subsequent cognition and action. From a cognitive science perspective, this hypothesis is testable. Whether the “subsequent awareness” arising from such experiences is psychologically beneficial, according to one or another conception of well-being, is also testable. Whether it is soteriologically beneficial is not empirically testable, but rather is a question for religion and philosophy. Empirical evidence from studies of the effects of mindfulness meditation practices on pain supports my hypothesis. The experience of pain has a sensory aspect, an affective aspect, and a cognitive aspect. The sensory aspect consists of the discriminability, localization, and felt intensity of the pain stimulus; the affective aspect consists of the emotional and motivational significance of the stimulus and the degree of experienced unpleasantness; and the cognitive aspect consists of the meaning of the stimulus—how the stimulus is appraised and the memories and anticipations it evokes. Studies examining mindfulness meditation, specifically the type of practice that cognitive scientists call “open monitoring” meditation,120 which cultivates a nonreactive awareness without attentional selection and attentional inhibition, indicate that this type of practice influences the sensory, affective, and cognitive aspects of pain in meditators compared to control subjects.121 The findings include long-term Tibetan Buddhist practitioners, long-term Zen practitioners, long-term Vipassana practitioners, and meditation-naïve individuals who receive  See Antoine Lutz, Heleen A.  Slagter, John D.  Dunne, and Richard J.  Davidson, “Attention Regulation and Monitoring in Meditation,” Trends in Cognitive Sciences 12 (2008): 163–169; Antoine Lutz, Amishi P.  Jha, John D.  Dunne, and Clifford D.  Saron, “Investigating the Phenomenological Matrix of Mindfulness-Related Practices from a Neurocognitive Perspective,” American Psychologist 70 (2015): 632–658. 121  See Joshua A. Grant, “Meditative Analgesia: The Current Status of the Field,” Annals of the New York Academy of Sciences 1307 (2014): 55–63. 120

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

203

four days of training in a secular breath-awareness form of mindfulness practice.122 A general finding across all the long-term mindfulness practitioners is a significant decrease in the self-reported unpleasantness of painful stimuli and a correlated reduction in neural activity in brain areas known to be associated with affective and cognitive aspects of pain, compared to control subjects, together with either no significant difference in the pain sensory intensity ratings between the meditators and the control subjects (in the case of the Tibetan Buddhist and Vipassana meditators) or reduced sensory intensity ratings for the meditators (in the case of the Zen practitioners). In other words, during the cultivation of an open, nonselective, and nonbiased awareness, stimulus intensity is accurately discriminated while experienced unpleasantness is reduced. Importantly, this effect does not occur during the practice of “focused attention” meditation, in which one places one’s attention selectively on an object, such as the breath or a mental image, and thereby in effect distracts oneself from the pain sensation. A further finding is that the long-term meditation practitioners, but not the control subjects, show an increase in neural activity in brain areas known to be associated with the sensory aspect of pain. This finding is striking because “no other known pain modulator reduces pain and at the same time results in increased activity of pain-related brain regions.”123 Finally, in the study with the meditation-naïve individuals, mindfulness practice led to reductions in both pain-intensity ratings and unpleasantness ratings, with reduced neural activity in sensory brain areas and increased activity in areas associated with affective and cognitive aspects of pain. This pattern of results compared to those found for the long-term meditators has been interpreted as suggesting “that beginners cannot inhibit automatic appraisal of their experience and actively reappraise it within the framework of mindfulness, whereas more experienced practitioners may actually achieve something closer to no appraisal.”124

 For the Tibetan Buddhist practitioners, see David M.  Perlman, Tim V.  Solomon, Richard J. Davidson, and Antoine Lutz, “Differential Effects of Pain Intensity and Unpleasantness of Two Meditation Practices,” Emotion 10 (2010): 65–71; Antoine Lutz, David R.  McFarlin, David M.  Perlman, Tim V.  Salomons, and Richard J.  Davidson, “Altered Anterior Insula Activation During Anticipating and Experience of Painful Stimuli in Expert Meditators,” Neuroimage 64 (2013): 538–546. For the Zen practitioners, see Joshua A.  Grant and Pierre Rainville, “Pain Sensitivity and Analgesic Effects of Mindful States in Zen Meditators: A Cross-Sectional Study,” Psychosomatic Medicine 71 (2009): 1–6-114; Joshua A. Grant, Jérôme Courtemanche, and Pierre Rainville, “A Non-Elaborative Mental Stance and Decoupling of Executive and Pain-Related Cortices Predicts Low Pain Sensitivity in Zen Meditators,” Pain 152 (2011): 150–156. For the Vipassana practitioners, see Tim Gard, Britta K. Hölzel, Alexander T. Sack, Hannes Hempel, Sara W. Lazar, Dieter Vaitl, and Ulrich Ott, “Pain Attenuation Through Mindfulness is Associated with Decreased Cognitive Control and Increased Sensory Processing in the Brain,” Cerebral Cortex 22 (2012): 2692–2702. For the meditation-naïve individuals, see Fadel Zeidan, Katherine T. Martucci, Robert A. Kraft, Nakia S. Gordon, John G. McHaffie, and Robert C. Coghill, “Brain Mechanisms Supporting the Modulation of Pain by Mindfulness Meditation,” Journal of Neuroscience 31 (2011): 5540–5548. 123  Grant, “Meditative Analgesia,” op. cit., p. 58. 124  Ibid., p. 60. 122

204

E. Thompson

Taken together, these results suggest that the long-term meditators have learned to decouple the alert monitoring of sensory events from the processes that lead these events to be anticipated and experienced as unpleasant.125 In other words, they have learned to unbias their attention to salient sensory events from habitual affective and motivational influences and approach-versus-avoid action tendencies.126 More generally, the results suggest that the long-term meditators have learned to cultivate a kind of vigilance that consists not in sustained concentration on a particular object, as happens in the case of “focused attention” meditation, but rather in what the Tibetan Buddhist Mahāmudrā meditation tradition, which is strongly influenced by Yogācāra, calls “mindfulness consisting in mere non-distraction.”127 In this style of practice, the meditator cultivates an awareness that disengages from the dualistic appearance of grasper and grasped and the approach-versus-avoid framework of obtaining the desirable and avoiding the undesirable. As we have seen, this dualistic framework is precisely the context required for concept-formation and conceptualization to occur, according to Dharmakīrti. Viewed from a cognitive science perspective, vigilance as present-moment centred “mere non-distraction” can be expected to attenuate autobiographical memory-dependent processes (Dharmakīrti’s imprints/vāsanās),128 the formation of mental narratives, self-related evaluations, and anticipatory self-projections, all of which involve conceptualization and are strongly concept-dependent. Of course, given that the meditators are evaluating and reporting their perceptions of sensory intensity and their experiences of the degree of unpleasantness, conceptualization must be present. Furthermore, it is not possible to know from these experiments the extent to which the long-term practitioners versus the control subjects categorize or conceptualize the stimuli as “pain,” and whether decreased conceptualization of the stimuli as “pain” predicts decreased unpleasantness, though this seems likely, given that the long-term practitioners are apparently minimizing appraisals of the stimuli. Another possibility is that the long-term meditators are engaging in an alternative conceptualization, one that conceptualizes the experience of the stimulus simply as a transitory mental event, rather than conceptualizing it by selectively attending to its content and construing that content as representational (as presenting an accurate depiction of reality). This alternative way of conceptualizing experience has been called “dereification,” and is a typical part of mindfulness meditation

 Ibid., p. 59.  See Jake Davis and Evan Thompson, “Developing Attention and Decreasing Affective Bias,” in Kirk Warren Brown, J. David Creswell, and Richard M. Ryan, eds., Handbook of Mindfulness: Research, Theory, and Practice (New York: Guilford Press, 2015), pp. 42–61. 127  See John D.  Dunne, “Buddhist Styles of Mindfulness: A Heuristic Approach,” in Brian D.  Ostafin, Michael D.  Robinson, and Brian P.  Meier, Handbook of Mindfulness and Self-­ Regulation (New York: Springer, 2015), pp. 251–270. 128  See Masahiro Fujino, Yoshiyuki Ueda, Hiroaki Mizuhara, Jun Saiki, and Michio Nomura, “Open Monitoring Meditation Reduces the Involvement of Brain Regions Related to Memory Function,” Nature Scientific Reports 9 (2018): 9968. https://doi.org/10.1038/s41598-018-28274-4. 125 126

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

205

practices.129 Dereification is the construal of thoughts, feelings, and perceptions as mental processes rather than as accurate representations of reality. Deification has been shown to reduce the effects of motivational states and traits on appetitive behaviours in approach-versus-avoid contexts, not only during formal practice sessions but also over subsequent time periods, in both the laboratory and in the field, and has been shown to have distinct neural correlates.130 Although dereification presumably is an effortful conceptual procedure in novice practitioners, it may eventually become effortless and nonconceptual (in the sense of not requiring active conceptualization) in experienced practitioners. Indeed, according to the bottom-up apoha theory, if dereification has become so thorough that the experienced meditation practitioner directly experiences the stimulus simply as a unique particular without any selective exclusion from that which is other, then the experience is nonconceptual. Investigating these issues requires that cognitive scientists, clinical scientists, philosophers, Buddhist scholars, and experienced meditation practitioners work together. In particular, more attention needs to be given to the cross-cultural philosophical issues about concepts discussed in this paper to clarify and advance the empirical investigation of mindfulness meditation practices. Although the issues discussed in this paper are primarily philosophical and exegetical, they have significant bearing on current research in the neuroscience of meditation. Without attention to these issues, the neuroscience of meditation runs the risk of using the terms “conceptual” and “nonconceptual” in vague and uncritical ways. To return to the main point, the empirical evidence just discussed is consistent with the hypothesis I proposed above. Open monitoring meditation appears to attenuate affect-biased attentional selection and attentional inhibition, affect-biased appraisal and anticipation, and approach-versus-avoid action tendencies.131 In Yogācāra terms, it weakens the grasper-grasped duality. Of course, we cannot know simply from the Yogācāra texts themselves whether their authors practiced anything like open monitoring meditation as described here.  Lutz et al., “Investigating the Phenomenological Matrix of Mindfulness-Related Practices from a Neurocognitive Perspective,” op. cit. 130  See Esther K. Papies, Lawrence W. Barsalou, and Ruud Custers, “Mindful Attention Prevents Mindless Impulses,” Social Psychological and Personality Science 3 (2012): 291–299; Esther K.  Papies, Tila M.  Ponk, Mike Keesman, and Lawrence W.  Barsalou, “The Benefits of Simply Observing: Mindful Attention Modulates the Link Between Motivation and Behavior,” Journal of Personality and Social Psychology 108 (2015): 148–170; and Lauren A. M. Abis, Esther K. Papies, Kaundinya Gopinath, Romeo Cabanban, Karen S.  Quigley, Venkatagiri Krishnamurthy, Lisa Feldman Barrett, and Lawrence W.  Barsalou, “A Shift in Perspective: Decentering Through Mindful Attention to Imagined Stressful Events,” Neuropsychologia 75 (2015): 505–524; Constanza Baquedano, Rodrigo Vargara, Vladimir Lopez, Catalina Fabar, Diego Cosmelli and Antoine Lutz, “Compared to Self-Immersion, Mindful Attention Reduces Salivation and Automatic Food Bias,” Nature Scientific Reports 7 (2017) Article number: 13839. https://doi.org/10.1038/ s41598-017-13662-z. 131  See also E. Fucci, O. Abdoun, A. Caclin, A. Francis, J. D. Dunne, M. Ricard, and R. J. Davidson, “Differential Effects of Non-dual and Focused Attention Meditations on the Formation of Automatic Perceptual Habits in Expert Practitioners,” Neuropsychologia 119 (2018): 92–100. 129

206

E. Thompson

On the one hand, it would be anachronistic at best and grossly inaccurate at worst to suppose that they did engage in open monitoring meditation as described here, because this practice has been strongly shaped by “Buddhist modernism.”132 On the other hand, from a cognitive science viewpoint informed by Buddhist philosophy and concerned with basic principles of human cognition, it seems eminently plausible to suppose that the Yogācāra philosophers would have been familiar with meditation methods for weakening (and, in their view, ideally freeing oneself from) the grasper-grasped duality of what we would call affectively and motivationally biased cognition. We can now return to the conceptualist-versus-nonconceptualist debate. None of the cases of nonconceptual experience that the nonconceptualist adduces is equivalent to the case of alert and acute awareness in which affectively biased perceptual categorization and approach-versus-avoid tendencies are either absent or attenuated, and hence conceptualization (exclusion from that which is other) is either absent or attenuated. In other words, none of the cases is equivalent to experiences in which the grasper-grasped duality is either absent or attenuated. Thus, the Yogācāra viewpoint provides an important and new (to the Western perspective) nonconceptualist framework. At the same time, the cognitive science viewpoint on attuned skillful action may work as a model for the Yogācāra conception of the “subsequent awareness.”133 Some characteristics of highly skilled action are unbiased and effortless attention, present-moment sensitivity, and spontaneous receptivity without anticipatory approach-versus-avoid tendencies.134 Jay Garfield thinks that this idea of action in the world without the grasper-grasped duality is more strongly developed in East Asian Buddhism than in Indian and Tibetan Buddhism, and he describes its full development and understanding as one of the reasons that “Bodhidharma had to go to the East.”135 It should be noted, however, that the idea of efficacious action without the grasper-grasped duality is one that the conceptualist may be able to acknowledge. Indeed, one way to think of the “subsequent awareness” is that it allows for a conceptual embrace or acknowledgement of nondualistic awareness. After all, one mark of great mastery is being able to conceptualize an experiential, activity-based practice or skill so as to teach others effectively.  See Robert Sharf, “Buddhist Modernism and the Rhetoric of Meditative Experience,” Numen 42 (1995): 228–283; Robert Sharf, “Is Mindfulness Buddhist? (And Why It Matters),” Transcultural Psychiatry 52 (2015): 470–484; and David McMahan, The Making of Buddhist Modernism (New York: Oxford University Press, 2008). 133  See Tzohar, “Enjoy the Silence,” op. cit. 134  See Edward Slingerland, Effortless Action: Wu-Wei as Conceptual Metaphor and Spiritual Ideal in Early China (New York: Oxford University Press, 2003); Edward Slingerland, Trying Not to Try: The Art and Science of Spontaneity (New York: Crown, 2014); and Brian Bruya, ed., Effortless Attention: A New Perspective in the Cognitive Science of Attention and Action (Cambridge, MA: MIT Press, 2010). 135  Jay Garfield, “Why Did Bodhidharma Go to the East? Buddhism’s Struggle with the Mind in the World,” Sophia 45 (2006): 61–80. 132

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

207

One conceptualist position congenial to these ideas comes from Alva Noë.136 Following Wittgenstein, his view is that concepts are not mental particulars but rather “techniques by which we secure our contact with the world,”137 that understanding consists of the mastery of such techniques, and that there are many modes of understanding, of which judgement is only one. We should not make judgement be the model for all understanding. There are forms of understanding, such as the practical understanding displayed in skillful action, in which we are not able, as a general rule, to formulate judgements adequate to capture what we do. Nevertheless, Noë argues, such modes of understanding are conceptual, because they require the use of concepts, understood not as mental representations of categories, but rather as techniques for having access to the world. My reservation about this viewpoint is not that it recasts concepts in Wittgensteinian terms as techniques of access to the world but rather that it leaves no place for nonconceptual awareness. On the one hand, it may be that we cannot properly speak of understanding in the absence of conceptuality, and hence that nonconceptual awareness is not a form of understanding. One may worry, however, that by rendering every form of attunement to the world a form of understanding, Noë stretches the notions of understanding and conceptuality too far, so that they become vacuous. On the other hand, even if nonconceptual awareness does not count as a form of understanding, it does not follow that it is not a form of experience (where, in Kantian terms, “experience” includes sensuous intuitions that do not need to be brought under concepts in order to present us with particulars138). Moreover, to suppose that nonconceptual awareness may be embraceable by one or another mode of understanding does not entail that it is itself conceptual.

6 An Enactive Approach to Concepts The paradox is that conceptual understanding must embrace precisely that which transcends conceptual understanding: “the grasping mind cannot grasp its ultimate inability to grasp; it can only cultivate its tolerance of that inability.”139 Likewise, the conceptual mind cannot conceive its inability to conceive, but it may be able to cultivate a tolerance of that inability and let itself be open to the nonconceptual. Such tolerance requires changing our view of concepts. Instead of taking concepts to be internal mental representations of an independent outside world, we can take them to be elements of a complex dynamic process in which the mind and the  Alva Noë, “Concept Pluralism, Direct Perception, and the Fragility of Presence,” in Thomas Metzinger and Jennifer Windt, eds., Open MIND: 27(T) Frankfurt am Main: Mind Group. https:// doi.org/10.15502/9783958570597. 137  Ibid., p. 2. 138  See Allais, “Kant, Non-Conceptual Content, and the Representation of Space,” op. cit. 139  Robert Thurman, trans., The Holy Teaching of Vimalakīrti: A Mahāyāna Scripture (University Park and London: Pennsylvania State University Press, 1976), p. 161. 136

208

E. Thompson

world are interdependent and co-emergent poles. Instead of asking how concepts represent the objective world inside the subjective mind, we can ask how mental concepts and object categories co-arise and mutually specify each other. This approach is the enactive approach to concepts, based on the idea from enactive cognitive science that cognition fundamentally is the enactment or bringing forth of relevance through embodied action.140 From an enactive perspective, cognizer and cognized are opposite poles of the same event. Continuous and reciprocal causal interactions between the cognitive being and its environment, over developmental and evolutionary timescales, produce affordances—perception-action saliences—that become recognition-­ expectation prototypes around which categories form. Eleanor Rosch explains the enactive perspective as follows: Since the subjective and objective aspects of concepts and categories arise together as different poles of the same act of cognition and are part of the same informational field, they are already joined at their inception. They do not need to be further joined by a representational theory of mind… Concepts and categories do not represent the world in the mind; they are a participating part of the mind-world whole of which the sense of mind (of having a mind that is seeing or thinking) is one pole, and the objects of mind (such as visible objects, sounds, thoughts, emotions, and so on) are the other pole. Concepts—red, chair, afraid, yummy, armadillo, and all the rest—inextricably bind, in many different functioning ways, that sense of being or having a mind to the sense of the objects of mind.141

Using the enactive perspective, we can recast the bottom-up apoha theory.142 Concepts in the mind and categories in the world co-arise through the activity of exclusion, under the influence of past experience, dispositions, motivational tendencies, and our body-given sense of similarities and dissimilarities. Dharmakīrti’s theory can be cast as offering a way to understand out how this co-arising happens. To recapitulate: Given our bodies—and, we can add, our culture—we pay selective attention to a certain perception-action salience x, which prominently stands out for us against the background of what we take to be non-x, such that we arrive at the dichotomy of x and non-x. Should our selective attention be drawn to non-x, we can make further divisions—y and non-y, z and non-z, and so on. Strictly speaking, each member of non-x is different from every other one, but we overlook these differences and take some members to be compatible with each other. We treat them in the same way, thus forming equivalence classes through the operation of exclusion from that which is taken to be other. This kind of exclusion already happens at the most fundamental sensorimotor level. Our body conflates differences and superimposes equivalences on them,

 See Varela et al., The Embodied Mind, op. cit., and Thompson, Mind in Life, op. cit. My thoughts about an enactive approach to concepts have been strongly influenced by Eleanor Rosch, “Reclaiming Concepts,” Journal of Consciousness Studies 6 (1999): 61–77, and Amita Chatterjee, “Funes and Categorization in an Abstraction-Free World,” in Siderits et  al., Apoha, op. cit., pp. 247–257. 141  Rosch, “Reclaiming Concepts,” op. cit., p. 72, italics in original. 142  See especially Chatterjee, “Funes and Categorization in an Abstraction-Free World,” op. cit. 140

What’s in a Concept? Conceptualizing the Nonconceptual in Buddhist Philosophy…

209

thereby generating invariants of sensorimotor response. Crucially, these sensorimotor invariants or equivalence classes are based on similarities of response and not on intrinsic similarities. The members of any given equivalence class C are similar because they evoke the same response; they do not evoke the same response because they are similar in themselves.143 Language requires such a stable sensorimotor world, in particular for our associating words with objects (such as colour terms with hue categories). From a cognitive science perspective, however, we can also allow that language possession may drive concept formation and category recognition, so that the exclusion process may not be only bottom-up, but rather may be both bottom-up and top-down with “looping effects.”144 Amita Chatterjee summarizes this enactive verson of the apoha theory as follows: Naming or the use of words (nāma-yojanā) paves the way for explicit learning and hastens the process of manipulating thought contents. Organisms become more and more skilful in abstraction… But even at this stage, we cannot claim that we have cut up the world at its natural joints or that our putative kinds are real natural kinds, since the external constraints come to us only in the form of affordances at the sensory-motor level. Our linguistic descriptions are somehow grounded in our sensory-motor invariants. Without implicit categorization, explicit categorization or learning cannot take place, yet that is no guarantee for grasping the world as it is. Dharmakīrti and some of his followers therefore rightly recognized the role of significant errors behind successful human practices. Cognitive scientists are not at all bothered by this kind of antirealism, because they do not presume to do ontology at all. If they do ontology, then they restrict their ontic claims to their terms of art.145

From an enactive perspective, however, concepts are not inherently erroneous, as they are for Dharmakīrti. Dharmakīrti’s view of concepts as erroneous rests on what has been called his “scheme-content dualism,” namely, “that we can and should make a clear and radical separation between what our linguistic-conceptual schemes create and impose upon an uninterpreted content and that content itself, which is real, accessible only to perception, outside the conceptual scheme, and hence free of the scheme’s distortions and coloring.”146 Given this viewpoint and Buddhist nominalism, concepts cannot but be inherently distorting, despite whatever pragmatic efficacy they may have in serving our ends. According to the enactive approach, however, concepts are a functional part of situations and have a participatory role in structuring them.147 To borrow Hilary

 For an extended and trenchant discussion of this point, see Mohan Matthen, Seeing, Doing, and Knowing: A Philosophical Theory of Sense Perception (New York: Oxford University Press, 2005). 144  See Gilles Fauconnier and Mark Turner, How We Think: Conceptual Blending and the Mind’s Hidden Complexities (New York: Basic Books, 2003). 145  Ibid., pp. 252–253. 146  Tillemans, “How to Talk About Ineffable Things,” op. cit., p. 51. The “scheme-content dualism” idea comes from Donald Davidson, “On the Very Idea of a Conceptual Scheme,” in Donald Davidson, Inquiries into Truth and Interpretation (Oxford: Clarendon Press, 1984), pp. 183–198. 147  See Rosch, “Reclaiming Concepts,” op. cit. 143

210

E. Thompson

Putnam’s words, “the mind and the world jointly make up the mind and the world.”148 The basic function of concepts is not to identify mind-independent things but to bring forth and stabilize saliences as figures on a ground, relative to the cognitive being and its mode of coupling with the world. Viewed from this perspective, as Chakrabarti and Siderits note, exclusion from that which is other “is not a logical operation we perform on sensory input, but a node in the interactive interplay of organism and environment”149 (or an equilibrium of a dynamical system comprising organism and environment). For the enactive approach, “there is nothing that a concept actually is; our notion of a concept merely picks out one facet of a complex dynamic process.”150 In short, if we use the enactive approach to recast the bottom­up apoha theory, then “the concept of a concept is the reification of a dynamic process in which mental particulars play a role and which can be analyzed post-hoc in terms of rule-governed behavior. This dynamic process is consistent over time for a cognizer, as well as across cognizers, because it leads to action that satisfies the interests of cognizers.”151 Here it no longer makes sense to suppose that concepts are inherently erroneous, because the scheme-content dualism underpinning this assessment has been abandoned in favour of a kind of pragmatism.152 What we call “concepts” are to be evaluated according to their efficacy in the pursuit of our aims and not according to how they represent a transcendent reality. What about the nonconceptual? As Rosch notes, “Concepts only exist against a non-conceptual background. We could not even think to talk about concepts and conceptualization without some contrast of what they are not.”153 From the enactive perspective, the co-arising of concepts (mind) and categories (world) presupposes a nonconceptual background against which they emerge. Nonconceptuality provides the inexhaustible field within which conceptuality functions. The obvious worry is that this contrast is a conceptual one, that it comes about through exclusion: the conceptual is that which is not nonconceptual, and the nonconceptual is that which is not nonnonconceptual. Here we need to remember that the concept of a concept is a reification of a dynamic process. The nonconceptual— nominalized and made into a concept—is also a reification. Nevertheless, if there is nothing that conceptuality actually is other than evolving nodes of a dynamic process, then there is nothing that nonconceptuality actually is other than everything else that makes these nodes possible.

 Hilary Putnam, Reason, Truth, and History (New York: Cambridge University Press, 1981), p. xi.  Chakrabarti and Siderits, “Introduction,” op. cit., p. 44. 150  Ibid. 151  Ibid., p. 46. 152  For a discussion of scheme-content dualism in relation to Putnam’s “internal realism,” see Evan Thompson, “Is Internal Realism a Philosophy of Scheme and Content?”, Metaphilosophy 22 (1991): 212–230. 153  Rosch, “Reclaiming Concepts,” op. cit., p. 75. 148 149

On Necessary Connection in Mental Causation––Nāgārjuna’s Master Argument Against the Sautrāntika-Vasubandhu: A Mādhyamika Response to Mark Siderits Sonam Thakchoe

1 Introduction The two traditional Indian Buddhist philosophers  – the Mādhyamika Nāgārjuna (c.150–250) and the Sautrāntika-Vasubandhu (c. 350–430) – agree that mental causation involves a causal relationship between successive consciousness moments in which the previous moments are causes and the latter moments effects. In this chapter, I investigate the nature of this relation at stake. Is it a type of relationship that requires (1) necessary connection between successive consciousness moments in which there is an internal causal connection between the previous and the latter conscious moments? (2) Alternatively, is the causal relationship between consciousness moments just a matter of constant conjunctions of intrinsic natures (svabhāvas) without any necessary causal relation?1 If the former, the two Buddhist philosophers need to explain which of the two truths – conventional truth or ultimate truth – does  There are two senses in which Buddhist philosophy of mind talks about a moment. The first sense of moment (kṣaṇa) refers to the smallest possible measure of time (dus mtha’ skad cig ma). In the context of mental causation, it is the smallest possible consciousness or thought-moment which the Sautrāntika says is simple, indivisible, ultimate, whereas the Mādhyamika says composite, divisible and conventional. By contrast, the second sense of moment has a temporal and pragmatic sense; it refers to the time required to complete an act of producing knowledge – say the time taken to produce direct perception of an object of knowledge fully.” For instance, the first 15 moments (kṣaṇa) of the 16 moments of transcendent insight constitutes the path of seeing (dṛṅmārga) because the truths, not previously seen (adṛṣṭa), are directly seen (dṛṣṭa). The subsequent knowledge of the truth of path constitutes the path of meditation because it is the conditioned result of the previous training (Jampaiyang 2019, 11). 1

S. Thakchoe (*) University of Tasmania, Department of Philosophy, Hobart, TAS, Australia e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_10

211

212

S. Thakchoe

this necessary relation belong and why? If the latter, they need to explain how mental causation  – a continuum of interrupted mental events in which the previous moment are causes and the subsequent moments effects – possible with intrinsic natures, absent any necessary causal connection between them? Either way, the Buddhist philosophers must address these two questions concerning causal relation adequately. The philosopher Mark Siderits defends the second position and attributes this view to Vasubandhu and Nāgārjuna. In his article Causation, ‘Humean’ Causation and Emptiness Mark Siderits (2014) explicitly rejects necessary connection in the Sautrāntika-Vasubandhu’s philosophy, claiming he is “Humean” about causation. The Sautrāntika-Vasubandhu is “Humean” about causation, according to Siderits, because akin to the Scottish philosopher David Hume for whom causation is just a matter constant conjunction of independent mental tropes without necessary connection, Vasubandhu holds that mental causation consists of constant conjunction of dharmas with intrinsic natures (svabhāvas) which can enter into causal relations, absent any internal necessary connection. Siderits puts the claim this way: Ultimately, causation is just a matter of constant conjunction: it always happens that events of this type are succeeded by events of that type. And to say that this is what causation ultimately is to say that it holds between dharmas with intrinsic natures. (Siderits 2014, 434–35)

According to Siderits, Vasubandhu holds the view that dharmas are ultimately real momentary entities with intrinsic nature, and there is no production based on direct causal relation. Causal relation which involves production requires appealing to a necessary connection between dharmas with intrinsic properties “there cannot be internal relation between any two dharmas” (Siderits 2014, 442). Even a third necessary connector could never succeed in fostering a connection between momentarily occurring tropes with intrinsic natures because they are causally independent. Whereas causal relation always presupposes that causes precede effects and that causes and effects are mutually dependent and momentary. Consequently, according to Siderits “the Abhidharmika presentist cannot claim that there are causal relations among dharmas when these are understood in the trope-theoretic fashion” (Siderits 2014, 442). This is because, “there is ultimately no such thing as alteration … There thus being no alteration, a dharma’s being caused to arise bearing its intrinsic nature cannot count as an alteration, so [Nāgārjuna’s Mādhyamika] argument against alteration has no purchase” (Siderits 2014, 14). The Mādhyamika’s argument is effective only against the ultimate causal realism of the Vaibhāṣika towards which it is directed, since it holds the view that the knowledge of ultimately true causal relation confers confidence in understanding why specific causal regularities should hold. But this critique will not work against Vasubandhu’s Humean causation according to which causation is no more than the occurrences of intrinsic natures as moments in which “a dharma’s being caused to arise bearing its intrinsic nature cannot count as an alteration” (Siderits 2014, 14). The implications, therefore, is that the Mādhyamika’s reasoning – that only those moments that are empty of intrinsic nature can dependently originate and those entities having intrinsic

On Necessary Connection in Mental Causation––Nāgārjuna’s Master Argument…

213

natures could not originate in dependence on other things  – will have no force against the causation view the Sautrāntika advocates. Moreover, Siderits claims that Nāgārjuna does not advance any positive account of mental causation because, like the Pyrrhonian sceptical method, he only uses reductio arguments against realist causal theories other philosophers have advanced, and he himself does not contribute any positive account of mental causation (Siderits 2014, 448). But Siderits is circumspective about this later claim: I am thus not certain that a successful Madhyamaka argument against the ‘Humean’ option is to be found in the Madhyamaka literature. Now I also believe there to be no such thing as a Madhyamaka ‘master’ argument, one that somehow proves that all things are necessarily empty of intrinsic nature. And perhaps no Mādhyamika author ever gave due consideration to the ‘Humean’ view and formulated a reductio to refute it. So even if I am not simply overlooking an existing Madhyamaka argument against the view, it remains possible that one might yet be devised. And Madhyamaka has many able defenders today. So I look forward to seeing what others have to say on the matter. (Siderits 2014, 448)

In taking up Siderits’ invitation to challenge his conclusions, I will argue both Nāgārjuna and Vasubandhu are the adherents of the first position. I will argue they both claim for the efficacy of necessary causal connection in mental causation. Indeed, as a Mādhyamika philosopher, Nāgārjuna advocates causal nonessentialism in which the efficacy of necessary connection is a matter of a conventional relationship. All relations must necessarily belong to the conventional category and that nothing could exist ultimately. As a spokesperson of the Sautrāntika, Vasubandhu advocates causal essentialism according to which causal efficacy of necessary connection is an ultimate relation. Only that is ultimately real can be causally effective, and conventional truth is only a fiction. Notwithstanding very significant ontological difference between the two positions, both philosophers agree that necessary connection determines the causal efficacy of mental causation, as it must involves necessary internal relation between the preceding causal consciousness moments and the subsequent effect consciousness moments. This is not surprising, and both philosophers take the Buddha’s definition of dependent origination seriously as it appears in the Śālistamba-sūtra: “Because this exists, this occurs because this arose, that arises, this is called ‘dependent origination’” (Reat 1993; Kamalaśīla 1995, 220). The commentaries on this passage may have offered slightly varying accounts of necessary causal connection, with various examples. Nevertheless they typically follow a similar pattern like this one: dependent on contact, feeling arises; dependent on feeling, craving arises; dependent on craving, attachment arises; with the cessation of contact, feeling ceases; with the cessation of feeling, craving ceases; with the cessation of craving, attachment ceases, etc. Necessary connection is the sole justification that warrants Nāgārjuna and Vasubandhu’s confidence in the Buddhist phenomenological and moral causality, karmic theory, soteriological projects (Nāgārjuna 2007, 1:v.25; Vasubandhu 1988, 4:v.34). Take away the efficacy of necessary causal connection; the Buddhist philosophers will have no rational justification to take seriously any of the fundamental Buddhist doctrines.

214

S. Thakchoe

For a start, even the doctrine of four noble truths would make no sense. If the truth of suffering and the truth of origin do not stand in necessary causal relation, the Buddhists will have no valid reason to warrant their confidence in the efficacy of their practice. What warrants the freedom from suffering (the truth of cessation) by walking on the truth of the path is the efficacy of necessary relation that stands between the suffering and the elimination of the origin of suffering. But we turn to the bigger question at stake here: do the two Buddhist philosophers succeed in providing us such proof of the efficacy of necessary connection in the context of mental causation? I will argue the Sautrāntika-Vasubandhu fails in this difficult task, whereas Nāgārjuna succeeds. Failing to provide proof of the efficacy of necessary causal relation between successive intrinsic consciousness moments throws into doubt the very basis of the central hypothesis the Sautrāntika-­ Vasubandhu is committed – that is, mental causation involves successive or series of consciousness events in which the preceding moments causally condition the successive moments. For the Sautrāntika-Vasubandhu to succeed in such attempt, he must provide a proof of the efficacy of necessary connection between two immediately consecutive intrinsic consciousness moments having a direct and internal causal connection wherein the former is a necessary cause of the latter which is a necessary effect arising from the first. Without an account of a necessary causal connection between two immediately consecutive intrinsic consciousness moments, Vasubandhu will have no proof to support the central Buddhist causal claim that there is an internal causal relation running across successive consciousness moments. And without this proof, he shall have no rational basis to merit any trust in the so-called natural justice principle embedded in karma as a moral theory and phenomenological theory. Nāgārjuna succeeds, I argue, because he has dependent origination argument as his ‘Master’ argument, which is the proof against the efficacy of intrinsic natures and the proof for the efficacy of dependent origination – the emptiness of intrinsic nature. Of the two forms of the origination argument, the deep dependent origination argument shows that intrinsic natures lack the efficacy to form any causal relations amongst consciousness moments, whereas emptiness of intrinsic natures – the dependent origination – affords Nāgārjuna a powerful natural explanation to account for the efficacy of necessary connection that stands between successive consciousness moments. The radical dependent origination argument cuts deeply into the Sautrāntika’s problem of the supposed causal efficacy of intrinsic natures by analysing if any synchronic relation between arising and cessing could be possible in each consciousness moment as intrinsic natures. It turns out intrinsic consciousness moments cannot form any synchronic relation between its arising and cessing, thus offering us no reasoned explanation that warrants our belief in the causal efficacy of intrinsic consciousness moments. When the same argument is deployed to analyse the Mādhyamika’s claim for the causal efficacy of consciousness moments as the emptiness of intrinsic nature by examining how synchronic relation between arising and cessing play out in each consciousness moment, the argument successfully explains the efficacy of synchronic relation between arising and cessing of consciousness

On Necessary Connection in Mental Causation––Nāgārjuna’s Master Argument…

215

moment, thus offering us reasoned explanation that warrants our confidence in the causal efficacy of dependent arising.

2 The Deep Dependent Origination Argument In his magnum opus, Fundamental Verse of Middle Way (Mūlamadhyamkakārikā) Nāgārjuna deploys his famous catuḥṣkoṭi – the four-cornered logic – against causal essentialism. (Nāgārjuna 2013) This argument we can reconstruct, in the context of mental causation, as follows: P1: Consciousness moments either have (P1.1) ultimate causes or (P1.2) conventional causes. P2: If consciousness moments have ultimate causes (P1.1), it is certain that there are three alternatives in which such ultimate causation could get going: the cause and its effect in mental causation will have either (P2.1) the same essence – intrinsic natures (svabhāva), or (P2.2) the cause and its effect will have distinct essences – extrinsic natures, or (P2.3) the cause and effect will have both intrinsic nature and extrinsic natures. P3.1: Suppose consciousness moments arise from the ultimate cause and the effect having intrinsic natures, requiring “no alteration.” Siderits attributes this view to Vasubandhu, stating: “on Vasubandhu’s presentist understanding of momentariness, there is ultimately no such thing as an alteration. For instance, what we call the motion of light is reductively analyzed into a causal series consisting of the occurrence of flame f1 at p1 at t1, the occurrence of f2 at p2 at t2, the occurrence of f3 at p3 at t3, etc. There thus being no alteration, dharma’s being caused to arise bearing its intrinsic nature cannot count as an alteration, so the argument against alteration has no purchase.” (Siderits 2014, 14). As we observe, Siderits makes two conflicting claims about intrinsic natures: (1) they are causally closed, with no alteration. This is the reason, he claims, the Mādhyamika’s critique against the alteration does not apply to the Sautrāntika. Although it does apply to the Vaibhāṣika’s view in which efficacy requires alternation of intrinsic nature’s temporally enduring character, what he calls “temporal thickness”. The Sautrāntika’s intrinsic natures is immune to such criticism because it understands intrinsic natures simply as occurrence of consciousness moments (Siderits 2014, 14) But then Siderits goes on to attributing (2) intrinsic nature a causal efficacy to causing other intrinsic natures to arise, analogous to the motion of a light forming causal series consisting of the occurrence of flame. Suppose the former is correct, the Mādhyamika critique applies. We would have to accept the efficacy of intrinsic nature as the given, existing intrinsically, independently of causes and condition (thus self-caused, akin to the causeless Īśvaraḥ) (P2.1). But this cannot be true. Causally closed intrinsic nature would have no efficacy to give rise to other intrinsic natures, much like an uncreated/uncaused Īśvaraḥ which cannot give rise to the universe. The occurrence of consciousness c0 at p0 at t0 would be completely non-causal. It follows, then, that the occurrence of consciousness c2 at p2 at t2 is temporally unrelated in any way, neither with the

216

S. Thakchoe

preceding occurrences of consciousness c1 at p1 at t1 nor with the subsequent occurrence of consciousness c3 at p3 at t3. The problem of causal nihilism. Thus without supposing causal relation between these successive intrinsic moments, Siderits cannot get his mental causation off the ground in the context of intrinsic nature. Even with the supposition that intrinsic consciousness moments (the cause) could give rise to other intrinsic consciousness moments (the effect), the two intrinsic consciousness moments would be identical (even though, two may not be simultaneous) since they do not change, immune to any alternation. If the subsequent intrinsic consciousness moment still requires arising from the previous, that arising would go on uninterruptedly and eternally. Since such arising is intrinsic and independent, there is nothing to stop the regress. The temporal thickness problem. Suppose the latter is correct and that intrinsic natures do have causal efficacy (like Īśvaraḥ with its supposed creative powers), then mental causation could be, as Siderits claim, series of thought moments caused to arise bearing their intrinsic natures. Upon such supposition, we could believe intrinsic natures to having causal efficacy to give rise to other intrinsic natures (much like people thinking Īśvaraḥ as having powers to create all beings and the universe). It follows, then, that the occurrence of consciousness c2 at p2 at t2 is necessary temporally related to both the preceding occurrences of consciousness c1 at p1 at t1 and the subsequent occurrences of consciousness c3 at p3 at t3. Only on the assumption that there were such causal relation operating between the successive intrinsic moments, in which the former is the cause and the latter the effect (hence granting the efficacy of necessary causal connection as the given), could Siderits succeed in driving mental causation in the context of intrinsic nature going. But to admit causal efficacy in necessary connection would require abandoning the claim that efficacy involves intrinsic natures, which is his chief hypothesis. On the Mādhyamika’s reasoning, Siderits’ argument fails to show the efficacy of intrinsic natures of occurrences of flame. On the contrary, the argument demonstrates the occurrence of flames is only dependently originated events – empty of intrinsic natures. The motion of a light form series of the occurrences of flame only because each flame’s occurrence is a result of an uninterrupted necessary relation formed between the arising of the flame and the cessing of the wax, wick, air including other flame supporting conditions. On this view, the motion of a light is causal series with necessary causal connection in which the arising of flame f3 at p3 at t3 f1 at p1 at t1 is dependent on the cessing of f2 at p2 at t2 which in turn is dependent on the cessing of f1 at p1 at t1 and so on (Schoening 1995, 2:63–65). P3.2: Supposed consciousness moments arise from the ultimate cause and the effect having extrinsic natures, such that consciousness moments would arise extrinsically, caused by ultimately distinct ‘other,’ extrinsic natures. (P2.2) The cause consciousness moments and the effect consciousness moments would then be utterly distinct, because it is not possible to have any causal relation between the successive consciousness moments which are ultimately ‘other’ extrinsic natures, analogous the right and the left horns of an ox, since they exist independently, cannot be said to cause one another. The Sautrāntika’s conception of consciousness moments as intrinsic natures faces the same problem as does extrinsic natures, as they are also

On Necessary Connection in Mental Causation––Nāgārjuna’s Master Argument…

217

supposed to be distinct, unique particulars (svālakṣaṇas). As such intrinsic consciousness moments (like extrinsic consciousness moments) utterly lack causal efficacy, when it comes to forming possible causal relation diachronically, for they are causally closed unique particulars without any capacity to relate to successive consciousness moments causally. Suppose extrinsic natures have their causal efficacy in mental causation, in spite of having no capacity to forming necessary connection between the cause and the effect, then, the problem of causal ‘slough’ (in the Cowherds’ language) seems inevitable: now anything could arise from everything else. After all, any consciousness moment would equally qualify the criterion of being ultimately distinct “other”. Constant conjunction of pitch darkness could cause a brightness, although two have no causal connection in any way. Constant conjunction of billions of confused thoughts arranged in a certain order, mixed up with anger, hatred, and aggression should bring about inner peace, calmness, and an end to hostility, a great peace-maker! The problem of temporal thickness – infinite regress – seems inevitable. Suppose an extrinsic consciousness moment (the cause) could give rise to another extrinsic consciousness moment (the effect), the two extrinsic consciousness moments would be ultimately distinct (even though, two may not be simultaneous), and they may not undergo any change, immune to any alternation. If the subsequent extrinsic consciousness moment still requires arising from the previous extrinsic consciousness moment, the same arising would go on uninterruptedly and eternally. Since such arising extrinsic and independent, there is nothing that could stop the regress (Nāgārjuna 1995, 2:63–64). Siderits claims that the Mādhyamikas arguments against causal essentialism is directed only on the Vaibhāṣika intrinsic nature, but not against extrinsic nature. He claims that the Mādhyamika endorses extrinsic natures (agreeing with Bhāvevika), as part of its ontology. But Nāgārjuna’s critiques of essentialism cuts through both intrinsic natures and extrinsic natures, both are the two sides of causal essentialism, expressed differently. P3.3: Consciousness moments are not both intrinsically self-caused and extrinsically caused by distinct ‘other’ (P2.3). Logical absurdities showed for P2.1, and P2.2 apply. P4: There is no fourth alternative relationship in which the ultimate cause could give rise to its effect in the context of mental causation. Therefore, the three alternative ultimate causal relationships are exhaustive. C1: Therefore, it is not possible for consciousness moments to exist as an ultimate, since all possible ways in which consciousness moments can arise ultimately – with intrinsic nature (self-caused), extrinsic nature (caused from another), or both intrinsic and extrinsic natures – are blocked as logical absurdities. It is not possible to establish any necessary connection between the ultimate causes – extrinsic natures and intrinsic natures – and the ultimate effects, and without necessary connection, ultimate causes cannot have efficacy get mental causation going.

Hence Nāgārjuna himself concludes Śālistambakakārikā this way: [consciousness effect] arises not from self (na svato), not from other (parato nāpi), not from both

218

S. Thakchoe

(na dvayoḥ), an agent (kartṛ), nor from time (kālataḥ); similarly, not created by Īśvara and so forth (īśvarādikṛtaṃ), nor from intrinsic reality (naivaṃ svabhāvānnā), nor arises causelessly (apyahetutaḥ) (Nāgārjuna 1995, v.18; Schoening 1995, 2:63–69). C2: It follows, therefore, from the premises P1–P5, there is only one way in which consciousness moments can exist– that is, must exist only conventionally (P1.2). Only whatever exists conventionally can dependently originate from necessary causal connections between specific causal and specific conditional relations which are empty of intrinsic/extrinsic natures. Considers visual consciousness moment as an example: It is true that each visual consciousness moment dependently (āśraya) arise from the appropriation of its conditions such as the visual sense organ (adhipati pratyaya), visible form/color (hetu pratyaya), appearance -i.e., mental image or representation of the form - (ālambana pratyaya) and immediately preceding mental attention directed towards the specific object which has just been ceased to exist (samanantara pratyaya) (Nāgārjuna 1995, vs.52–54; Schoening 1995, 2:76). Absent any one of these conditions, we find it is not possible for any visual consciousness moments to arise, and without one moment, there cannot be causal series. As we can appreciate the Mādhyamika’s deep dependent origination argument is not entirely reductive, does not stop at a categorical rejection of the Sautrāntika’s assertion that consciousness moments are ultimately real – independent, intrinsic natures or extrinsic natures – and yet causally effective. The reductive method is only a means to arrive to the positive conclusion that, upon analytic inquiry of mental causation, we cannot find any intrinsic/extrinsic consciousness moments to support the efficacy of ultimate causation. But where Siderits errs is his supposition that the Mādhyamika’s dependent origination argument is purely reductive and that its rejection of ‘ultimate’ causal connection should also entail the rejection of “conventional” or “nominal” necessary connection. But this entailment clearly does not hold. For a start, as we can appreciate the first premise (P1) of the Mādhymika’s reductio argument has two sub-premises (P1.1, and P1.2). Considering the doctrine of the two truths as the ontological framework within which the argument operates, the reductive conclusion holds only for P1.1: the realist assertation that consciousness moments exist ultimately, but the argument is not a reductio against P1.2: the Mādhyamika’s own view that consciousness moments exist only conventionally. And this is precisely the constructive philosophical view Nāgārjuna defends with the Mādhyamika’s ‘Master’ argument.

3 The Radical Dependent Origination Argument It stands to reason that some major philosophical works of Nāgārjuna have received very little or no attention from Siderits. Those works include Verses on Rice Seedling / Śālistambaka-kārikā, and Detailed Commentary on the Verses on Rice Seedling /

On Necessary Connection in Mental Causation––Nāgārjuna’s Master Argument…

219

Śālistambaka-vistarākhyāṭīkā (Nāgārjuna 1995), traditionally considered to be commentaries on the famous Śālistambasūtra (Reat 1993). The Heart of Dependent Origination (Pratitīyasamutpādahṛdaya) and Commentary on the Heart of Dependent Origination (Pratītyasamutpādahṛdayavyākhyāna) (Nāgārjuna 2001), Precious Garland (Ratnāvalī), Nāgārjuna’s independent works. In these important philosophical treatises, we observe Nāgārjuna’s constructive philosophical project coming out forcedly, through his moral phenomenology, philosophy of mind, soteriology, and mental causation. We see him arguing persuasively for mental causation and the efficacy of necessary connection in the context of dependent origination. In these works, supplying three premises, Nāgārjuna defends the Mādhyamika’s theory of causal efficacy of the emptiness of intrinsic natures (alias dependent origination) as the conventional truths and critiques the Sautrāntika theory of causal efficacy of intrinsic/extrinsic natures as the ultimate. The radical dependent origination argument is a proof of the causal efficacy of cessation and its capacity to forming necessary causal connections at the subtlest and radical ontological level, and thereby demonstrating causal inefficacy of intrinsic/extrinsic natures at the deep level. In the Śālistambasūtra-tīkā (Commentary on the Sūtra of Seedling) in particular, Nāgārjuna fully develops a constructive account of dependent origination in which he describes dependent origination as constituting the reality (dharmatā), true nature of mental processes, the non-deceptive truth, and the actuality (bhūtatā). The knowledge of dependent origination, he says, constitutes the truth (satyatā), the objective reality (tathatā), the knowledge which is freed of distortions (aviparītatā), one which accords with the reality and is incontrovertible (aviparyayatā). He describels dependent origination as the enduring law (dharmasthitato), the infallible law (avipraṇāśadharmatā), the natural law (dharmaniyama) that governs the regularities of natural causal processes and causal events based on definitive successions of conditionality (idaṁpratyayatākramaniyāmatā) in accordance with the sequence of causes and conditions (Nāgārjuna 1995, v.12; Kamalaśīla 1995, 243–264). When appropriate causes and conditions come together, they mutually activate each other, their complementarity or necessary relation results in the efficacy in mental causation. The efficacy of necessary relations holds true between all successive stages of mental causation, because without necessary causes (hetvabhāve) conditions could not be activated (āpravṛttiḥ), and neither could the causes be activated without the presence of necessary conditions (Nāgārjuna 1995, v.12). Likewise, in the Sautrāntika’s mental causation, one which Vasubandhu develops in chapter IX of Abhidharmakośabhāṣya (Vasubandhu 1988), each moment of consciousness stands in direct causal relation with others temporally bearing a direct causal relation to the next moment of consciousness, and the next to the subsequent ones. Accordingly, mental causation involves causal continua or series (saṁtāna) of the material and mental aggregates (skandhas) succeeding uninterruptedly in a row. The occurrences of successive consciousness moments are however differentiated on account of the evolution (pariṇāma), or transformation of the series. Of the

220

S. Thakchoe

consecutive moments “the last consciousness moment possesses a special efficacy, the capacity of immediately producing the result: it is distinguished, in this regard, from other thought moments; it is then termed viśeṣa, or the ultimate moment of evolution.” (Vasubandhu. 1992, 4:v.34 ;1991: 1353) When a consciousness moment at the time of death, for example, is associated with desires, it possesses the efficacy of giving rise to a new life. This consciousness has for its antecedent many mental actions of all types of ethical qualities, nevertheless it is the efficacy projected by a weighty action of the last thought moment which exerts a maximum causal influence. So, a previous consciousness moment is a cause of the latter consciousness moment; the consciousness is then a cause of consciousness. (Vasubandhu 1988, 1343) There is little doubt that the efficacy of necessary connection is one which the Sautrāntika-Vasubandhu takes seriously, as he tries to develop here, notwithstanding to the contrary position Siderits attributes. The most radical of the Mādhyamika’s dependent origination arguments is the one that proves causal interdependence of the subtlest consciousness moments. The Mādhyamika and the Sautrāntika agree that consciousness is always in a flux of motion, invariably changing, in the process of becoming from the moment it comes into existence. But they disagree profoundly about what this change amounts to. For the Mādhyamika, even the subtlest consciousness moment is only conventionally real, has no ‘existence’ or no being, apart from constant becoming, flux, change: momentariness, emptiness. There is not a single moment in which consciousness could remain unchanged, and fully becomes what it is, or acquires the static being. Chang is its reality. Whereas for the Sautrāntika consciousness moment does have a momentary existence, or momentary being which it describes as: ultimate, simple, inexpressible, intrinsic nature, and unalterable. The Mādhyamika’s radical dependent origination argument shows that each consciousness moment, considered as an individual event, is necessarily an interdependent (entirely lacking intrinsic nature and independence), is causally conditioned, both by its arising from appropriate causes as well as its ceasing. As a causally dependent event, consciousness moments could only occur interdependently upon the ceasing of the preceding conditions, in conjunction with the cessation of other necessary conditions. Just as a seed dependently arises from its necessary causes and conditions, so does consciousness moment dependently arise from its respective causes and conditions. Analogously, just as the ceasing of the barley seed allows the barley sprout to emerge from it, only upon the ceasing of the previous consciousness moment (in virtue of the cessation of the conditions supporting its existence in conjunction with other conditions), does the subsequent consciousness emerges out of it. In this way, mental causation necessarily requires the efficacy of both arising and ceasing, and the two are mutually empowering activities, together constituting a single moment of consciousness. Arising and ceasing to constitutes the character of each consciousness moment, analogous to ascending and descending of the two ends (A and B) of a scale, mutually causing each other. The A’s descend causes the B to ascend, and vice versa, the A’s ascends causes the B to descend. Similarly, arising causes the ceasing of consciousness, and the ceasing likewise causes the arising

On Necessary Connection in Mental Causation––Nāgārjuna’s Master Argument…

221

of consciousness. In this way, both arising and ceasing synchronically and diachronically constitutes the becoming (not the ‘being’) of a consciousness moment in which two are equally causally significant: conditioned and causally efficient. Just in the same way, the arising of a flame and the cessation/burning of a wick are simultaneous in a motion of light. The burning of the wick causes the flame, and the flame causes the burning of the wick, the two must go together to generate the motion of light. The Sautrāntika-Vasubandhu, by contrast, attributes causal efficacy only to the arising of consciousness and denies causal efficacy to the ceasing consciousness. Accordingly, Vasubandhu claims whatever is a conditioned must disintegrate: spontaneously and causelessly. Consciousness disintegrates spontaneously because a conditioned consciousness does not exist beyond the acquisition of it’s: it perishes on the spot where it arises; it cannot go from this spot to another. Consciousness ceases causelessly, because it’s ceasing is intrinsic, not caused by anything external; it does not depend on an extrinsic cause. Moreover, the ceasing of consciousness is not causally productive, is a simple absence, negation, not an effect. Only an effect, something ‘done’ or ‘created’ can be causally effective. (Vasubandhu. 1992, 4:v.2) Thus, the ceasing and arising of each consciousness moment in mental causation are mutually excluding, not synchronic and not diachronic, on the view of the Sautrāntika-Vasubandhu. The Mādhyamika postulates three key premises to support the logical and phenomenal expressions of the radical dependent origination in mental causation (Nāgārjuna 1995, v:41; Kamalaśīla 1995, 285–286, 324–326; Schoening 1995, 2:74). The first is the premise from the principle of non-interruption or non-­ annihilation (nocchedataḥ / ched pa ma yin pa) which is that the arising of the latter consciousness moment necessarily entails the ceasing of the previous consciousness moment and the ceasing of the previous consciousness moment necessarily entails the arising of the latter consciousness moment. The arising and ceasing causally condition each other, and, thus, diachronically sustaining the successive chain of causes and effects uninterruptedly. A- effect consciousness becomes a cause, and the cause-consciousness becomes an effect. The arising consciousness is causally efficacious, as it causes it’s ceasing as one of the effects. The ceasing consciousness is causally efficacious, as it causes it’s arising in the next moment. The arising consciousness does not occur from the prior complete cessation of its causes. Therefore causes are not annihilated. Nor without ceasing, therefore, causes are not eternal, just as the ascending end of the scale does not ascend prior to, or without the descending of the descending end of the scale. But when the former consciousness is ceasing in virtue of the ceasing of its causes and conditions, just at that time, concurrently, as for the effect of that ceasing, which is comprised of the arising of consciousness in the next moment, occurs. The Sautrāntika-Vasubandhu’s account too, in theory, complies with the principle of non-interruption, or non-annihilation (nocchedataḥ / ched pa ma yin pa). In practice, however, its commitment to the svabhāva ontology directly undermines the principle. A causal interruption is inevitable when each intrinsic consciousness moment ceases and go out of existence instantly as soon as it comes into existence.

222

S. Thakchoe

Moreover, this ceasing of intrinsic consciousness, the Sautrāntika claims, is uncaused  – not causally dependent, does not arise from any cause. Nor does the ceasing consciousness produce any effect, thus, deprived of any causal efficacy. Consequently, there is no direct diachronic relation between the ceasing of the previous intrinsic consciousness and the arising of latter intrinsic consciousness. The ceasing consciousness does not stand in any causal relation  – neither as an effect nor as a cause – exerting no causal influence whatsoever over the arising of the subsequent consciousness moments. The inevitable interruption leads to causal nihilism, as it undermines causal efficacy, causes rupture and discontinuity in mental causation (Vasubandhu 1988, 4:v.2b). Here causal nihilism is one of the unavoidable consequences of the Sautrāntika’s view according to which mental causation is supposed to be series of occurrences of intrinsic consciousness moments in which only the arising matters in its causal process and the ceasing has no role, whatsoever, to play. We can articulate the problem of causal nihilism by turning to Siderits’ analogy of motion of light against the view he defends. It is not possible for the motion of light to continue as a causal series consisting of merely as occurrences of flame (contrary to Siderits claim) because each occurrence of flame is interrupted by a cessation of the occurrence of flame (due to the cessation of the causes supporting the flame) and this cessation, the Sautrāntika claims, is always causally inert – it is not produced by any cause, and it cannot produce any effect. Thus, the cessation of the previous flame occurrence f1 at p1 at t1 along with the cessation of the wax, the wick and the air which its consumed would render impossible the occurrence of flame f2 at p2 at t2 to succeed the previous flame occurrence, for the cessation of the previous flame occurrence f1 at p1 at t1 would leave an unbridgeable causal gap which interrupts the causal continuum for a motion of light to proceed. Likewise, all the subsequent occurrences of flame such as f3 at p3 at t3 would not be possible because of the causal gaps due to the cessations of the previous flame occurrences. Therefore, the motion of light is impossible simply from the occurrences of flame. Otherwise, we will have to accept the occurrence of flame without depending on the causal efficacy of the cessation. It, therefore, follows that there will be no motion of light without the efficacy of cessation considered. Without the cessation of the previous occurrence, no subsequent flame could occur, and without it, there could not be causal continuum or series, thus the inevitable temporal nihilism. The second premise of radical dependent origination is the principal causal resemblance relation. (Nāgārjuna 1995, v.18) According to this premise habits forming mental capacities, dispositions, seeds, desires, prejudices etc., associated with the previous consciousness do not get entirely terminated or interrupted in the process of causal change, because the successive moments arise resembling the previous ones in virtue of being causally conditioned by them, and that there is a clear internal causal connection continued through the successive consciousness moments. Each consciousness moment comes into existence by its own necessary causes and conditions, some of which are antecedent and the others present. Both the antecedent causes – the ceasing of the precious consciousness – and the present

On Necessary Connection in Mental Causation––Nāgārjuna’s Master Argument…

223

conditions collectively give rise to the future consciousness, this way the subsequent ones resemble the former. Consider the visual perception of blue (subject). Its production has both antecedent cause and the present necessary conditions (thesis) because it cannot arise merely from the presence of the visual faculty and a blue object (reason). Whenever a visual perception of blue arises, it must arise resembling some aspect of the three necessary conditions – visual sense faculty (present, dominant condition), a previous moment of thought processes – desires, biases, etc., (preceding condition) and a blue object (present, objective condition). It is not possible for the visual perception of a blue to arise, if anyone of the necessary conditions is not satisfied, therefore there must be all three conditions including the previous moment of mental consciousness which disposes perception its ability to carry out perceptual judgement based on the past prejudices, experiences, memories, desires etc. On this view, consciousness owes its very character to its causes and conditions, as it can only dependently originate resembling the preceding conditions. It cannot exist any other way: neither intrinsically nor extrinsically, nor as a unique particular (svālakṣaṇa) – because it is not possible for any subsequent consciousness to exist as intrinsic/extrinsic natures while at the same time being produced resembling the previous consciousness moments. For this reason, the Mādhyamika argues, all consciousness moments in any aspects of causal chain (all the way up to the highest-­ order cognitive processes and all the way down to the subtlest atomic consciousness moments) must be empty of intrinsic natures (niḥsvābhavaḥ) for them to follow the regularities of dependent origination and the principle of resemblance. It is the emptiness of intrinsic/extrinsic natures that renders possible consciousness moments to dependently originate because it is an emptiness which makes possible forging necessary connections. The third premise is the principle of non-transference (na saṁkrāntitaḥ), which is a reason against any form of ontological transference, despite the importance of causal resemblance. Nāgārjuna insists that there is no transference whatsoever in mental causation (Nāgārjuna 1995, v.20). There is a resemblance between each consciousness moment as far as the latter consciousness moments dependently arises from the ceasing of the previous consciousness. No aspect whatsoever of the previous consciousness transmigrates or transfers to the subsequent consciousness as its effect. There is only the appearance of an effect-consciousness resembling the cause-consciousness. This resemblance occurs in virtue of non-deficiency of similar causes and conditions sustaining the previous and the latter consciousnesses, but they are not identical. It is like the reflection of a face seen in a well-polished mirror. The face does not transfer into the mirror, nor does the mirror take up the face’s space. And yet there is the appearance of the face. Of course, not just any face, indeed, the reflection only of the face resembling the face of the person standing in front of the mirror, because the necessary conditions, amongst which the face of that person is one, required for that specific reflection on taking place in the mirror, are satisfied. Again, it is clear how causal resemblance operates out in the Sautrāntika’s account without undermining the principle of non-transference. It has two

224

S. Thakchoe

contradictory commitments: all consciousness moments are intrinsic natures (unique particulars), but the subsequent consciousness moments resemble the preceding moments. It could argue the earlier intrinsic consciousness causally condition the arising of the latter intrinsic consciousness, and that the latter intrinsic consciousness become what they are ontologically due to the power of the preceding intrinsic consciousness. In this way, the latter intrinsic consciousness could resemble the preceding ones. But in the context of consciousness moments as intrinsic natures (also unique particulars), it is contradictory to posits resemblance relation between two intrinsic consciousnesses wherein the former causes the latter, in spite of the fact that, the natures of both the former and the latter consciousness moments are supposed to be intrinsic, unique, ultimate and unalterable. If we suppose there is a causal resemblance between the successive moments of intrinsic consciousness, then the former intrinsic moments would necessarily wholly transfer onto the latter moments, undermining the principle of non-transference, thus the problem of eternalism. Moreover, the transference of the earlier intrinsic moments to the latter moments would imply that the latter moments are not intrinsic natures, for the intrinsic natures of the earlier moments would completely replace the latter moments, annihilating the latter intrinsic moments, thus the problem of nihilism. Suppose mental causation requires the cessation of the previous moment of intrinsic consciousness, even then, all the previous consciousness moments will have to relinquish their intrinsic natures, because on this view, for consciousness to exist is for it to exist as intrinsic nature, absent intrinsic nature, no consciousness could exist, thus causal nihilism. Suppose mental causation does not require the cessation of the previous moment of intrinsic natures, in that case, the previous moments of intrinsic consciousness would exist eternally, taking the place of the subsequent moments. Now all moments of consciousness in mental causation will have identical intrinsic natures temporally and spatially since the identical intrinsic natures would eternally pass on from the earlier to the later moments. This is causal eternalism. We turn to Siderits’ analogy of motion of light to articulate the problem at stake here. According to the view Siderits defends, it is possible for the motion of light to continue as a causal series consisting of them merely as occurrences of flame f1 at p1 at t1, the occurrence of f2 at p2 at t2, the occurrence of f3 at p3 at t3, etc., without ever depending on the cessation of the previous flame occurrence along with the cessation of the wax, the wick and the air which it’s consumed. The occurrence of flame f2 at p2 at t2, will not depend on the cessation of occurrence of flame f1 at p1 at t1, and likewise, all subsequent occurrences of flame such as f3 at p3 at t3 would not depend on the cessation of the previous flame occurrences. Therefore, the motion of light generated from the occurrences of flame would not be causally dependent. It, therefore, follows that the motion of light could continue as a causal series eternally without ever depending on the causes and conditions, thus the temporal thickness. Both the problems are unavoidable for the Sautrāntika. What matters to mental causation is arising of consciousness moments and has no causal relevance to the cessation. The arising of subsequent consciousness does not depend on its ceasing of the previous conscious moment and nor is the ceasing of the previous

On Necessary Connection in Mental Causation––Nāgārjuna’s Master Argument…

225

consciousness dependent on the arising of the subsequent consciousness moments. In mental causation, the arising of intrinsic consciousness moment successively and continuously is possible without ever depending on the cessation. The arising of consciousness cannot depend on the ceasing of consciousness; if it does, they would have to abandon the claim that the ceasing is causally ineffective. This is one of the consequences of the Sautrāntika’s view according to which mental causation is series of occurrences or arising of consciousness moments in which only the arising matters in its causal process and the ceasing has no role, whatsoever, to play.

4 Conclusion I have argued necessary connection is the heart of Nāgārjuna’s mental causation; it determines the success or the failure of dependent origination arguments in the context of mental causation. I have argued that there is indeed such a thing as a Mādhyamika’s ‘Master’ argument against intrinsic natures which we can draw from Nāgārjuna’s philosophical corpus which thus far received no or very little attention from the philosophical studies of Buddhism. In these texts Nāgārjuna deploys dependent origination argument as the central proof to demonstrate the efficacy of necessary connection between successive consciousness moments which are empty of intrinsic natures, as the very basis of mental causation. I suggest Siderits’ reductive analysis (following the Sautrāntika’s ontological reductionism) fails to prosecute the dependent origination argument to its full conclusion, abruptly quitting analytic inquiry at the level of dharmas, presupposing the intrinsic natures of momentary occurrences of consciousness moments, and prematurely concluding that mental causation involves the arising of intrinsic natures as consciousness moments which are not subject to alteration (Siderits 2014). But the Mādhyamika asks: why should we stop analytic inquiry at the level of momentary occurrences of consciousness moments? Only on the supposition that these momentary occurrences are somehow immune to the analytic inquiry that Siderits could arrive at the conclusion that consciousness moments are intrinsic natures, ultimate, unalterable, independent (following Hume and Vasubandhu). But upon prosecuting analytic inquiry all the way through, subjecting even the subtlest consciousness processes under the glare of same analytic force, Nāgārjuna, however, only finds consciousness moments to be utterly empty of intrinsic natures, finding no basis for the supposition of causal efficacy of intrinsic nature. On the contrary, upon a thorough prosecution of analytic inquiry, discovering that there is nothing that withstands analytic enquiry, Nāgārjuna arrives at the most radical conclusion: even the most subtle consciousness moments are strictly relational process – dependently originated – entirely lacking any intrinsic nature, including the emptiness of the consciousness moment itself is also empty. The Mādhyamika’s analytic enquiry destroys the credibility of Siderits’ and the Sautrāntika’s supposition of the causal efficacy of intrinsic consciousness moments, but it validates the efficacy of necessary connection embedded in regularities of successive

226

S. Thakchoe

consciousness moments which are strictly dependently originated natural processes. Thus, I contend that an incomplete analytic inquiry, not surprisingly, fails Siderits to fully appreciate the full force of the Mādhyamika reducio argument against his supposition that intrinsic natures are logically irreducible and that constant conjunction of intrinsic natures can deliver efficacy in mental causation, absent any necessary causal connection. Deploying the deep dependent origination argument, I have argued intrinsic natures lack both diachronic and synchronic efficacy to form any causal relations amongst consciousness moments because they could not form necessary causal relations crucial for mental causation. I have defended the Mādhyamika’s claim that only consciousness moments which are empty of intrinsic natures –dependent originated – can afford them the diachronic and synchronic efficacy to stand in a necessary causal relation to account for mental causation. I take the analyses a step deeper through the radical dependent origination argument which enables us to closely examine the nature of mental causation at work at the subtlest moments of consciousness, focusing on the possible relationship between the arising and cessing of each consciousness moments. Through the premises of non-interruption, or non-annihilation (nocchedataḥ / ched pa ma yin pa), of resemblance and of non-transference (na saṁkrāntitaḥ) I have defended the Mādhyamika’s radical dependent origination thesis, argued for the efficacy of necessary connection between the arising and cessing of each consciousness, all forms of the subtlest consciousness moments are also synchronically and diachronically causally dependent. Deploying these premises against Siderits’ and the Sautrāntika’s position, I have argued to the effect that it is not possible to forge any synchronic and diachronic causal connection between the arising and cessing of intrinsic consciousness moment, because the two activities are contradictory properties – only arising has efficacy, and cessation is causally inert. The supposition that intrinsic natures are causally effective, even though they cannot forge any causal connection amongst themselves both diachronically and synchronically, I argue, faces unavoidable consequences: temporal eternalism and nihilism. Siderits attempts to deflect the Mādhyamika’s criticism of the Sautrāntika’s conception of intrinsic/extrinsic nature, in my view, also fails. For such an attempt to succeed, I suggest, Siderits would require abandoning the marrying of efficacy/ causation with intrinsic/extrinsic natures altogether. As the consequent of bringing the two together which he must do to get going his Humean version of the Sautrāntika-Vasubandhu’s mental causation, as he does here, Siderits cannot avoid slipping back into the causal slough  – temporal nihilism and temporal thickness traps – which he is keen to avoid. Therefore, I recommend taking the Nāgārjuna dependent origination argument more seriously. I think it is effective against any conception of realism, after all, it’s the Master argument – the ‘King of all arguments’, as the Tibetans would often say.

On Necessary Connection in Mental Causation––Nāgārjuna’s Master Argument…

227

References Jampaiyang, Chim. 2019. Ornament of Abhidharma: A commentary on Vasubandhu’s Abhidharmakosa. Trans. Ian James Coghlan. Somerville: Wisdom Publications. Kamalaśīla. 1995. Commentary on the Śālistamba-sūtra (Śālistamba-ṭīkā). In Commentary on the Śālistamba-sūtra (Śālistamba-ṭīkā). Trans. Jeffrey Davis Schoening. Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. Nāgārjunaa. 1995. The Śālistamba Sūtra and its Indian Commentaries: Tibetan editions. Edited by Jeffrey Davis Schening. Vol. 2, 2 vols. Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. Nāgārjuna. 2001. Nāgārjuna verses on the great vehicle and the heart of dependent origination. Trans. R.C. Jamieson. 2001 edition. New Delhi: D.K. Printworld. ———. 2007. Nāgārjuna’s precious garland Buddhist advice for living and liberation. Trans. Jeffrey Hopkins. Ithaca: Snow Lion Publications. ———. 2013. Nāgārjuna’s Middle Way: Mulamadhyamakakarika. Trans. Mark Siderits and Shoryu Katsura. Boston: Wisdom Publications. Reat, Noble Ross, trans. 1993. The Śālistamba-sūtra: Tibetan original, Sanskrit reconstruction, English translation, critical notes (including Pāli parallels, Chinese version, and ancient Tibetan fragments). Delhi: M. Banarsidass Publishers. Schoening, Jeffrey Davis. 1995. The Śālistamba Sūtra and its Indian commentaries. Vol. 1, 2 vols. Arbeitskreis für Tibetische und Buddhistische Studien Universität Wien. Siderits, Mark. 2014. Causation, ‘Humean’ causation and emptiness. Journal of Indian Philosophy 42: 433–449. https://doi.org/10.1007/s10781-­013-­9206-­3. Vasubandhu. 1988. Abhidharmakosabhasyam, Vol. 4. Trans. Louis de La Vallée Poussin and Leo M. Pruden. 4 volume edition. Berkeley: Asian Humanities Press. ———. 1992. Abhidharmakośabhāṣyam, Vol. 2. Trans. Louis de La Vallée Poussin. 4 vols. Berkeley: Asian Humanities Press.

Part II

Metaphysics

A Post-Reductionist Buddhism? Matthew MacKenzie

1 Introduction Perhaps more than any other contemporary scholar, Mark Siderits has illuminated the deep connections between ontology, explanation, epistemology, and philosophy of language in Indian Buddhist philosophy. His ground-breaking interpretations of Abhidharma and Madhyamaka—particularly concerning reductionism, emptiness, and the two truths—have largely set the terms of debate in Anglophone Buddhist philosophy. This chapter is very much in the spirit of Siderits’ work, though it will reach conclusions somewhat at odds with his own. The first part of the chapter will examine the ontological and explanatory reductionism of much Abhidharma thought. The thorough-going reductionism of Abhidharma yields a two-tiered ontology of ultimately real and (merely) conventionally real entities, and a correspondingly two-tiered account of ultimate and conventional truth. The second part of the chapter will take up the Madhyamaka critique of Buddhist reductionism and the philosophical consequences of the view that all things are empty (śūnya) of inherent existence (svabhāva). If all things are empty, then arguably Abhidharma reductionism is undermined. That is, if everything is empty of inherent existence, then the project of reducing things to that which exists inherently is doomed. Moreover, as Sidertis has so forcefully argued, the Madhyamaka view of emptiness undermines the Abhidharma distinction between two truths. As he provocatively puts it, for the Mādhyamika, “the ultimate truth is that there is no ultimate truth” (Siderits and Katsura 2013, p. 273). Hence, if all things are empty, all real entities are conventionally real, and all truths are conventionally true. Yet, as discussed in part three of the chapter, this raises the specter of the ‘dismal slough’—a pernicious ontological and epistemic flattening that threatens to undermine the critical and revisionist force M. MacKenzie (*) Colorado State University, Department of Philosophy, Fort Collins, CO, USA e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_11

231

232

M. MacKenzie

Buddhist philosophy. How might a Mādhyamika avoid the dismal slough? One approach, explored in Siderits’ more recent work, is to, as it were, transpose Buddhist reductionism into a Mādhyamika key. Here the strategy is to show that while Buddhist reductionism cannot provide an account of ultimate reality, it still offers the best account of conventional reality. It will be argued that recent work in the philosophy of science gives reasons to doubt this approach. Finally, section four of the chapter will explore an alternative, post-reductionist strategy for avoiding the dismal slough. This alternative is based on three key ideas: 1. Ontological parity: there is no ontologically privileged level of reality. 2. Conceptual-explanatory pluralism: we use and need an irreducible plurality of conceptual and explanatory schemes. 3. Epistemological pragmatism: an embrace of fallibilism and grounding warrant in successful practice.

2 Buddhist Reductionism What Siderits calls ‘Buddhist Reductionism’ (1997) is based on a rational reconstruction of certain key ideas of Indian Abhidharma thought.1 On this account Abhidharma Buddhists were ontological and explanatory reductionists concerning composite entities such as persons or chariots. As Siderits explains: To be a reductionist about a certain kind of thing is to hold that the existence of a thing of that kind just consists in the existence of certain other sorts of entities, entities that are thought to be in some sense more basic or fundamental … To be a reductionist about persons (a Reductionist) is thus to hold that the existence of a person just consists in the existence of a large number of psychophysical elements: body parts, feelings, perceptions, desires, and so on. (Sidertis 2016, p. 264)

The ‘psychophysical elements’ here are dharmas, the ephemeral mental and physical elements to which persons are reducible. On the Abhidharma account, suitably related dharmas can be understood as forming the five skandhas (bundles): rūpa, vedanā, saṃjñā, saṃskāra, and vijñāna. The existence of persons, then, just consists in these dharmas, suitably connected. Further, Siderits remarks: The continued existence of a person must then be said to consist in a causal series of sets of suitably arranged psychophysical elements: these body parts only exist for a while but cause similar successor parts to arise; this feeling only exists for a while but causes a successor desire, and so on. Buddhist Reductionists hold, then, that the existence of a person just consists in the occurrence of a causal series of psychophysical elements. (Sidertis 2016, p. 265)

Now, since persons are reducible to impersonal, ephemeral psychophysical elements, this raises the question of the ontological status of persons. In order to see  More specifically, Buddhist Reductionist ideas are developed in the Sarvāstivāda and Sautrāntika Abhidharma schools. 1

A Post-Reductionist Buddhism?

233

how the Buddhist Reductionists answer this question, we first need to understand their account of the doctrine of two truths and two realities. As Vasubandhu writes in the Abhidharmakośa, “An entity, the cognition of which does not arise when it is broken and, mentally divided, is conventionally existent like a pot. Ultimate existence is otherwise” (Pradhan 1975, p. 334).2 That is, composite entities are reducible to their basic parts and therefore exist conventionally (samvṛtisat). In contrast, the irreducible elements are dravyasat (“foundationally existent”) and therefore exist ultimately (paramārthasat). Conventional truths (samvṛtisatya) concern conventionally existent entities such as pots or persons, whereas ultimate truths concern ultimate entities or the way things ultimately are. On this account, then, persons are merely conventionally real and the truths concerning persons are mere conventional truths. Furthermore, according to Siderits’ account of Buddhist Reductionism, we can say that persons to do not belong in our ultimate ontology and that ‘person’ is a convenient designator (1997, p. 463). Yet this formulation is somewhat ambiguous. It might imply a layered ontology consisting in a foundational level of basic entities as well as a layer (or layers) of derivative entities. On this view, the conventional reality of persons carries an ontological commitment to persons, but denies that persons are among the basic entities in the ontology.3 On the other hand, the formulation might imply that the ultimate ontology is exhaustive, and therefore that there is no ontological commitment to merely conventionally real entities such as persons. That is, there aren’t really any such entities as persons, but there may be reason to think and talk as if there are. While both of these views have some plausibility as interpretations of Abhidharma, I take Siderits to be committed to the second interpretation. On his account, the Reductionist is not ontologically committed to conventionally real entities. In this regard the Reductionist and the Eliminativist agree on matters of ontology. The difference between the Reductionist and the Eliminativist, then, is not ontological, but semantic. The Reductionist holds that it is still useful to talk about persons—or more technically, to retain a conceptual scheme and a discourse that make reference to persons—while the Eliminativist does not. This way of understanding the conventional/ultimate distinction immediately raises two questions. First, if there is no ontological commitment to conventional entities, why retain a discourse that treats them as real, albeit conventional? If there aren’t really any composite entities (no persons, chariots, or what have you), then why bother? Second, if there is no ontological commitment to conventional entities, then how can there be conventional truths? Isn’t it just false that there are persons? With regard to the first question, on Siderits’ answer is straightforwardly pragmatic. The criterion of conventional truth is consistent successful practice (Sidertis 2007, p.  74). While persons may not ultimately exist, a conceptual scheme and many of the practices that treat persons as real facilitate successful practice and are

 ‘yatra bhinnena tadbuddhir anyāpohe dhiyā ca tat | ghaṭārthavat saṃvṛtisat paramārthasad anyathā ||’ AK 6.4. 3  Something along these lines is developed in Ganeri (2007) Ch. 6. 2

234

M. MacKenzie

thereby defensible on Buddhist grounds. But what constitutes successful practice? For Siderits, the answer is utilitarian. “Surely everyone could agree,” he writes, “that successful practice is practice that brings about more pleasure and happiness, and less pain and suffering” (Sidertis 2007, p. 75). For instance, ultimately there is no personal identity over time, because there are no persons. However, it can make sense to have practices that presuppose personal identity because of the causal connections between some sets of skandhas at one time and other sets of skandhas at a future time are relevant to how much impersonal suffering occurs in the world. In teaching a child to brush her teeth regularly, one might help her to see that if she doesn’t take her of her teeth now, she might have painful cavities (and dental work) in the future. The child comes to identify with her future self and learns long-term prudential concern. Yet from a Buddhist Reductionist point of view, this is merely a conventional conceptual fiction. What really matters here, and what justifies the fiction, is the causal connections between skandhas and the amount of suffering avoided. As Siderits puts it, “The conventional truth is that we are persons. This is conventionally true because it is ultimately true that these present skandhas are the cause of the future skandhas in this series. So what these skandhas do will affect the welfare of those future skandhas. This is why thinking of ourselves as persons results in greater overall welfare” (2007, p. 75). With regard to the second question, on Siderits’ view, the absence of an ontological commitment to conventionally entities does not entail the falsehood of statements that refer to (or quantify over) such entities. That is, as seen above, ‘there are persons’ can be conventionally true even if, ultimately, there are no persons. ‘There are persons’ is conventionally true because believing it consistently leads to successful practice. So is it conventionally true, but ultimately false that there are persons? No. According to Siderits, at the ultimate level, ‘there are persons’ is neither true nor false. Any statement that asserts or presupposes the existence of conventional entities is neither ultimately true nor ultimately false (Sidertis 2007, p. 143). There is, therefore, a semantic partition between the conventional and ultimate levels of discourse. For the Buddhist Reductionist, a statement is ultimately true only when it corresponds to ultimate reality. The truth-makers of ultimate truths are ultimately real entities. Furthermore, Buddhist Reductionists are committed to bivalence at the ultimate level. Hence, on Siderits’ account, Buddhist Reductionists are metaphysical and semantic realists at the ultimate level. In contrast, conventional discourse allows for bivalence failure and involves a pragmatic rather than a correspondence account of truth. Indeed, this is to be expected, since the conventional level “always involves accommodations to human interests and cognitive limitations” (Sidertis 2015, p. 158). Buddhist Reductionists, then, are metaphysical and semantic anti-realists at the conventional level. Yet, while the Buddhist Reductionist maintains sharp ontological and semantic distinctions between the conventional and the ultimate, it is important to note that the domains are not entirely separate. Despite being the realm of convenient fictions, the conventional does not simply float free from the ultimate. Rather, for Buddhist Reductionists, the conventional is indirectly grounded in the ultimate. On this view, dharmas are paramārthasat and therefore genuinely real and fundamental

A Post-Reductionist Buddhism?

235

(dravyasat). Further, they are dependently originated (pratītyasamutpanna) and the causal connections between them form the causal structure of ultimate reality. Sentient beings interact with and are constrained by ultimate reality in the pursuit of their desires and interests. So, while ultimately real entities and their causal relations cannot serve as the truth-makers of conventional claims, they can causally ground and constrain the conventional. In short, if the conventional were to float free from the ultimate, it could not consistently lead to successful practice in the long run.

3 The Madhyamaka Critique As discussed in the previous section, Buddhist Reductionism constitutes a type of anti-realism about everyday composite entities, including persons. Such entities may be pragmatically or conventionally real (samvṛtisat), but they are not ultimately real (paramārthasat). The being of these entities is fully accounted for in terms of more basic entities; they are fully analytically and ontologically decomposable. Thus, they have a merely derived nature (parabhāva), rather than their own irreducible intrinsic nature (svabhāva). Further, conventionally real entities must be epiphenomenal because if they were to have their own causal powers, they would not be completely reducible. Hence, according to the Buddhist Reductionists, all causation is microcausation—that is, real causation4 occurs only between simple, momentary dharmas. Further, the genuine causal powers of these entities are determined by their intrinsic natures. Notice, then, that this two-tiered ontology rests on a radical dichotomy between the entities with a purely extrinsic nature (parabhāva) and those with a purely intrinsic nature (svabhāva). It is this view that is forcefully criticized by Nāgārjuna and subsequent Madhyamaka thinkers. While it is not always clear which Abhidharma view is being targeted by Nāgārjuna, there can be no doubt that his primary philosophical target is the notion of svabhāva. An entity has svabhāva when it is ontologically independent of other objects, has an intrinsic and fixed nature or essence, and can be individuated mind-independently. Thus, the semantic range of ‘svabhāva’ overlaps not only with our notions of ‘substance’ and ‘essence’, but also with our notions of a ‘thing-in-itself’ and an absolutely or ‘really real’ existent. Importantly, then, to say that an entity is (or has) svabhāva is not simply to claim that it exists, but rather to specify its mode of existence: it is claimed to exist independently or absolutely. Mādhyamikas deny that anything could have this mode of existence and point out that the deep assumption that to be real is to be svabhāva inexorably leads to paradox. To say that an entity lacks svabhāva is just to say that the entity is empty (śūnya). Moreover, the emptiness of phenomena is said to be an implication of

 I include regularity theories of causation among accounts of real causation here.

4

236

M. MacKenzie

phenomena being dependently originated (pratītyasamutpanna) (and vice versa). As we see in Nāgārjuna’s Mūlamadhyamakakārikā (24.18–19): Dependent origination we declare to be emptiness. It [emptiness] is a dependent concept; just that is the middle path. There being no dharma whatsoever that is not dependently originated, It follows that there is no dharma whatsoever that is non-empty. (Siderits and Katsura 2013, pp. 277–278)

Yet, if all dharmas are empty of svabhāva, then no dharmas are ultimately real. Indeed, since according to Nāgārjuna, nothing could have svabhāva, nothing could be ultimately real in the sense specified by Buddhist Reductionists. And if nothing could be ultimately real, what becomes of the distinction between ultimate and conventional? According to Buddhist Reductionism, ultimately real entities are required to serve as the truth-makers of ultimate truths and as the indirect bases for conventional truths. By pulling the rug out from under the Buddhist Reductionist view, it would seem that the conventional-ultimate distinction collapses. However, Nāgārjuna too is committed to the two truths view (24.8–10): The Dharma teaching of the Buddha rests on two truths: Conventional truth and ultimate truth. Who do not know the distinction between the two truths, They do not understand reality in accordance with the profound teachings of the Buddha. The ultimate truth is not taught independently of customary ways of talking and thinking. Not having acquired the ultimate truth, nirvana is not attained. (Siderits and Katsura 2013, pp. 272–273)

So, rather than reject it, the Mādhyamika must reinterpret the two-truths view. In his commentary on MMK, Candrakīrti gives three distinct glosses on “conventional” (samvṛti). The first is “concealing,” that is, those ways of thinking and speaking (vyavahāra) that conceal the ultimate nature of things. The second is “mutual dependence,” such as between cause and effect, or part and whole. The third refers to the various customary practices of ordinary people in the world. As Sidertis and Katsura characterize it, “conventional truth is a set of beliefs that ordinary people (loka) use in their daily conduct, and it is conventional (samvṛti) because of its reliance on conventions concerning semantic and cognitive relations” (Siderits and Katsura 2013, p.  272). Note here that these glosses on the conventional are not based on the divisibility or mereological reducibility of conventional phenomena. With regard to ultimate truth, it consists in “the faultless realization of the noble ones (āryas).” Here Siderits and Katsura offer two possible ways to understand this idea. On the first interpretation, the realization is an insight into the true nature of things, namely that all dharmas are empty and therefore conventional. Ultimate reality, then, would be absent of all dharmas. On this view, there is a real way

A Post-Reductionist Buddhism?

237

ultimate reality is and the ultimate truth is content of an enlightened being’s grasp of this reality. On the second interpretation, “the ultimate truth according to Madhyamaka is just that there is no such thing as the way reality ultimately is. Or to put this in a somewhat paradoxical way, the ultimate truth is that there is no ultimate truth. On this reading, what the āryas realize is that the very idea of how things really are, independently of our (useful) semantic and cognitive conventions, is incoherent.” (Siderits and Katsura 2013, pp. 272–273) I take it that the second interpretation is the one Siderits prefers. On this view the ultimate truth is not ultimate because it corresponds to the ultimately real (svabhāvic) entities or because it corresponds to the true nature of ultimate reality independently of our conventional modes of thinking and speaking. Rather, the ultimate truth is a kind of meta-truth concerning the incoherence of the very idea of a trans-­conventional way things are. On this reading, Madhyamaka is a form of metaphysical anti-­ realism, that is, it constitutes a thorough-going rejection of metaphysical realism. According to Hilary Putnam, metaphysical realism can be characterized as follows: On this perspective, the world consists of some fixed totality of mind-independent objects. There is exactly one true and complete description of ‘the way the world is’. Truth involves some sort of correspondence relation between words or thought-signs and external things and sets of things. [We can] call this perspective externalist perspective, because its favorite point of view is a God’s Eye point of view. (1981, p. 49)

Elsewhere Putnam elaborates on the first point by claiming that the metaphysical realist is committed to the belief in a ‘ready-made world’: the idea that the world is uniquely and mind-independently partitioned into objects, properties, and relations. Moreover, these objects are ‘things-in-themselves’ that have fixed intrinsic natures independently of our interest, concepts, and descriptions. As Putnam says of this view, “the world divides itself up into objects and properties in one definite unique way” (1998, p. 190). There will be a uniquely complete and correct description of reality and the correctness of this description will involve correspondence between the description and the metaphysically pre-given way the world is. According to the anti-realist reading of Madhyamaka, the view that all things are empty of svabhāva implies the rejection of each of the above claims. For Nāgārjuna and later Mādhyamikas, emptiness implies the absence of ontological independence and with it the idea that a phenomenon could have a fixed or absolute nature. Some Tibetan commentators on Madhyamaka philosophy have distinguished three dimensions of dependence relevant for understanding emptiness (Westerhoff 2009). First, there is causal dependence: an entity depends for its existence on its causes and conditions. Second, there is mereological dependence: a composite entity existentially depends on its parts. We may also include in this category other non-causal dependence relations such as the interdependence between object and quality or agent and activity. Third, and according to these Tibetan commentators the most subtle form, is conceptual dependence: an entity is dependent on a basis of conceptual designation, a designating term or concept, and a designating cognitive event. Hence, to claim that an entity is dependently arisen (and therefore empty) is to do

238

M. MacKenzie

more than to point out its causal origin; it is to claim that it is constitutively embedded in a network of causal, ontic (including mereological) and conceptual relations. On the anti-realist interpretation, to hold that all things are empty is to give up the picture of a ready-made world. If all things are empty, there is no unique, mind-­ independent partitioning of the world into objects, properties, and relations (or, for that matter, dharmas, skandhas, and santānas). Yet, pointing to an entity’s causal and mereological dependence does not obviously tell us anything about its ultimate ontological status. To say that a tree depends on its causes and conditions and on its parts or constitutive processes does not seem to entail a rejection of metaphysical realism about trees. On the other hand, if trees are not ontologically fundamental phenomena, we will need to look elsewhere for the target of this line of reasoning. Indeed, a main target of Madhyamaka dialectic is precisely those entities that are meant to provide the ontological foundation on which less fundamental entities rest. That is, the real candidates for being svabhāva are those entities, such as the Buddhist Reductionists’ dharmas, that are supposed to be at the terminus of a series of ontological dependence relations. Such entities are ontologically independent (svabhāva) precisely in contrast to all those other entities which ontologically dependent (parabhāva). If they are to be ontologically foundational or ultimately real, they must be ontologically independent, otherwise one will need to keep following the chain of ontological dependence to its ultimate foundation elsewhere. Moreover, an ontologically independent entity will need to have an independent nature or essence as well—it must be what it is independently of what other things are. Thus, the totality of svabhāvic entities will constitute the ultimately real foundation of the world and it will be a world that, as Putnam says, ‘divides itself up’ in an absolute way.

4 Dismal Slough One of the most interesting implications of the Madhyamaka rejection of metaphysical realism is that it opens up the possibility of a rehabilitation of the conventional world and conventional truth (MacKenzie 2008). If there are no ultimately real entities and no ultimate truth-makers, then all entities and all truths are conventional. Even the claim ‘all things are empty’ is conventional in that it quantifies over conventional entities and it is formulated within a particular language and conceptual scheme. The upshot is that the conventional is no longer epistemically, semantically, or ontologically second-rate in comparison to the ultimate. Thus, conventional entities are as real as anything could be and conventional truths are just plain true (Sidertis 2015). Again, a comparison with Putnam’s internal realism is instructive.5 He writes:

 Of course, it is well known that Putnam eventually abandoned the internal realism he defending in the 1980’s. 5

A Post-Reductionist Buddhism?

239

Realism with a capital ‘R’ doesn’t always deliver what the innocent expect of it. If there is any appeal of Realism which is wholly legitimate it is the appeal to the commonsense feeling that of course there are tables and chairs, and any philosophy that tell us that there really aren’t—that there are really only sense data, or only ‘texts’, or whatever, is more than slightly crazy. And yet, in searching for those things that are real in the metaphysical realist’s sense— those things that are really real—what the Realist delivers is not the common sense world of tables and chairs. Rather, the Realist admits, all there really is … is what ‘finished science’ will say there is-whatever that may be. [The innocent] is left with a promissory note for She Knows Not What, and the assurance that even if there aren’t tables and chairs, still there are some Dinge an sich that her ‘manifest image’ (or her ‘folk physics’, as some Scientific Realists put it) ‘picture’. (Putnam 1998, pp. 3–4)

Yet, once metaphysical realism is rejected, there would seem to be little basis for the kind of ontological foundationalism that promises the really real and the expense of the everyday world of our experience. First, because, as mentioned above, only that which is really real—real in an absolute way—could serve as an ontological foundation. But the rejection of svabhāva and of metaphysical realism entail the rejection of anything existing in the absolute way. Second, because the idea of a really real ontological hierarchy of entities—a uniquely correct, well-founded chain of ontological dependence relations bottoming out in an ontological foundation—also makes little sense once metaphysical realism is abandoned. Thus the rejection of metaphysical realism leads to a kind of ontological parity, that is, the view that there is no ontologically privileged level of reality. And this ontological parity opens up the possibility of a rehabilitation of conventional reality in that the conventional world is now seen to be as real as anything could be.6 Shifting from the ontological to the semantic domain, we can see a similar pattern. In rejecting metaphysical realism, Madhyamaka rejects the idea that truth consists in correspondence with metaphysically real (svabhāvic) truth-makers. No truths are ultimate in that sense. Some truths may be soteriologically ultimate, but if they are discursively articulable, then they are also conventional.7 In giving up metaphysical realism, Putnam originally endorsed an epistemic account of truth (Putnam 1981). However, Sidertis (2016) has argued persuasively that this is not the most promising route for a contemporary philosophical reconstruction of Madhyamaka. Indeed, I agree with Priest et al. (2011) that a deflationist view of truth is more promising in this regard. A deflationist account of truth affirms uncontroversial semantic features of truth, such as that

is true if and only if p. However, the deflationist refuses to offer a philosophically substantive account of truth in terms of, for example, correspondence to reality or idealized warranted assertability. Rather the deflationist will merely affirm that “Devadatta is in the house” is true if and only if Devadatta is in the house, and that “all things are empty” is true if and only if all things are empty. On this interpretation of Madhyamaka,

 Though it is worth emphasizing again that there is an alternative reading of Madhyamaka as offering a kind of global error-theory (Tillemans 2011). 7  I will leave aside here the complex question of non-discursive, non-conceptual (niṣprapañca) truths. 6

240

M. MacKenzie

some truths may be soteriologically second-rate, but none are semantically second-­ rate (Sidertis 2011). My argument so far has been that, in rejecting the metaphysical realism entailed by the svabhāvic paradigm, Madhyamaka may have the philosophical resources to develop a (deflationary) rehabilitation of conventional reality and conventional truth.8 And yet, this deflationist rehabilitation of the conventional may come at the significant cost of sliding into the ‘dismal slough’ of pure conventionalism (Tillemans 2011). In the Prasannapadā, Candrakīrti cites from the Ratnakūṭa the following passage: The world (loka) argues with me. I don’t argue with the world. What is agreed upon (saṃmata) in the world to exist, I too agree that it exists. What is agreed upon in the world to be nonexistent, I too agree that it does not exist. (Tillemans 2011, p. 151)

Taken as a gloss on conventional truth in Madhyamaka, this passage implies that it is agreement with the world (lokaprasiddha) that establishes conventional truth as truth. This way of rehabilitating the conventional, though, has a significant cost. As Tillemans frames the problem: Most of us would agree that the potential flattening of the normative roles of truth and knowledge … is indeed quite dismal. It is a trivialization of the idea of truth to think that we could somehow settle what is true by periodically taking inventories for determining of what people believe to be true at given times and places. (2011, p. 152)

The worry then, is that what I have called ontological and semantic parity may turn out to be a mere ‘flattening’ that collapses to the distinction between real and unreal and between truth and falsity. Yet it is important to see that the kind of deflationism sketched above is not committed to the lokaprasiddha account of conventional truth or reality. The deflationist about truth does not propose that

is true if and only if most people accept p. Rather, she proposes that

is true iff p. Moreover, in offering a substantive criterion of truth, the lokaprasiddha account of conventional truth goes against the grain of deflationism every bit as much as a correspondence or coherence theories. Likewise, the ontological deflationist countenances ordinary first-order existence claims, such as ‘there a five chairs in the room’ or ‘Denmark is a real country’. And again, there is no necessary connection between the truth of those claims and what the majority of people happen to accept. Ordinary existence is to be established through ordinary means, which rarely involved opinion polling. On the other hand, the ontological deflationist will also want to avoid questions such as, ‘Does the chair exist, or are there only simples arranged chair-wise?’ So, I think the Madhyamaka deflationist can avoid the dismal slough of pure conventionalism. However, Siderits raises a more difficult problem for the philosophically quietist deflationist. On his view, the deflationist may still have “reason to worry about the flattening of truth.” His statement of the problem is worth quoting at length:

 For more on the notion of ontological deflationism in this context see my (2008).

8

A Post-Reductionist Buddhism?

241

We know that reductive methodologies have led to spectacular advances in the sciences. The techniques of modern medicine that resulted in improvements in health, and the electronics that made advanced information technologies possible, were developed by applying reductive methods, in biochemistry and solid state physics respectively. If the cell is not just an assemblage of organic molecules, why should treating as such prove so effective at disclosing ways to intervene in its behavior? Buddhist antirealism may not be vulnerable to the charge of relativism; it may have the resources to explain how to resolve such disputes as whether childhood vaccination causes autism. But a Buddhist quietism may lack the resources to account for the science behind the development of vaccines. (Sidertis 2015, p. 192)

The problem here is not one of the dismal slough of pure conventionalism. Rather, on Siderits’ account, the problem concerns explanatory and practical power. Buddhist quietism, such as the deflationist Madhyamaka currently under consideration, can affirm that childhood vaccination has large health benefits and that there is no causal link between vaccination and autism. But, according to Siderits, it is reductive methods that have the power to drive new discoveries, deeper explanation, and effective interventions. Madhyamaka deflationist quietism may avoid skepticism about conventional reality and ‘arguing with the world’ at the expense of any deeper understanding or effectiveness in the world.9 One possible solution to this problem is to adopt reductionism as a pragmatic method, rather than as a metaphysical position. In order to avoid the dismal slough of pure conventionalism as well as the mere explanatory shallows of deflationist quietism, Siderits recommends reinstating “something like the distinction between the two truths as understood by Abhidharma” (Sidertis 2011, p. 178). However, on this picture, rather than having all conventional truths, on the same explanatory level, there is a nested hierarchy of explanatory levels. There are conventional truths about the cell which are grounded in ultimate truths about the molecules to which the cell is reducible. Yet these ‘ultimate’ truths are themselves treated as conventional relative to deeper truths about atoms, and so on. Furthermore, those entities treated as ‘ultimate’ in a particular explanatory context will also be treated as having an intrinsic nature (svabhāva).10 Like the Abhidharma, there is a contrast between conventional and ultimate that tracks reducibility. However, in this new kind of reductionism there is an explanatory hierarchy of deeper levels of explanation without an ontologically or explanatorily most fundamental level. Instead, “that there is a final, ultimate truth is just another bit of upāya, something it is useful for us to believe insofar as it makes us strive to resolve our disagreements” (Sidertis 2011, p. 179)

 Of course, it would be fair for the quietist to reply that the whole point of quietism is to eschew the misguided desire for metaphysically deep explanation. In that case, we may say that there is a trade-off between quietist anti-skepticism and explanatory power. 10  See pp. 175–178 of Sidertis (2011) for a fuller discussion of the role of intrinsic natures in this account. 9

242

M. MacKenzie

5 Post-Reductionism So, unlike the forms of reductive metaphysical naturalism common today, Siderits’ argument for a rehabilitation of reductionism within Madhyamaka is explanatory and pragmatic, not metaphysical. We should adopt a form of reductionism, not because it carves nature at the joints, but because it is indispensable (or our best approach) to our explanatory and practical ambitions. But is reductionism really indispensable to scientific discovery and our general understanding of the world, even after the rejection of metaphysical realism? I have my doubts, and in this section I will sketch a possible alternative post-reductionist view consistent with those aspects of Madhyamaka with which we have been concerned. Now, there are many different conceptions of reduction, but our main concern here is explanatory reduction. According to Steven Horst, at the heart of the traditional notion of explanatory reduction are two distinctive characteristics. “First,” he writes, “it is explanation of a system in terms of its parts—chemical reactions in terms of the properties of the atoms, for example. Second, it is explanation in which, once one has an adequate theory of the lower-level system in hand, one can then demonstrate that all of the salient features of the higher-level system are a necessary consequence of these” (Horst 2014, p. 201). To revisit Siderits’ example, a reductive explanation of the cell is one that fully explains the relevant features of the cell in terms of the organic molecules which are its parts. In this sense, the cell would be just an assemblage of organic molecules. Why might one hold that reductionism is explanatorily indispensable? First, one might take reductionism to be a kind of second-order empirical generalization. It turns out that reductive explanations have been successful and will likely continue to be so for the range of phenomena studied in the sciences (Oppenheim and Putnam 1958). Second, one might take reductionism to be a norm or regulative ideal governing genuinely scientific forms of inquiry. On this view, there may be other fruitful forms of explanation, but to be truly scientifically serious our inquiries must aim for truly reductive explanations. Third, one might take a more metaphysical approach, adopting a layered or hierarchical ontology whereby higher levels metaphysically supervene11 on lower levels, bottoming out in the ontologically fundamental reduction base. These reasons are not mutually exclusive, but only the first two are available to the Mādhyamika reductionist, since the third relies on a metaphysical picture she must ultimately reject. In contrast, several strands of work in the philosophy of science provide reasons to be skeptical of the traditional explanatory reductionist picture. For instance, work on scientific explanation has found that primary forms of intra- and inter-theoretic explanations do not fit the traditional reductionist picture (Toulmin 1974; Churchland 1986; Darden and Maull 1977). Work concerning the autonomy of the special sciences, such as biology, too have called into question the reductionist view (Lewontin 1983; Rosen 1991). Finally, work at the intersection of philosophy of science and 11

 One might, of course, adopt some other form of ontological grounding, such as realization.

A Post-Reductionist Buddhism?

243

metaphysics has developed sophisticated challenges to reductionism and to the hierarchical view of the world (Cartwright 1983; Dupre 1993; Horst 2007). Indeed, Horst argues that: If there is a mainstream view about reduction in philosophy of science today, it is probably that reductionism, at least as conceived through most of the twentieth century, was a failed project, motivated by a misguided apriorism from Positivist philosophy of science, that the special sciences are autonomous and self-justifying, and that there are in fact a variety of types of inter-theoretic relations to be found at the intersections of various sciences, few of them amounting to inter-theoretic reduction. (2014, pp. 205–206)

Recall that the challenge for the Mādhyamika deflationist was to account for the kind of improvement in our epistemic practices and general understanding of the world seen in the development of the sciences. Siderits’ has argued that a rehabilitation of a kind of reductionism may provide the best response to this challenge. Perhaps he is right in this. But supposing that Horst’s account of developments in contemporary philosophy of science is broadly accurate, we have reason to think that reductionism is not necessary to meet the challenge. We can and have developed modes of inquiry with great explanatory depth and power that do not rely exclusively or even primarily on traditional explanatory reduction. So, what might a post-reductionist Madhyamaka look like? As mentioned in the introduction of this chapter, the post-reductionist reconstruction of Madhyamaka involves three basic elements: (1) ontological parity, (2) conceptual-explanatory pluralism, and (3) epistemological pragmatism. With regard to the first element, recall Nāgārjuna’s claim that “There being no dharma whatsoever that is not dependently originated, It follows that there is no dharma whatsoever that is non-empty.” The point is that the ubiquity and interminability of dependent origination implies that all phenomena are empty of svabhāva.12 Madhyamaka thus arrives at an ontological anti-foundationalism in which there is no well-founded hierarchy of dependence relations bottoming out in an ultimate independent foundation. As inquirers, we find ourselves perpetually in the midst of an open-ended network interrelated phenomena no domain of which has absolute priority over all the others. All of our explanatory work takes place within this horizon of dependent origination. With regard to the second element, conceptual-explanatory pluralism, Mādhyamikas argue that phenomena are dependent not just on their causes and conditions and on their the non-causal relations to other phenomena, but also on their ‘conceptual designation’. More specifically, one might say an entity depends on a basis of designation, a term or concept, and a cognitive event of designation. The basis of designation is a set of phenomena to which the concept can be applied. In the current context, we can think of the basis in terms of discernable features of the causal nexus. The concept brings the phenomena under some classification (‘person’, ‘object’, ‘cause’, ‘condition’). Finally, concepts don’t apply themselves  This claim about the ubiquity of dependent arising does not imply that everything is connected to everything else, but only that each phenomenon is related to at least one other phenomenon. See Sidertis (2011) for further discussion of this point. 12

244

M. MacKenzie

(if they did they would be svabhāva), so the cognitive event is the specific application of the concept within a specific domain. Hence, we have a kind of triangulation between a domain, a concept (or framework of concepts), and an agent using the concept to make sense of the domain. On this view, domains are not ready-made, but are articulated or partitioned relative to a conceptual framework and the cognitive activity of agents. This account of conceptual dependence suggests a form of conceptual relativity wherein terms such as ‘exist’, ‘object’, ‘relation’, and ‘cause’ have no absolute or independent meaning or application. Rather, they are themselves, as Nāgārjuna says, ‘dependent designations’ (upādāya-prajñapti). They are conventional (samvṛti) or transactional (vyavahāra) terms that get their sense and application from their role within a conceptual framework, which itself is grounded in on-going cognitive activity by sentient beings. Even the concept of emptiness itself is not absolute, but is itself empty. For Madhyamaka, as I have argued, the myth of the ready-made world is to be rejected—how we carve up the world is inextricably shaped by how minds like ours cognize that world. How then can the post-­ reductionist Mādhyamika understand those cognitive enterprises we call the sciences? Here again, the post-reductionist Mādhyamika could take Horst’s view that the sciences are cognitive “enterprises of modeling local features of the world (and of ourselves) in particular representational systems. Such models are local and piecemeal. They are also idealized in a variety of ways that can present principled barriers to their wholesale integration into something like a single axiomatic system” (2007, p. 5). The upshot here is that our ontology will be inextricably bound to our modes of inquiring, modeling, and intervening in specific domains. Thus neither a naïve folk ontology nor the absolute ontology of metaphysical realism is to be accepted. Rather the post-reductionist Mādhyamika will adopt a critical pragmatic ontology, wherein ontological commitments are driven by the overall project of making sense of various aspects of the world. This approach does not rule out the use of reductive methods of explanation in specific domains, but it does eschew the type of global explanatory reductionism historically associated with the unity of science program (Silberstein 2002). In this way the post-reductionist Mādhyamika can agree with Jay Garfield that in our inquiries we “pay attention to pratītyasamutpāda—to interdependence, and its multiple, multidimensional, interand intra-level character—and let a thousand entities bloom, requiring of each only that it genuinely toil and spin, accomplishing some real explanatory work” (2002, p. 75).13 With regard to the third element, epistemological pragmatism, there are two features I want to emphasize. First, the epistemological pragmatist understands epistemic warrant in terms of (long-run) pragmatic success. Indeed, from this perspective the point of cognition is to facilitate adaptive engagement between the cognitive

 That my hypothetical post-reductionist Mādhyamika would agree with Garfeld here, does not, of course, entail that Garfield would in turn agree with my overall account. 13

A Post-Reductionist Buddhism?

245

agent and its environment.14 Relatively simple cognizers may represent the world in fixed and simple ways. However, human cognizers have the ability to deploy a wide variety of representational systems. Moreover, we have the ability to both revise current and develop new modes of representing. This leads to the second important feature of epistemological pragmatism, namely fallibilism. On the picture I am sketching here, there is no certain foundation on which to build our knowledge. Further, the deep conceptual and explanatory pluralism of post-reductionist Madhyamaka rules out global coherentism as well. Our inquiries are plural, pragmatically constrained, and provisional, and so is the epistemic warrant that arises from those inquires.

6 Conclusion In this chapter, I have argued that the Madhyamaka critique of svabhāva, if successful, undermines reductionism, ontological foundationalism, and metaphysical realism. In doing so, Madhyamaka opens up the possibility of a (deflationist) rehabilitation of the conventional. Yet, it also threatens a slide into the dismal slough of pure conventionalism. In order to avoid the dismal slough and to give an account of the success of our epistemic practices, Siderits’ has argued for a rehabilitation of (a suitably modified) Buddhist reductionism within the Madhyamaka framework. The resulting picture is both philosophically interesting in its own right and an excellent example of power of Siderits’ brand of fusion philosophy. My aim here has not been to refute this view, but rather to explore the possibility of a post-­ reductionist form of Madhyamaka philosophy that could avoid the dismal slough and give a plausible account of our epistemic practices and their improvement over time. The post-reductionist view I have sketched here is in some ways quite different than Siderits’, but I hope that my indebtedness to his rigorous and insightful work is obvious.

References Cartwright, Nancy. 1983. How the laws of physics lie. Oxford: Oxford University Press. Churchland, Paul. 1986. Neurophilosophy: Toward a unified science of the mind-brain. Cambridge: MIT Press. Darden, Lindley, and Nancy Maull. 1977. Interfield theories. Philosophy of Science 44: 43–64. Dunne, John. 2004. Foundations of Dharmakīrti’s philosophy. Somerville: Wisdom Publications. Dupre, John. 1993. The disorder of things: Metaphysical foundations of the disunity of science. Cambridge: Harvard University Press.

 Both John Dunne (2004) and Laura Guerrero (2013) have interpreted Dharmakīrti’s notion of arthakriyā in similarly pragmatist terms. 14

246

M. MacKenzie

Ganeri, Jonardon. 2007. The concealed art of the soul: Theories of self and practices of truth in Indian ethics and epistemology. Oxford: Oxford University Press. Garfield, Jay. 2002. Empty words: Buddhist philosophy and cross-cultural interpretation. Oxford: Oxford University Press. Guerrero, Laura. 2013. Truth for the rest of us: Conventional truth in the work of Dharmakīrti. Dissertation. ProQuest. Horst, Steven. 2007. Beyond reduction: Philosophy of mind and post-reductionist philosophy of science. Oxford: Oxford University Press. ———. 2014. Beyond reduction: From naturalism to cognitive pluralism. Mind & Matter 12: 197–244. Lewontin, Richard. 1983. Biological determinism. Tanner lectures on human values. Vol. 4. Salt Lake City: University of Utah Press. MacKenzie, Matthew. 2008. Ontological deflationism in Madhyamaka. Contemporary Buddhism 9: 197–207. Oppenheim, Paul, and Hilary Putnam. 1958. The unity of science as a working hypothesis. In Concepts, theories and the mind-body problem. Minnesota studies in the philosophy of science II, ed. H.  Fiegel, M.  Scriven, and G.  Maxwell, 3–36. Minneapolis: University of Minnesota Press. Pradhan, Prahlad. 1975. Abhidharmakośabhāṣyam of Vasubandhu. Tibetan Sanskrit works 8. Patna: K. P. Jayaswal Research Institute. Priest, Graham, Mark Siderits, and Tom Tillemans. 2011. The (two) truths about truth. In Moonshadows: Conventional truth in Buddhist philosophy, ed. Cowherds, 131–150. Oxford: Oxford University Press. Putnam, Hilary. 1981. Reason, truth, and history. Cambridge: Cambridge University Press. ———. 1998. Renewing philosophy. Cambridge: Harvard University Press. Rosen, Robert. 1991. Life itself. New York: Columbia University Press. Siderits, Mark, and Shoryu Katsura. 2013. Nāgārjuna’s middle way: Mūlamadhyamakakārika. Somerville: Wisdom Publications. Sidertis, Mark. 1997. Buddhist reductionism. Philosophy East and West 47: 455–478. ———. 2007. Buddhism as philosophy: An introduction. Indianapolis: Hackett Publishing. ———. 2011. Is everything connected to everything else? What the Gopīs know. In Moonshadows: Conventional truth in Buddhist philosophy, ed. Cowherds, 167–180. Oxford: Oxford University Press. ———. 2015. Personal identity and Buddhist philosophy: Empty persons. 2nd ed. London: Ashgate. ———. 2016. Studies in Buddhist philosophy. Oxford: Oxford University Press. Silberstein, Michael. 2002. Reduction, emergence, and explanation. In Blackwell guide to the philosophy of science, ed. P. Machaner and M. Silberstein, 80–107. Oxford: Blackwell. Tillemans, Tom. 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, ed. Cowherds, 151–166. Oxford: Oxford University Press. Toulmin, Stephen. 1974. The structure of scientific theories. In The structure of scientific theories, ed. F. Suppe, 127–134. Urbana: University of Illinois Press. Westerhoff. 2009. Nāgārjuna’s Madhyamaka: A philosophical introduction. Oxford: Oxford University Press.

Madhyamaka, Ultimate Reality, and Ineffability Jay L. Garfield and Graham Priest

1 Introduction Mark Siderits’ contributions to Buddhist philosophy, and to the enterprise he likes to call “fusion philosophy,” are legion. We write this essay in celebration and warm appreciation of his career and his impact on the area. Philosophy is an odd subject: as Aristotle showed us in his discussion of Plato, the way one Western philosopher shows respect for other philosophers is to take their work seriously enough to disagree with it! So we will show our respect for Siderits in this time-honoured way. The subject concerns the interpretation of Nāgārjuna’s Mūlamadhyamakārikā (MMK) and the Madhyamaka tradition it inspired. We hold that according to this there is an ineffable ultimate reality/truth (satya). According to Siderits there is no ultimate reality, and a fortiori, no ineffable ultimate reality. Indeed, in his (1979) he argues that this is a crucial

J. L. Garfield (*) Smith College, Department of Philosophy, Northampton, MA, USA e-mail: [email protected] G. Priest CUNY Graduate Center, Department of Philosophy, New York, NY, USA © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_12

247

248

J. L. Garfield and G. Priest

difference between Madhyamaka and Advaita.1 In what follows, we will explain why we disagree.2

2 The MMK Let us start with the MMK. Here, Nāgārjuna states MMK XXIV: 8–10:3 The Buddha’s teaching of the Dharma Is based on two truths: A truth of worldly convention And an ultimate truth. Those who do not understand The distinction between these two truths Do not understand The Buddha’s profound truth. Without a foundation in conventional truth The significance of the ultimate cannot be taught. Without understanding the significance of the ultimate Liberation cannot be achieved.

Now, with texts as difficult as the MMK, differences of interpretation are always possible. However, a plain statement to the effect that there is an ultimate satya, of the kind Nāgārjuna makes here, leaves very little room for any interpretation according to which there is not (and, we might note, none of Nāgārjuna’s Indian or Tibetan commentators adopt any such interpretation). Now, this is not to say that any Mādhyamika thinks that the ultimate is ultimately existent—no Mādhyamika thinks that anything, including emptiness, is ultimately existent. But as Nāgārjuna says, there is an ultimate, even if it exists only conventionally, and even if to say so is only conventionally true. That the ultimate is ineffable requires more work.4 MMK XXV takes up the question (posed to the Buddha in some of the sūtras) of what happens to an enlightened person after they die. There, we find, MMK XXV: 17, 18:

 See that essay for references to those who have taken different sides on the issue.  We note that Jan Westerhoff (2018: 212), the editor of a volume of Siderits essays, holds that the existence of such a reality is the crucial difference between Madhyamaka and Yogācāra. Clearly, we cannot endorse this claim either, though the relationship between Madhyamaka and Yogācāra is not on the agenda here. Indeed, both Mādhyamikas and Yogācārins argue that there is an ultimate truth; all take it to be empty; and all take that emptiness to be a negation. The only difference between the proponents of these two schools is what they take to be the object of that negation. Mādhyamikas all take the object of negation to be intrinsic existence, or essence. Yogācārins, on the other hand, take the object of negation to be subject-object duality, or externality, and understand emptiness through the rubric of the three naturelessnesses adumbrated in the Saṃdhinirmocana-sūtra. See Garfield (2015, chapter 3). 3  Unless otherwise noted, translations from the MMK are from Garfield (1995). 4  What follows comes from Priest (2018: 4.7). 1 2

Madhyamaka, Ultimate Reality, and Ineffability

249

Having passed into nirvāṇa, the Victorious Conqueror Is neither said to be existent Nor said to be non-existent. Neither both nor neither are said. So when the victorious one abides, he Is neither said to be existent Nor said to be non-existent. Neither both nor neither are said.

And in MMK XXII, the matter is generalised to other predicates, MMK XXII: 11–12: ‘Empty’ should not be asserted. ‘Nonempty’ should not be asserted. Neither both nor neither should be asserted. They are used only nominally. How can the tetralemma of permanent and impermanent, etc., Be true of the peaceful? How can the tetralemma of the finite, infinite, etc., Be true of the peaceful?

The tetralemma (catuṣkoṭi) provides an exhaustive taxonomy of the statuses that a statement about some subject may have. It follows that nothing can be said of the status of the post mortem enlightened being (Tathāgata). A few verses later, we are told that reality and a Tathāgata have the same nature, MMK XXII: 16ab: Whatever is the essence of the Tathāgata This is the essence of the world.

The world here is reality; and it is not conventional reality: there is much to be said about that. Hence, it is ultimate reality, and we are being told that it, too, is ineffable.5 That nothing which language could express (mental fabrication) could apply to the ultimate is made quite explicit at MMK XVIII: 9, which says: Not dependent on another, peaceful and Not fabricated by mental fabrication, Not thought, without distinction. That is the character of reality.

If reality is without distinctions (and the context makes it clear that it is ultimate reality that is in question), one cannot say that it is thus or so, as opposed to thus and so. That is, it is ineffable.  In Siderits and Katsura (2013: 302), Siderits says that none of the four kotis of the cauṣkoṭi applies to the status of the post mortem enlightened person, not because it is ineffable, but because to say anything about them presupposes that they exist, which they do not, and so we have a case of presupposition failure. The problem with this suggestion is that presupposition failure does not take us out of the catuṣkoṭi. On standard treatments of presupposition failure, things which have false presuppositions are either false or neither true nor false (second and fourth koṭis). Nor does it help to say that such a sentence fails to express a proposition. It does so, since it can occur in a propositional context (e.g., Mark believes that the King of France is bald). Indeed, the standard definition of presupposition is that A presupposes B iff A and ~ A both entail B. Since we have an entailment, we must have propositions. Finally, and to drive the final nail in, in the case of ultimate reality, we cannot have reference failure, because, as we have seen, it does exist. 5

250

J. L. Garfield and G. Priest

Indeed, in all these matters Nāgārjuna is merely echoing what one finds in the Prajñāpāramitā Sūtras. In the Astahaṣrika, it is said:6 When again Subhuti has said, “surpassing the world with its gods, men, and Asuras, that vehicle [Mahāyāna] will go forth,” what then is the world with its gods, men and Asuras? The world of sense desire, the world of form, the formless world. If the world of sense desire, the world of form, or the formless world were Suchness, Non-falseness, unaltered Suchness, if they were the Unperverted, Truly Real, True Reality, That which is as reality is, the Permanent, Stable, Eternal, Not liable to reversal, existence and non-existence, then that great vehicle would not go forth, after having surpassed the world with its gods, men, and Asuras. But because the world of sense desire, the world of form, and the formless world have been constructed by thought, fabricated from fictions and feigned, because they are not as reality really is, but entirely impermanent, unstable, not eternal, liable to reversal, and non-existence, therefore the great vehicle will go forth, after having surpassed the world with its gods, men, and Asuras.

And in the Vajracchedikā Sūtra, we have:7 [The Buddha said]: Subhūti, words cannot explain the real nature of the cosmos. Only common people fettered with desire make use of this arbitrary method.

So, at least in the Madhyamaka tradition—both according to the sūtras that constitute its foundation and according to Nāgārjuna’s own treatise—there is an ultimate truth, and it is indeed ineffable.

3 Paradox Given all this, the obvious next question is why Siderits does not accept this reading of Nāgārjuna. The answer is simple. The ultimate is ineffable, but, as we have just seen, both Nāgārjuna and the Prajñāpāramitā Sūtras say a good deal about it. Now, talking of the ineffable is a contradiction. It is not a contradiction in the text, but it is a clear performative contradiction, as when someone says ‘I cannot speak’. Nor is this contradiction recondite: it is just plain obvious, and Nāgārjuna must have been aware of it. Nonetheless, nowhere does he do anything to suggest that he wishes to avoid it. Nāgārjuna, then, according to this understanding of the text, accepts this paradox. Perhaps this is hardly surprising; after all, Nāgārjuna deploys the catuṣkoṭi, a principle which allows for the possibility of things that are both true and false. Nāgārjuna, we conclude, is a dialetheist.8 Siderits is unhappy with the possibility that Nāgārjuna might accept a contradiction. After all, he argues, many verses of the MMK argue by reductio (albeit a four-­ case reductio, since each of the four cases generated by the catuṣkoṭi is considered); and, he asks, doesn’t a reductio argument depend on contradictions being false?:9  Conze (1979: 182).  Price and Wong (1990: 51). 8  Further, see Garfield and Priest (2003). 9  Siderits (2008: 118). In what follows, page references are alway to reprints, where these exist. 6 7

Madhyamaka, Ultimate Reality, and Ineffability

251

But the more interesting question is what dialetheism [the claim that some contradictions are true] would do the Mādhyamika’s ability to argue for their claim that all things are empty. I think that the answer is that the result would be rather dire. Nāgārjuna’s strategy is to use only reductio arguments. He seeks to demonstrate that the opponent’s various theses concerning the ultimate nature of reality invariably lead to contradiction, and so cannot be maintained. Now suppose he took this to show that ultimate reality has a contradictory nature, for instance in having an inexpressible nature that is expressible, or in having as its nature that it lacks a nature. If he is willing to countenance true contradictions, then the opponent might insist on revisiting his reductio arguments…. [T]he Mādhyamika is not well positioned to claim that only those contradictions that favour their own positions are true, while the contradiction derived from the opponent’s thesis is false. In that case the modus tollens argument to the falsity of the opponent’s thesis cannot get off the ground. The Mādhyamika would be left without a way of showing that all things are empty.

Siderits points out correctly that Nāgārjuna and his Prāsaṅgika followers rely on, and explicitly endorse relying on, reductio arguments.10 And, he argues, dialetheism would make it impossible to advance these arguments, since they typically draw inconsistent consequences from the opponent’s conclusions. If contradictions can be true, Siderits argues, then there is no reason to believe that the consequences adduced are false, and so no force to the apparent reductios. There are at least two problems with this argument.11 First, in order for a reductio to succeed, it is not necessary (nor, we would add in the present context, is it sufficient) for the consequence adduced from the opponent’s thesis to be contradictory. It is necessary only that it be unacceptable. After all, the argument is reductio ad absurdum, not reductio ad contradictionem. And some things are more absurd than some contradictions. Thus, the claim that Mark Siderits is a frog—although it is assuredly not a contradiction—is more absurd than the claim that the liar sentence is both true and false, which is contradictory. So, as long as Nāgārjuna infers unacceptable consequences from the target of his reductio, his use of reductio arguments can succeed, whether or not those consequences are contradictions; moreover, the simple fact that a contradiction is inferred does not by itself constitute a reductio ad absurdum. Second, Nāgārjuna’s (as well as other Mādhyamikas’) reductio arguments are often ad hominem. That is, they adduce consequences unacceptable to the opponent, showing the opponent that their position is unacceptable by their own lights. In such a case, nothing follows about what Mādhyamikas themselves might think. So, even if a Mādhyamika accepted some contradictions as true, when arguing with an opponent committed to consistency, the demonstration that their thesis entails a contradiction—even a contradiction acceptable to a Mādhyamika—would be a successful

 Whether in fact all of their arguments are reductio arguments is another matter, but one that is beside the present point. 11  Oddly enough, essentially these points are made by Siderits himself (2008: 119). 10

252

J. L. Garfield and G. Priest

ad hominem reductio. It follows that a commitment to reductio reasoning does not preclude dialetheism.12

4 Commentators on the MMK So much for Nāgārjuna and the MMK. Let us now turn to the Indian commentators on the text.13 Siderits holds that they reject the interpretation on the MMK which we have endorsed. The major argument for this is, again, that contradiction is unacceptable to them. We begin with two preliminary points. The first is that to take how commentators some hundreds of years later interpret a text as evidence as to how the text should be understood, though it carries some weight, is hardly definitive. To see this, one has only to remember how the Neoplatonists interpreted Plato’s texts. As most would now agree, this is not Plato’s interpretation. Next, we note that a sea change comes over Buddhist logic around the sixth century, with the work of Dignāga and Dharmakīrti. Under the influence of their debates with philosophers of the Nyāya school, they came to accept the Principle of Non-­ Contradiction, PNC (and the Principle of Excluded Middle). Subsequent Indian Buddhist philosophers followed them down this path. It is therefore unsurprising that after this time we find Buddhist philosophers endorsing interpretations of the MMK which do not commit it to contradiction. In exactly the same way, under the influence of the PNC, we find twentieth century philosophers advocating consistent interpretations of Hegel, against the patently obvious dialetheic interpretation of his texts. So we may restrict ourselves to interpreters of the text before the influence of Dignāga and Dharmakīrti was felt. Let us look at two of these, Bhāviveka and Candrakīrti (whom history, codified in Tibetan doxography, came to regard as having different interpretations of the text, Svātantrika and Prāsaṅgika, respectively). Let us start with the former. Bhāviveka is clearly on our side. Siderits himself quotes him as saying:14 [The nature of things] is cognized non-conceptually, and is entirely inexpressible, because of the fact that nothing is really arisen…

 In (2008: 119), Siderits has a rather different argument against contradictions being true: they do not express propositions, and a fortiori do not express true propositions; but on the same page Siderits endorses classical logic. In classical logic, contradictions express propositions, uniformly false ones. Indeed, since contradictions enter into valid inferences (such as Explosion and Conjunction-Elimination), they must express propositions. 13  Chinese commentators are a quite different matter. But since these are not on Siderits’ horizon, we need not go into the matter here. 14  Siderits (2013: 383). 12

Madhyamaka, Ultimate Reality, and Ineffability

253

Here, Bhāviveka not only explicitly asserts the existence of an ultimate reality—he is obviously not talking about conventional reality—but also its ineffability, explaining that concepts cannot be applied to it. Siderits tries to reinterpret Bhāviveka’s words:15 The point here is a subtle one that is easily overlooked: it is that the ultimate nature of reality is something that is inexpressible and only cognizable non-conceptually because there is no such thing as the ultimate nature of reality.

It hardly needs pointing out that to say that something is cognizable (only) non-­ conceptually (as Bhaviveka and Siderits both say) is hardly compatible with saying that it does not exist: it entails that it does.16 Let us now turn to Candrakīrti.17 Candrakīrti is as clear as Nāgārajuna that there are two satyas. In his Madhyamakāvatāra he says:18 6.23: All phenomena are understood to have two natures As understood through deceptive and correct perception, respectively. Reality is the object of correct perception. The object of deceptive perception is the conventional truth. Thus the transcendent buddhas understand without any error the nature of the two truths. All internal and external phenomena—such as cognitive dispositions and sprouts—are presented as having two natures: a conventional and an ultimate nature. Of these, the ultimate nature is the object of the distinctive insight of those who have achieved correct perception. Although it has no essence of its own, it is a nature. The other nature is that apprehended by ordinary people whose perception is clouded by the cataracts of ignorance, and who take things to have intrinsic existence. Thus, all things are seen to have two natures. (Candrakīrti 1992)

And Candrakīrti affirms that the ultimate truth is ineffable: in commenting on the homage verses of MMK, Candrakīrti writes:19 And not the slightest word or combination of letters is apposite, since the way things are is beyond conceptualization.

Now, one might nonetheless argue that Nāgārjuna and Candrakīrti are still no dialetheists, no matter how much it appears that they are committed to paradox. Thus, there are passages in the MMK in which Nāgārjuna might be thought to endorse the PNC explicitly. Thus, at XXV:14 we have:20 How could both non-being and being pertain to nirvāṇa. Just like light and darkness, both are not present in the same place.

 Siderits (2013: 383).  For further discussion of Siderits’ paper, see Deguchi et al. (2013b). 17  In his (1994), Siderits agrees with us on what follows. However, he walks back from this in the postscript to (1979), as it appears in (2016: 299–300). He should have stuck to his guns. 18  Candrakīrti (1992: 98). 19  Gelugpa Students’ Welfare Committee (2003: 4). Translations from this are ours. 20  Tillemans (2009: 74). 15 16

254

J. L. Garfield and G. Priest

But this is hardly decisive. First of all, it is not obvious that light and dark being in the same place at the same time is a contradiction. (It is certainly not an explicit contradiction.) It may just be an appropriate absurdum. And even if it is a contradiction, it hardly shows that Nāgārjuna accepts the PNC. After all, this contradiction may indeed be absurd by Nāgārjuna’s lights, or by those of his opponent, and it is the absurdity, not the inconsistency that is the issue. Moreover, to conclude that Nāgārjuna rejects all contradictions because he rejects one is a patently fallacious inference. (Compare: Mark Siderits is not a frog, so nothing is a frog.) In his gloss on this passage in Prasannapadā, Candrakīrti comments:21 For being and non-being too, there is no possibility for the two mutually contradictory things (parasparaviruddha) to be present in one place—that is, nirvāṇa. Thus, “how could could both being and non-being pertain to nirvāṇa?” The point is, they could not at all.

As this comment makes clear, the point just made about contradiction is specific to the predicate of being/non-being as applied to nirvāṇa; it is not a general logical point about the unacceptability of contradictions. Indeed, we are in the midst here of an argument for the conclusion that nirvāṇa is ineffable; so on those grounds alone, neither being nor non-being pertains to nirvāṇa, let alone both of them. Elsewhere in Prasannapadā, Candrakīrti says:22 You do not accept that there is repeated arising, and you do not accept that there can be an infinite regress. Therefore your argument makes no sense. You have contradicted your own thesis. We have asked about our opponent's account of arising. When the thesis and example are presented together with what follows from them as the argument is set out, the opponent cannot accept them. If the opponent is willing to contradict his own thesis, we can't argue with madmen. Therefore, without any embarrassment, we can say that the thesis and example, being contradictory, cannot be advanced.

But this comes immediately after a passage where he has shown that the Sāṃkhya position is committed to an infinite regress. And the Sāṃkhya (‘you’) did accept the PNC. Hence what we have here can clearly be interpreted as an ad hominem argument. There is no need for Candrakīrti himself to accept the Principle. Siderits cites this passage (without reference) to argue that Mādhyamikas regard as madmen those who endorse contradictions.23 But this interpretation is unwarranted. Candrakīrti calls this opponent a madman, and this is because the opponent takes himself to be advocating a consistent view, and yet contradicts himself; that is indeed mad (and this, perhaps, is an aberration which Siderits himself shares).24

 Cited in Tillemans (2009: 75), who gives this as evidence that Candrakīrti endorsed the PNC.  Gelugpa Students’eWelfare Committee (2003: 10). This is also cited as evidence that Candrakīrti accepted the PNC by Tillemans (2013: 86). 23  Siderits (2008: 118). 24  For further discussion of these matters, see Deguchi et al. (2013a). 21 22

Madhyamaka, Ultimate Reality, and Ineffability

255

5 Advaita Vedānta So much for the MMK and Madhyamaka. As we noted at the start, Siderits takes the non-existence of an ineffable ultimate reality to be a major difference between Madhyamaka and Advaita Vedānta, which certainly does endorse the existence of such a thing.25 And we agree that there is a difference between Advaita Vedānta and Madhyamaka. So, we might now ask, what is that difference? The answer is simple. For a thing to be empty (śūnya) is for it to depend for being what it is on something else. The core thesis of Madhyamaka is that everything is empty. And the everything includes the ultimate. Thus, one of the most crucial verses of the MMK (XXIV: 18) says: Whatever is dependently co-arisen That is explained to be emptiness. That, being a dependent designation, Is itself the middle way.

Lines 3 and 4 tell us that emptiness (‘That’) is as ontologically dependent as anything else (and that the middle way consists in neither reifying it nor taking it to be non-existent). ‘Emptiness’ (śūnyatā) is frequently used in the tradition to refer to the ultimate. Thus, the Heart Sūtra, tells us:26 Therefore, Śāriputra, in emptiness there is no form, no feeling, no perception, no dispositions, no consciousness; no eye, no ear, no nose, no tongue, no body, no mind; no visible object, no sound, no smell, no taste, no tactile sensation, no mental object; no sensory awareness; no cognitive awareness; no object of cognitive awareness.

It is clear that ‘emptiness’ is referring to some kind of reality, even if that reality is itself empty; and that is obviously not conventional reality; it is the ultimate.27 How to understand the claim that the ultimate is empty is a delicate matter, but we can say some things about it (despite its ineffability!). Emptiness is not a self-­ standing thing, an independent reality standing behind the conventional, as Brahman is the reality standing behind the world of illusion in the Vedānta system. Emptiness, as the Heart Sūtra tells us (‘form is empty; emptiness is form’) is simply the emptiness of phenomena; no phenomena; no emptiness. Since emptiness depends on (other) empty phenomena, it itself is empty. This is a very different account of the ultimate from the one we find in Advaita Vedānta, according to which the ultimate is not empty. The point that Brahman—the ultimate reality—has a nature, and is independent is made throughout the Upaniṣads, but we might point to the remarks that Śaṅkara quotes from the Chāndogya Upaniṣad in his Brahmasūtrabhāṣya:28

 So we agree with Siderits at least here!  For this translation, see Garfield (2016). 27  And even if Madhyamaka were not resolutely nominalist, it would clearly make no sense in the passage to take ‘emptiness’ to be referring to the property of being empty. 28  Potter (1991, Vol. III: 137–138). 25 26

256

J. L. Garfield and G. Priest

Brahman cannot lack its true nature

or later in the same Upaniṣad: the world originates from Brahman.

In the Kaṭha Upaniṣad, we find this remark:29 That indeed is the Pure. That is Brahman. That indeed is called the Immortal. On it all the worlds do rest…

Śankara, in his commentary on the Vedānta sūtras, I.1.2 writes:30 Brahman is that from which the origin of this world proceeds.

We could multiply citations endlessly, but the point is already clear: in the Vedānta system, it is important that Brahman is ultimately real, is the foundation of all other reality, and has its nature intrinsically. So, even though Siderits is correct that Brahman is as ineffable in Advaita as we have argued that śūnyatā is in Madhyamaka, Brahman is not śūnyatā. It is not even śūnya. And that—not the presence of an ultimate—is what distinguishes Advaita Vedānta from Mahdhyamaka (or at least is one distinction between these two schools).

6 The Ultimate Truth is That There is No Ultimate Truth Siderits is well known for his slogan: the ultimate truth is that there is no ultimate truth. This was first formulated in Siderits (1989), but as that essay and subsequent essays make clear, Siderits himself does not intend this as a paradox. ‘There is no ultimate truth’ is meant literally enough; but ‘The ultimate truth is that…’ simply means that ‘an awakened person is aware that…’. And so the slogan, he argues, should be read as the unparadoxical claim that “an awakened person is aware that there is no ultimate truth.” We prefer a more natural and straightforward—and paradoxical— understanding of the slogan. As we have argued, there is an ultimate reality. Nonetheless, since it is ineffable, there are no truths about it (or falsehoods, for that matter: if there were falsehoods, their negations would be true). But that statement (that there are no truths about it) is itself ultimately true; so there are. Understood this way, Siderits’ slogan is a direct expression of the paradox of ineffability; so undertstood, it is faithful to the words of the Prajñāpāramitā Sūtras, to Nāgārjuna’s MMK, and to the readings of many of its influential commentators. Siderits’ slogan is therefore best understood as the paradox it appears to be, his own auto-commentary on it to the contrary notwithstanding. (As Gadamer pointed out, authorial intent is not the same thing as meaning, and authors may not be their own best readers!) 29 30

 Radhakrishnan and Moore (1973: 48).  Radhakrishnan and Moore (1973: 511).

Madhyamaka, Ultimate Reality, and Ineffability

257

7 Conclusion We do not, of course, expect Siderits to agree with any of this. He, and those like him who subscribe to the PNC, will struggle with the text to find an interpretation which does not commit it to a violation of the Principle. This, we think it simply a mis-application of the principle of charity: an attempt to read Nāgārjuna through the eyes of Western shibboleths. In the interest of making Nāgārjuna sound cogent because consistent, it renders him incoherent. Better to believe simply what the text says, and, indeed, what Siderits himself so elegantly said—even though he may not have meant it.

References Candrakīrti. 1992. dBu ma la ‘jug pa’i bshed pa. Sarnath: Kagyud Relief and Protection Committee. ———. 2003. dBu ma rtsa ba’I ‘grel ba tshig gsal ba zhes bya ba gzhugs so. Sarnath: Gelugpa Students’ Welfare Committee. Conze, E., trans. 1979. The Large Sutra on Perfect Wisdom. Delhi: Motilal Banarsidass Publishers Pvt Ltd. Deguchi, Y., J. Garfield, and G. Priest. 2013a. How We Think Mādhyamikas Think: A Response to Tom Tillemans. Philosophy East and West 63: 462–435. ———. 2013b. Does a Table Have Buddha Nature? A Moment of Yes and No. Answer! But not in Words or Signs: A Response to Mark Siderits. Philosophy East and West 63: 87–98. Garfield, J., trans. 1995. The Fundamental Wisdom of the Middle Way. New  York: Oxford University Press. ———. 2015. Engaging Buddhism: Why it Matters to Philosophy. New  York: Oxford University Press. ———. 2016. The Heart Sūtra: Bhagavatī-Prajñāpāramitā-Hṛdaya-Sūtra. https://jaygarfield. files.wordpress.com/2016/08/heart-­sutra.pdf. Garfield, J., and G.  Priest. 2003. Nāgārjuna and the Limits of Thought. Philosophy East and West 53: 1–21. Reprinted as ch. 5 of Garfield’s, Empty Words. New York: Oxford University Press (2002), and ch. 16 of Priest’s Beyond the Limits of Thought, 2nd edn. Oxford: Oxford University Press (2002). Potter, K. 1991. Encyclopedia of Indian Philosophy, Vol. 3: Advaita Vedānta up to Śaṁkara and his Pupils. New Delhi: Motilal Banarsidass. Price, A.F., and M.L. Wong, trans. 1990. The Diamond Sūtra and the Sūtra of Hui-Neng. Boston: Shambala. Priest, G. 2018. The Fifth Corner of Four. Oxford: Oxford University Press. Radhakrishan, S., and C. Moore. 1973. A Sourcebook of Indian Philosophy. Princeton: Princeton University Press. Siderits, M. 1979. Distinguishing the Mādhyamika from the Advaita: a Field Guide. In Essays in Indian Philosophy, Jadavpur Studies in Philosophy, ed. S.R. Saha, 129–143. Calcutta: Allied Publishers. Reprinted as 6.2 of Siderits (2016). ———. 1989. Thinking on Empty: Madhyamaka Anti-Realism and Canons of Rationality. In Rationality in Question: On Eastern and Western Views of Rationality, ed. B.A. Scharfstein, 231–249. Leiden: Brill. Reprinted as 1.2 of Siderits (2016). ———. 1994. Matilal and Nāgārjuna. In Relativism, Suffering, and Beyond: Essays in Memory of Krishna Matilal, ed. P. Bilimoria and J.N. Mohanty, 69–92. Delhi: Oxford University Press. ———. 2008. Contradictions in Buddhist Argumentation. Argumentation 22: 125–133.

258

J. L. Garfield and G. Priest

———. 2013. Does a Table Have Buddha Nature? Philosophy East and West 63: 373–386. ———. 2016. Studies in Buddhist Philosophy, ed. J. Westerhoff. Oxford: Oxford University Press. Siderits, M., and S. Katsura. 2013. Nāgārjuna’s MiddleWay. Boston: Wisdom Publications. Tillemans, T. 2009. How do Mādhyamikas Think? Notes on Jay Garfield, Graham Priest, and Paraconsistency. In Pointing at the Moon: Buddhism, Logic, Analytic Philosophy, ed. J.  Garfield, W.  D’Amato, and T.  Tillemans, 83–100. New  York: Oxford University Press. Reprinted as ch. 3 of Tillemans (2016). ———. 2013. “How do Mādhyamikas Think?” Revisited. Philosophy East and West 63: 417–425. Reprinted as ch. 4 of Tillemans (2016). ———. 2016. How do Mādhyamikas Think? And Other Essays on the Buddhist Philosophy of the Middle. Somerville: Wisdom Publications. Westerhoff, J. 2018. The Golden Age of Indian Buddhist Philosophy. Oxford: Oxford University Press.

The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits and I Really Disagree Dan Arnold

1 Introduction: On Taxi Cabs and Two Truths In the course of a 2016 conference in Taiwan, I chanced to share a cab with Mark Siderits. We were, I recall, en route to some stunning coastal attraction, and, as typically when Mark is around, the break from the conference’s rigors involved no rest from intense philosophical conversation. Long having counted Mark as a cherished interlocutor, I found it familiar thus to be hashing out our philosophical differences over this and that. At one point, though, the familiar scenario was punctuated by an altogether unexpected development: Mark nonchalantly allowed that he and I didn’t much disagree in our understandings of Madhyamaka, after all. This left me agape. (Two other colleagues also shared the cab; I have witnesses!) Now, I myself had long thought Mark and I aren’t all that far apart on this. I’ve always thought, for example, that Mark’s Personal Identity and Buddhist Philosophy elaborates a cogent reading of Madhyamaka on which that school’s principal upshot, as on my own reading, is to recuperate conventional truth. The book’s concluding words eloquently express just the idea I would emphasize as upshot of my own interpretation: “If there is just truth, then it seems we might say that persons do after all exist.... That rivers and mountains are empty becomes the simple fact that there are rivers and mountains. That persons are empty becomes the simple fact that we are persons.”1

 Siderits (2003: 192–93, 202); see, too, p.195, note “o,” for acknowledgment of the famous Zen saying echoed here. It should be noted that the second edition of the book (2015) differs here. 1

D. Arnold (*) University of Chicago Divinity School, Chicago, IL, USA e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_13

259

260

D. Arnold

But despite my own sense of us as largely simpatico, I had long been under the impression that Mark, at least, thought we differed sharply, and so was gobsmacked by his taxi cab concession.2 Naturally, then, when I was invited months later to honor Mark’s groundbreaking work in Buddhist philosophy by contributing to the present volume, I got to thinking: just what do we disagree on? One thing over which we surely disagree is the tenability of reductionism. Mark has long argued that the Buddhist tradition’s principal commitments are compatible particularly with a physicalist version of reductionism,3 whereas I am instead inclined to think the Madhyamaka tradition of thought advanced what amounts to a devastating critique of any such approach. Among the difficulties in discussing this with Mark is that he well understands the philosophical challenges confronting reductionism. One of the most basic of these is well developed in one of my favorite of Mark’s essays: “Is Reductionism Expressible?” (2009) Insofar as reductionism is paradigmatically exemplified in Buddhist traditions by Abhidharma philosophy, that essay’s title question identifies a problem centering (as typically when interpreting Buddhist philosophy) on the two truths: Abhidharma’s reductionist redescription of conventionally real existents is proposed as ultimately true  – but the entertaining of that claim is itself intelligible only as conventionally real. After all, it is persons who entertain claims, and “persons” are just the kinds of things that Abhidharma analysis reduces. Against the view that reductionism therefore  cannot coherently be conceived, one might adopt a metalinguistic strategy aimed at showing the self-contradiction to be only apparent; the apparent contradiction merely reflects epistemic facts about creatures like us, and therefore cannot be taken as showing that the proposed account cannot make sense as ultimately true. The problem, though, is that this move just reproduces the original problem; if, as Mark puts it, the reductionist’s claim regarding persons is that “there ultimately are no such creatures, then no ultimately true statement can invoke facts about their interests and cognitive limitations and their resulting speech habits” (2009, 66). Despite the cogency with which he has himself scouted this challenge, Mark is optimistic regarding the possibility of a reductionist’s meeting it; I, on the other hand, take it that the self-referential incoherence sketched here is devastating.4 Whether or not I am right to think so, the burden of this essay is to warrant my exegetical case for thinking Madhyamaka is best understood as pressing much the same argument; that reductionism is not coherently intelligible is, on my reading,

 See Siderits (2006: 20).  See, e.g., Siderits (2001). 4  Mark’s essay here (2009: 69n14) issues a promissory note referring the reader to Siderits (2003, chapter 3). Regarding my reading of reductionism, Mark has said I confuse that with eliminativism (2006: 20); I would say, rather, that I am persuaded by Baker (1987: 9) that both approaches crucially “share the view that there are no irreducibly intentional entities or properties.” 2 3

The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits…

261

just what Madhyamaka aims to show.5 I want to argue, moreover, that appreciating as much is relevant to another issue over which Mark and I might differ (though this is rather less clear to me): whether or how Madhyamaka can propose its claims as ultimately true. On my view, there is no contradiction in thinking it really true that being a person cannot be exhaustively explained by any reductionist analysis. To that extent, I have long resisted Mark’s predilection for considering Madhyamaka a kind of anti-­ realism.6 Whether that is apt depends, of course, on just what “realism” means, and it may be only that terminological issue over which Mark and I finally differ. That cannot be decided, though, without doing a bit of heavy-lifting, and so I will begin by exploring the significance of Mark’s interpretation of Madhyamaka as chiefly motivated by semantic considerations, which seems to me to lend plausibility to his sense of Madhyamaka as generally anti-realist. My wager is that this characterization is less tempting if conventional truth is understood not as chiefly semantic, but instead as most saliently involving a personal level of description. On that reading, what Madhyamaka emphasizes is that conventional truth is intelligible only given an always already situated “taking” of things – a thought, I suggest, not all that far from the early Heidegger’s idea that being-in-the-world is an ontological fact about Dasein.7 Among the upshots of this is that any supposedly significant difference between semantic and ontological considerations will get less traction from the outset. The idea that personality resists statically “semantic” or “ontological” conceptions is reflected, I take it, in Candrakīrti’s thought that action, as theorized in the Sanskrit grammatical tradition, is the “personal” category, par excellence. Taking our bearings from the thought that “persons” are, above all, subjects of action,8 the idea that a person could make sense as anything like a mereological sum can immediately be recognized as misguided. That, I will suggest, is as Candrakīrti shows in arguing, in perhaps his most famous discussion, that things like chariots will not admit of exhaustively reductionist analyses – an argument he makes, perhaps unwittingly, in terms clearly shot through with the “Personalist” (pudgalavādin) schools of Buddhist thought that had by Candrakīrti’s time already been marginalized. While I am especially unlikely to persuade Mark that Madhyamaka thus has a basically personalist bent, I hope at least to suggest the sense it makes, on my  Here I’ll stipulate that by “Madhyamaka” I mean the thought of Nāgārjuna and Candrakīrti as I understand them. I would also note that in doing philosophy in conversation with Buddhist philosophers, Mark and I commonly face both philosophical and exegetical challenges, which represent the basic poles of the hermeneutic circle. Insofar as I was intellectually socialized in the field of religious studies, the balance of my work surely skews more towards the exegetical. 6  This characterization is rather a moving target; Mark has variously characterized Madhyamaka, also calling it, e.g., semantic non-dualism. Nevertheless, I find anti-realism an ideal-typical characterization that is helpful in scouting possible readings of Madhyamaka, and so will here emphasize that. 7  The reader may notice various Heideggerian elements to the reading here developed; see Arnold (2019) for further reference to Being and Time as warranting some of this. 8  Consider as analogous Locke’s thought that “person” is a chiefly forensic category. 5

262

D. Arnold

reading, to think Madhyamaka has no problem keeping the world in view. My interest in Madhyamaka’s unmistakably personalist contours thus reflects my inclination to resist reading Madhyamaka as any kind of anti-realism. I will begin by scouting the latter point in a more narrowly circumscribed way; regardless, then, of one’s conclusions regarding Madhyamaka’s affinities with Pudgalavāda, I hope in the next section at least to convey a sense of the different intuitions it’s reasonable to have regarding the possible truth of Madhyamaka claims.

2 On How It Can Be Ultimately True That There Is No Ultimate Truth For Madhyamaka as for most schools of Indian Buddhist philosophy, a lot of interpretive work hinges on the school’s understanding of the two truths: the conventional truth typical of everyday experience, and the ultimate truth taken by Buddhists as the unique purview of a Buddha. Many modern interpreters have followed Mark Siderits in affirming that what Madhyamaka affirms about this is that the ultimate truth is that there is no ultimate truth. Part of the appeal of Mark’s formulation, no doubt, is that it is redolent of the word-play typical of Mahāyāna Buddhism’s “Perfection of Wisdom” literature, which gleefully flirts with paradox by saying things like “a Tathāgata can be discerned by the absence of characteristics which is characteristic of him.”9 As traditionally understood by Buddhist commentators and philosophers, such statements only seem self-contradictory; the same is true for Mark’s signature formulation, as the statement’s two iterations of the expression “ultimate truth” have, in fact, different referents. Explaining as much in his excellent book Buddhism as Philosophy, Mark clarifies that we are thus to understand his statement as saying “the ultimate truth1 is that there is no ultimate truth2,” where that means: the “fact that must be grasped in order to attain full enlightenment” is that there is no “statement that corresponds to the ultimate nature of mind-independent reality” (2007: 202). On this reading, the point is thus to deny particularly that anything exists in a way that could make statements true. It is this focus on statements that makes Mark’s a basically semantic line of interpretation; there are no statements, on this view, backed by ultimately real truth-makers. Much in Madhyamaka literature is reasonably taken to warrant such an interpretation’s emphasis – most famously, no doubt, verse 29 of the Vigrahavyāvartanī, where Nāgārjuna denies that he is advancing any “thesis” (pratijñā). On a resolute reading, Nāgārjuna point is that Madhyamaka should not be understood as affirming anything at all about what is really the case. Mark’s semantic anti-realist reading is

 I’m here quoting the Vajracchedikā, but most any paragraph from any of the Prajñāpāramitā Sūtras likely includes something comparable. Compare Harrison’s translation (2006: 144): “it is by virtue of the featurelessness of his distinctive features that a Realized One can be seen.” 9

The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits…

263

akin to this, holding that insofar as Madhyamaka’s point is that no statement can correspond to mind-independent reality, it would court self-contradiction to make any statement even to that effect. Mark’s orienting alertness to such semantic issues recurrently shows up in my conversations with him, with Mark often protesting that my preferred interpretation of Madhyamaka as making transcendental arguments cannot be right, since that would be for Madhyamaka to have a “master argument” for its claims – an argument, that is, that itself has svabhāva.10 Despite, however, Nāgārjuna’s apparently straightforward disavowal of any “thesis,” that claim will admit of various interpretations. Note, first, that the Sanskrit word pratijñā has a philosophically technical sense, denoting such a “thesis” as figures in the kinds of arguments that later tradition would formalize as requiring warrant with reference to concomitant and contrapositive examples. Given Nāgārjuna’s principled concern to eschew all the philosophical approaches he encountered, it is reasonable to understand pratijñā as denoting (as we might even translate) a philosophically typical thesis – a thesis, I have emphasized before, of the sort thought to admit of a posteriori justification.11 The point in that case is that all the terms of any such procedure have their status as warrants only insofar as it is presupposed that they are essentially real; since Madhyamaka’s point just is to argue that nothing exists in that way, it stands to reason that Madhyamaka’s “thesis” could not be of the sort that requires such justification. Given that Madhyamaka thus aims to show that no statements can be backed by ultimately existent truth-makers, that can be shown only by way of a basically different kind of argument. Nonetheless, the argument can have ontological significance: there are no such truth-makers just because there are no such existents. The argument is not, moreover, that we just haven’t yet found any “ultimately real existents” – it is that nothing can coherently be thought to exist that way. As universal in scope, the truth of that claim must be presupposed by any exercise of our empirical faculties, which therefore cannot themselves be invoked as warrants. This is the gist of my view of Madhyamaka as making, against the Indian philosophical mainstream, basically transcendental arguments, generally to the effect that nothing at all makes sense except given the truth of Madhyamaka’s chief claim: viz., that all existents are empty of essence. That, I take it, is why it makes sense for Nāgārjuna to say (at Mūlamadhyamakakārikā 24.14) that “everything makes sense if emptiness makes sense; nothing makes sense if emptiness does not make sense.”12 Insofar as the claim thus concerns a transcendental condition of the possibility of all existents and practices, Madhyamaka must therefore eschew philosophical business as usual; that, I take it, is what Nāgārjuna means by disavowing any thesis.13  Cf. Siderits (2006: 20).  See Arnold (2005: 143–152) for further development of this line of thought. 12  Compare Siderits and Katsura (2015: 275–276). 13  That Tibet’s Gelukpa tradition similarly understands Nāgārjuna seems to me reflected in its view that Nāgārjuna does, in fact, have a distinctive thesis: viz., that “emptiness” means “dependent origination.” (See Matsumoto 1990: 33.) For further development of this train of thought, see Arnold (2005: 143, ff.), et passim. 10 11

264

D. Arnold

Also relevant here is that the Sanskrit word satya is ambiguous; though typically meaning “true” or “truth,” the word derives from the present participle sat, “existent,” and one commonly sees reference not only to saṃvṛti- and paramārtha-satya (conventional and ultimate “truth”), but also to saṃvṛti- and paramārtha-sat (conventional or ultimate existents).14 On my reading, the latter sense is paramount for the Abhidharma stream of thought that Madhyamaka constitutively resists. Abhidharma thus amounts to a project in basic ontology, with conventional truth attributed to statements referring to conventionally existent entities, and ultimately true claims being those that refer only to the ultimately existent ontological primitives, or “dharmas” in the Abhidharma literature. Insofar as it is Abhidharma ontology that Madhyamaka chiefly rejects, it stands to reason that the existential sense of sat/satya would be most basically at issue. In that case, however, the alternating sense of “ultimate truth” in Mark’s signature formulation will look different: what is ultimately true (the “fact that must be grasped in order to attain full enlightenment”) is not the semantic claim that no statement corresponds to ultimate reality; it is the ontological claim that there are no ontologically basic existents. If Madhyamaka’s cardinal claim is understood this way, the interpretive problems we face differ from those that most preoccupy Mark; there is, e.g., no contradiction in thinking it ultimately true that there are no ultimately real existents. I thus take it as not obviously right that Madhyamaka is “anti-realist” in regard, at least, to the truth of its own claims. Insofar as its proponents chiefly target accounts on which the truth of claims requires backing by any kind of privileged “truth-maker” – something, e.g., that is what it is independently of any context of inquiry  – Madhyamaka is best understood not as jettisoning truth, but only that understanding thereof. This point is integral to the sense Madhyamaka makes as a “middle way” that eschews not just reification, but also nihilism; the claim that no existent has an essence (svabhāva) means not that there are therefore no existents, but that a particular criterion of existence makes no sense. While Mark joins me in thus rejecting readings of Madhyamaka as embracing nihilism,15 it seems he would nevertheless resist the way I have just put that; he has said that “Madhyamaka accepts the svabhāva criterion of ultimate reality – something is ultimately real only if it bears its nature intrinsically – and also claims to show that in fact nothing bears an intrinsic nature” (2016: 132). For Mark, it seems, Madhyamaka thus accepts standard accounts of truth-makers, arguing only that because there are no such things, no statements can be made true thereby. Indeed, Mark has said in conversation that a realist conception of truth inevitably involves a finally causal explication of the truth-making relation.16 My guiding thought, in contrast, is that Madhyamaka means to challenge just that presumption. I take it that Madhyamaka claims to say something really true, and it thus seems clear their

 Cf. Kapstein (2001: 211, ff.).  See, e.g., MacDonald (2009, 2015). 16  See Arnold (2006: 15note13). 14 15

The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits…

265

arguments against ultimately existent truth-makers are best understood as targeting that account of truth – hence, my thought that Madhyamaka rejects not realism, per se, but rather a philosophically technical understanding thereof. Our divergent intuitions, however, may not make much difference here. Mark allows, for example, that the basically deflationary account of truth he finally attributes to Madhyamaka amounts to “a kind of principled endorsement of common-­ sense realism with respect to truth” (2003: 185) – and that is much as I would say. We might, to that extent, disagree only over the terminological question of whether “common-sense realists” should cede the term realism to foundationalists. My view in this regard is that insofar as Madhyamaka cogently shows the incoherence of foundationalism, Madhyamaka itself has the better claim to saying anything about “reality.” If, in other words, it is shown that conventionally real is all the more “real” anything can be, it seems reasonable to conclude that Madhyamaka’s recuperation of conventional truth represents the only sort of realism worth the name. To that extent, there is no reason to think Madhyamaka has any special explaining to do when it comes to the possible truth of its own claims. Indeed, Madhyamaka’s point is arguably that a “realist” conception of truth makes sense only in the context of conventionally real discursive transactions – and that is just what is denied by anyone who thinks we require a philosophically technical account of truth-makers. Why, then, should it be the latter whom we call “realists”?17

3 On How to Understand the Conventional: Candrakīrti’s Vestigial Personalism While it’s thus unclear to me how significantly we really differ over “realism,” I turn now to a thought that Mark, like many scholars of Madhyamaka, surely resists. Having suggested that Madhyamaka chiefly rejects Abhidharma’s idea of conventional truth as finally superseded by ultimate truth, I now want to explore the question of what we ought to think most salient about conventional truth. Here I want to argue that it makes a difference whether conventional truth is characterized chiefly in terms of semantic considerations, or instead as involving a personal level of description. Buddhists have traditionally emphasized in this regard that conventional truth is preeminently conceptual  – a point perhaps most often implied by the tradition’s emphatically valorizing a Buddha’s gnosis as constitutively non-conceptual. There are several reasons why this is as Buddhists would affirm. For one thing, the paradigm case of a “conceptual construct,” for Buddhists, is the self we all habitually

 I take it Jay Garfield (2006: 5) agrees: “to call a position according to which the reality of everything that, say, an idealist, or for that matter a materialist, denies is real, is real, ‘anti-realist’ seems to be at best misleading.” 17

266

D. Arnold

grasp; there is no real thing in the world corresponding to our desperate attachment, which therefore pertains only to an artifact of our conceptual activity. For a Buddha to apprehend things non-conceptually, then, is above all for a Buddha to apprehend things as selfless. The tradition’s emphasis on the nonconceptual character of a Buddha’s awareness also makes sense given that Buddhism is a basically gnostic tradition, in the sense that the summum bonum is taken as consisting in a transformative understanding of reality. To that extent, the tradition has a stake in emphasizing that the “understanding” at issue is unique; it cannot be anything like mere assent to a proposition that constitutes a Buddha’s understanding, which must consist, rather, in a complete and irreversible psychic transformation. The tradition’s wager is thus that our profound habituation to self-grasping runs far deeper than can be expressed or denied by anybody’s officially held philosophical views. Rightly understanding this or that Buddhist argument is, then, far from sufficient for extirpating the soteriological error at the heart of saṃsāra; however well one inferentially grasps the meaning of Buddhist doctrine, that is not at all the same as living, non-inferentially, out of a veritably perceptual awareness of the truth that there are no selves.18 To that extent, it makes sense that what most saliently characterizes the conventional truths of ordinary experience is that they are conceptually held – and surely it thus makes sense to hold, with Mark, that semantic concerns must be chiefly at issue for Buddhists. Nonetheless, my intuition is that it matters that the conventional be characterized as chiefly involving a personal level of description. To be sure, this keeps in play many of the same considerations; after all, persons, I take it, are among other things constitutively responsive to reasons, and to that extent this characterization of conventional truth surely comprises semantic considerations. A personal level of description, however, significantly picks out more than just conceptual or semantic considerations; persons not only exhibit responsiveness to reasons, they also, inescapably, act; the necessarily situated character of action makes it apt to say, with Heidegger, that the idea of a person just is that of a being-in-a-world.19 That we can aptly say this with Candrakīrti, too, seems to me evident in what is perhaps Candrakīrti’s most famous text: his treatment, at Madhyamakāvatāra 6.151–163, of the familiarly Buddhist idea that the reductionist analysis of persons is helpfully analogized to that of chariots.20 This passage is among the many places where Candrakīrti emphatically affirms statements that are unmistakably redolent of some “Personalist” schools of Indian Buddhist thought – and what he takes as most significant about such formulations is their entailment that constitutively situated actions are necessarily presupposed by any analysis of conventionally existent  While this, I think, is the reason that really motivates the tradition’s emphasis on a Buddha’s nonconceptual awareness, that emphasis nonetheless became philosophically significant; particularly in the epistemological tradition epitomized by Dharmakīrti, the question of how essentially nonconceptual perception can relate to any propositional knowledge loomed large. See, e.g., Dreyfus (1997), Arnold (2012a), Siderits et al. (2011). 19  My emphasis on action is informed by Moran (2001) and, especially, Kachru (2015). 20  Cf. Siderits (2007: 52–54). 18

The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits…

267

entities like chariots and persons. It is thus based on eminently personalist reasoning that Candrakīrti, in arguing that no reductionist account can coherently be taken as ultimately true, gives one of his most illuminating accounts of conventional truth. Now, it ought not, perhaps, to be surprising that Madhyamaka could think a personal level of description most basic for our ordinary epistemic practices. Given Madhyamaka arguments to the effect that ultimately real existents cannot coherently be conceived, it turns out that conventionally real existents are the only kind in play. To that extent, Madhyamaka can be said to effect a recuperation of conventional truth – a recuperation, we might also say, of what Wilfrid Sellars christened the “manifest image,” as against a “scientific image” that is reasonably thought analogous to the Buddhist tradition’s “ultimate truth.” And for Sellars, “there is an important sense in which the primary objects of the manifest image are persons.”21 It thus makes perfectly good sense to think Madhyamaka centrally concerned to argue that reductionism turns out to be unintelligible except relative to the personal level that Abhidharma was supposed to have reduced; what could be problematic in saying so? The main reason why both traditional and contemporary scholars of Madhyamaka are apt to balk is that Personalism (pudgalavāda) names a family of Buddhist schools that later tradition would almost unanimously judge veritably heretical. Buddhist schools that affirmed the reality of persons were widely taken to have smuggled “selves” back into a Buddhist account, surely anathema for any self-­ respecting Buddhist. Despite, however, the extent to which traditional opinion thus hardened against the Personalists, Candrakīrti’s famous discussion of chariots clearly reflects the perhaps unwitting influence of pudgalavāda on Madhyamaka.22 We can approach this by first scouting the lengthy critique of pudgalavāda famously ventured by Vasubandhu, an influential proponent of Abhidharma traditions and later of the idealist Yogācāra school of thought. In chapter 9 of his Abhidharmakośabhāṣyam, Vasubandhu critiques pudgalavāda as represented by the Vātsīputrīya school of thought. He starts by saying that to count as existent (sat),23 a person must exist in one of the only two ways Abhidharma philosophers thought possible: as substantially existent (dravya-sat), or else notionally existent (by way of a preliminary translation of prajñapti-sat).24 These respectively correspond to “ultimately existent” (paramārtha-sat) and “conventionally existent” (saṃvṛti-sat). Vasubandhu’s guiding thought is that these represent (as Matthew

 Sellars (1963: 9). On Buddhism’s two truths vis-à-vis Sellars, see the essays in Garfield (2019), Part I. 22  Though controversial, the idea of Pudgalavāda influence on Madhyamaka is not novel; see Vetter (1982, 1992); Kapstein (1987: 88–114); Walser (2005: 245–253) (et passim). It’s worth noting that it’s anachronistic to think Pudgalavāda obviously “unorthodox”; the Chinese pilgrim Hsuan-tsang calculated, at the time of his travels in India in the seventh century ce, that 25% of the subcontinent’s Buddhists were adherents of Personalist schools (Williams et al., 2012: 92). See, too, Cousins (1994). 23  The Vātsīputrīyas are thus represented as “accepting an existent person” (pudgalaṃ santam icchanti; Pradhan 1975: 460). 24  I will be returning to the sense of the word prajñapti in the last section. 21

268

D. Arnold

Kapstein says) “two mutually exclusive alternatives, and that between them there is no tertium quid.”25 Vasubandhu represents his personalist interlocutors, however, as refusing these alternatives and instead affirming a distinctive formulation: “A person shows up (prajñapyate) relative to (upādāya) currently appropriated (upātta) bodily aggregates.”26 Vasubandhu, in challenging this statement’s tenability as expressing an alternative to the two possibilities allowed by Abhidharma, first drills down on the word upādāya, here rendered as “relative to.” The word does indeed have that sense, but it will become important to recognize that it is literally the gerund form of the verbal root upa  +  ā  +  √dā, “to acquire” or “appropriate; literally, the Personalist’s formula thus says that it is “having appropriated” its constituent parts that a person shows up. Vasubandhu doesn’t note that, but he does consider two different ways the gerund might be understood: for a person to show up “relative to” the aggregates could be for it to show up insofar as the latter are taken as object of awareness (glossing upādāya with ālambya), or insofar as the construct person causally depends upon them (glossing upādāya as equivalent to pratītya).27 Vasubandhu thus effectively introduces the two basically different kinds of considerations in play in the previous section: a person’s showing up can be relative to the aggregates in a semantic or mind-dependent sense (insofar as the aggregates, as objects of awareness, are what is taken for a “person”); or in the ontological sense of causally depending on them, with Vasubandhu’s gloss of the latter option (pratītya) echoing the familiar term pratītya-samutpāda, “dependent” origination. Indeed, that upādāya commonly has the same sense as pratītya is reflected in Tibetan translations, which typically render both words the same way.28 Be that as it may, Vasubandhu’s point is just that the Personalists have the same problem either way: regardless of which kind of dependence is involved, only the aggregates are really referred to, since only these ultimately exist.29 Candrakīrti, we will see presently, embraces just the formulation that Vasubandhu thus took to distinguish Personalism; indeed, for Candrakīrti the same formulation  Kapstein (1987: 95) thus represents the Abhidharma idea as parallel to Derek Parfit’s view that reductionism and non-reductionism are mutually exclusive. For translations of Vasubandhu’s critique of pudgalavāda, see Kapstein (2001: 349–375), Duerlinger (2009). 26  ādhyātmikān upāttān vartamānān skandhān upādāya pudgalaḥ prajñapyate (Pradhan 1975: 460). I will explain my translation of prajñapyate as “shows up” in the last section of this essay. 27  Pradhan’s edition has a couple of mistakes here (1975: 461, lines 19–20), but the right readings are clear both from context and from Woghihara’s edition of Yaśomitra’s commentary (1936, Part II: 699, lines 30–33). 28  On the connection between the gerunds upādāya and pratītya, see Salvini (2011), Arnold (2005: 165–67) (with discussion of several texts in which Candrakīrti riffs on this connection). 29  Before taking leave of Vasubandhu, I would note that when he represents his Personalist interlocutors as trying to salvage their view following his initial interrogation of the word upādāya, he considers pudgalavāda’s characteristic preoccupation with fire and fuel as best exemplifying the kind of relation they have in mind – an example that is central, too, for Nāgārjuna and Candrakīrti, as has been noted by those who have argued for pudgalavādin influence on Madhyamaka (see note 22, above; see, too, Arnold 2012a: 224, ff.). 25

The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits…

269

expresses his considered opinion on what should be said about chariots in the wake of his Madhyamaka analysis. The analysis in question begins at Madhyamakāvatāra 6.151, where Candrakīrti enumerates the seven possibilities that he takes to exhaust the ways in which chariots could be related to their parts.30 The arguments of the ensuing verses show that none of these can coherently be maintained. As so often in Madhyamaka texts, that conclusion seems veritably intended to elicit an objection to the effect that nihilism is therefore the inescapable conclusion  – an objection, recurrently entertained in Madhyamaka texts, that almost invariably affords occasion to explain precisely how Madhyamaka avoids nihilism. Here is how Candrakīrti imagines an interlocutor leveling the charge he thus invites: If all this is as you say  – if a chariot, when sought according to the seven-fold method you’ve expressed, turns out not to exist – then, there being no chariots, ordinary usage and reference involving ‘chariots’ would be annihilated. We regularly observe, however, that people say such things as ‘bring the chariot,’ ‘buy a chariot,’ and ‘prepare the chariot.’ Since, therefore, they are well-known to everyone, things like chariots do exist.31

If nothing is ultimately real, this objection goes, then nothing about our ordinary discursive practices could make sense. Vasubandhu, we saw, arguably shares this intuition, and thus held that what’s problematic for Personalists is that there is, in fact, an ultimately existent referent available to ground their reference to “persons”; hence, the problem is that however one understands the Personalists’ characteristic claim, it is the aggregates that are really referred to, with nothing left over for “person” to refer to.32 Candrakīrti’s statement of the Personalists’ signature formulation comes in responding to the foregoing objection. Introducing the verse that echoes the Personalists, his commentary first turns the tables on his interlocutor: This is a problem only for you. A chariot does not make sense when sought according to the above-described seven-fold method, and yet you remain intent on rigorously demonstrating its existence. Given that you cannot affirm any way of demonstrating this other than these seven, how can you make sense of ordinary usages like ‘bring a chariot’? This is not a problem for us.

The intelligibility of our commonsense epistemic intuitions is imperiled, then, not by Candrakīrti’s analysis, but precisely by the foundationalist’s demand for backing by metaphysically unimpeachable existents (by, that is, things that have svabhāva).

 “A chariot, for example, is not admitted as distinct from its parts, nor as indistinct, nor as possessing them. It is not in its parts, nor are the parts in it; it is not a mere collection of them, either, nor is it their arrangement.” (I have translated Candrakīrti’s Sanskrit verses from Li (2015); I have translated passages from his commentary, still extant only in Tibetan translation, from La Vallée Poussin (1970a).) 31  Translated from La Vallée Poussin (1970a: 276.12, ff). 32  Earlier in the same chapter of the Madhyamakāvatāra (at 6.126), Candrakīrti represents Abhidharma philosophers as saying just that: “Some of our fellow Buddhists say that just because there is no substantiating a self that is distinct from the aggregates, the aggregates themselves are the object of the notion of self. Some such Buddhists affirm all five aggregates as the basis of the notion of self, others just one (i.e., thought).” 30

270

D. Arnold

Given, however, that there are no such existents, the manifest fact that ordinary discourse goes on can only mean that persistence in demanding such backing makes no sense. It’s not Candrakīrti’s analysis that spells trouble, but the foundationalist’s criterion of the “real”; this amounts to the point that conventionally real is all the more real anything can be.33 Now exemplifying the very turn-of-phrase that Vasubandhu took to typify Buddhist Personalism, Candrakīrti explains at Madhyamakāvatāra 6.158 why Madhyamaka doesn’t face the foundationalist’s problem: Even though it is not made intelligible (either ultimately or ordinarily) in any of the foregoing seven ways, a chariot – in ordinary terms alone, not subject to rigorous analysis – shows up relative to its proper parts.

Like Vasubandhu’s Personalists, Candrakīrti thus affirms that the right thing to say about how chariots exist is as “showing up relative to” (upādāya prajñapyate) something. Commenting on the verse, Candrakīrti presently introduces a term that can be supplied as the most general accusative for this expression, standing in for whatever it is relative to which any existent shows up. Thus, the Personalists affirmed that persons show up relative to the impersonal aggregates (skandha) of Abhidharma analysis, while Candrakīrti’s verse says a chariot shows up relative to “its proper parts” (svāṅgāni); a more general formulation of the principle at issue results from substituting for both accusatives the word upādāna, denoting anything “taken up” or “appropriated.” The word upādāna is familiar to students of Buddhism as naming the ninth link in the 12-fold chain of dependent origination; there, “appropriation” (upādāna), itself caused by “desire” (tṛṣṇā), in turn gives rise to “being” (bhava), which in turn causes birth, etc. The centrality of the “appropriation” metaphor is reflected in the tradition’s common reference to the aggregates particularly as the upādāna-­ skandhas, so called because it is the aggregates that are “taken up” by habituated self-grasping. (The same usage is reflected in Vasubandhu’s statement of the Personalists’ view, which, recall, says persons show up relative particularly to those aggregates that are currently “appropriated,” upātta  – another form of the same word.) It thus stands to reason that Madhyamaka texts would regularly take upādāna as shorthand for all the impersonal categories said in Abhidharma literature to be the only ultimately real existents.

 Candrakīrti’s argument here has affinities with Mark Johnston’s variously advanced case for the reality of the “manifest” world; Johnston similarly argues that “observing that the facts of personal identity do not involve superlative entities is not itself a criticism of the practices organized around identity. It would be a criticism only if such practices had to limn the metaphysical joints in order to be justified. But this is [an] impossibly strong condition on justification, [a] condition that would produce an all too automatic victory over ordinary life” (Johnston 1992: 618; emphasis original). Johnston’s argument is redolent of “sceptical solutions” to philosophical problems, as characterized by his teacher Saul Kripke: “A sceptical solution of a sceptical philosophical problem begins … by conceding that the sceptic’s negative assertions are unanswerable. Nevertheless our ordinary practice or belief is justified because … it need not require the justification the sceptic has shown to be untenable” (Kripke 1982: 66–67). 33

The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits…

271

Now, however, the gerund upādāya becomes newly salient; for like upādāna, this is another form of the verbal root upa + ā + √dā. The relation is immediately clear in the Sanskrit: ordinary existents show up upādānam upādāya. What’s most significant for Candrakīrti is that this makes it natural to emphasize that the formulation at issue implies reference to a whole situation or event of “taking up” or “appropriating.” As the Sanskrit grammatical tradition held in theorizing the idea that language chiefly tracks actions, reference to such an act entails at least implicit reference to all the kārakas – all the kinds of things, denoted by the Sanskrit case-­ endings that relate nouns to verbs, that are instrumental in bringing about an action. Among other things, that means the statement necessarily implies some subject of the act; there can be no appropriating without an “appropriator” (upādātṛ).34 Developing this point, Candrakīrti’s next verse (Madhyamakāvatāra 6.159) indicates the kinds of things one is therefore committed to admitting in virtue of affirming that anything like a chariot or a person shows up: “The common usage is that this very chariot, the subject of the verb (kartṛ), is at the same time a whole that possesses parts, also familiar to all as being the appropriator (upādātṛ).” Strikingly, Candrakīrti thus affirms just the kinds of abstractions typically refuted by Buddhists, who tend to have no truck with anything like the idea of “wholes” somehow existing over and above their parts. Indeed, mereological reductionism regarding such items is virtually a bedrock principle in Buddhist philosophy, which surely makes Candrakīrti’s apparent flaunting of this contentious.35 But Candrakīrti’s point in doing so is not, of course, to affirm any such things as ultimately existent; indeed, quite the converse. He agrees that “wholes” and “appropriators” and the like will indeed admit of reductionist analysis, and that such items therefore count as conventionally existent. His point is just that the basic constituents to which Abhidharma would reduce such things are not finally any more real than the conventional existents they were supposed to explain. The reason for this, harking back to the problem scouted in Mark’s “Is Reductionism Expressible?”, is that Abhidharma’s categories are themselves intelligible only relative to the very things they were posited to explain. Candrakīrti’s comment on verse 6.159 says as much: Insofar as they mistakenly understand the meaning of scripture, some mistakenly explain everyone’s settled convention like this: “Only collections of parts exist, but wholes do not exist in any way at all, since they are not apprehended over and above the parts. Likewise, only parts exist, but not part-possessors; only actions exist, but not actors; only what is appropriated exists, not the appropriators thereof – and this because in each case the latter is not apprehended over and above the former.” That reasoning entails, however, that parts themselves do not exist, either.36

 On the philosophical significance of the grammarians’ kāraka analysis, see Matilal (1990: 40–48).  As Mark says in this regard, “The Buddhist view of persons is based on … mereological reductionism, the view that wholes are reducible to their parts.” (2007: 105, et passim) 36  Translated from La Vallée Poussin (1970a: 278, lines 9–18). 34 35

272

D. Arnold

Candrakīrti’s point regarding Abhidharma’s impersonal categories, then, is just that none of them makes sense except in the context of some “taking” of things. That is, the very idea of ultimately real existents is incoherent just insofar as any candidates for that status turn out to be intelligible only relative to what is conventionally true – and, I’m suggesting, that most basically means relative to a level of description involving the actions of persons. To that extent, the conclusion to be drawn is aptly expressed as an exhortation (made in the last quarter of verse 6.159) that veritably epitomizes Candrakīrti’s understanding of Madhyamaka: “Do not ruin the conventional that is familiar to everyone!” To persist in thinking our ordinary epistemic practices as good as non-­ existent if they lack metaphysical grounding is just to render them unintelligible. The truth of the matter is that no account of ultimately existent truth-makers can be made coherent – and insofar as anything proposed as existing that way turns out to depend for its intelligibility on those very practices, it cannot be thought a problem that “conventionally” is all the more real anything can be. What’s problematic is just the idea that there must be something more “real” than the manifest world of ordinary experience. Accepting as much means “settling” for a lesser sort of reality only in comparison with the kind we habitually wish were existent. Here, then, we get at the sense this makes as a Buddhist position. What must be understood is that the kind of reality we most basically crave for our selves is incoherent; insofar as we yearn for our personalities to be somehow completely present, nothing could make sense as satisfying that desire. Regarding the sense this reading of Madhyamaka as elaborating Buddhist commitments, I would thus emphasize not only that a personalist reading of Madhyamaka is not inconsistent with the tradition’s cardinal “without-self doctrine” (anātmavāda), but indeed that it counts as elaborating that. It is, then, right to say, with the Pudgalavādins, that person denotes an idea basically different from that of self. While I am not, then, making any claim to the effect that the historical thinkers Nāgārjuna and/or Candrakīrti were card-carrying adherents of any avowedly Personalist school of thought, it’s here worth briefly considering what historical Pudgalavādins might have held. Madhyamaka would surely have a stake in rejecting Personalism if that were the view that persons must be reckoned as ultimately existent entities on the same level as Abhidharma’s ontological primitives. What historically pudgalavādin thinkers are commonly taken to have held, though, is just that the status of person is “inexpressible” (avaktavya). Despite the whiff of mystification this might suggest, Amber Carpenter (2015) has cogently argued that the Personalists likely meant only that the status of persons cannot be expressed in the existential terms allowed in Abhidharma literature. If that’s right, then the Personalists thus share with Madhyamaka a guiding concern to reject Abhidharma’s criteria of existence. In this case, the point is that while of course persons no more make sense as “substantially existent” (dravyasat) than selves do, the “notionally existent” (prajñaptisat)

The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits…

273

alternative is nonetheless inadequate to the distinctive way in which reference to “persons” is ineliminable.37 These alternatives can be thought commonly problematic insofar as both presuppose that it could only be as present or existent (sat) that persons make sense. Much as the early Heidegger argued in terms of Dasein, the rival intuition is that being a person cannot be thought to consist in being any kind of thing at all – neither as the enduring substance of Brahmanical ātmavādins, nor as a mereological sum that could be completely present. In this regard, I suggest that it matters that the grammatical tradition’s theorization of action figures centrally in Candrakīrti’s argument. The reason conventional truth cannot finally be explained is just that any explanation will be intelligible only relative to some taking of things – and as thus entailing (inter alia) a taker, a situated act is always already presupposed by any explanation. We can thus say that Brahmanical ātmavādins and Abhidharma philosophers alike posit ontologies in which something worth the name upādāna alone is real – a self in the former case, impersonal dharmas in the latter. Either way, that is to say that upādāna effectively functions as autonomously intelligible substance. Against that, Candrakīrti argues that no such thing is intelligible apart from a situated act of “appropriating” (upādāna), which entails reference to everything conventionally thought necessary in realizing any act. Emphasizing that description of his argument, we can say of the conventional truth that he shows to be ineliminable both that it is constitutively personal, and – what amounts to the same thing – that it is constitutively centered not on substances but on action. It is, then, just because persons are thus to be individuated as actors that no account of them as wholly existent can capture all there is to being a person.38 All of this means, again, that there is no problem with Madhyamaka’s thinking it true that there are thus no ultimately real existents. On a personalist reading, the truth-claim at issue is that a certain explanatory agenda vis-à-vis persons cannot, in principle, be realized. That makes sense, moreover, as a claim that cannot be made true by (e.g.) anything “substantially existent”; the truth that is presupposed by any explanation at all just is that all existents are empty. And that is, I’ve been  Chief among the Personalists’ concerns, Carpenter argues, is that Abhidharma’s reductionism makes sense of basic facts about personal identity by affirming that causal relations among aggregates occur in relatively discrete continua (santāna) – but these continua cannot themselves be individuated without reference to the “persons” they constitute. Thus, “The Pudgalavādin recognizes and avoids [the] circularity by acknowledging that in one central case, our convenience does not determine but rather tracks the different ways in which aggregates are related” (2015: 19). 38  This represents, I think, a good way to make sense of Jonardon Ganeri’s similar contention that Candrakīrti affirms a “performativist account of the language of self.” As Ganeri says of the “appropriation” Candrakīrti has emphasized, that should be understood “as an activity of laying claim to, not the making of an assertion of ownership. Grammatical form notwithstanding, the avowal or self-ascription of a mental state, ‘I have a pain’, is not a two-place relation between me and my pain…. When I say ‘I am in pain,’ I do not assert ownership of a particular painful experience; rather, I lay claim to the experience within a stream.” (Ganeri 2007: 202; see Thompson 2014: 356–66 for a similar reconstruction of Madhyamaka’s position.) Stanislaw Schayer (1929–30) has in this regard said that Nāgārjuna’s critique of motion is more generally meant as targeting action; see Arnold (2012b: 559−560) for discussion of this. 37

274

D. Arnold

emphasizing, to say that any explanation presupposes the eminently personal phenomenon of taking things some way or another, apart from which no existents count as essentially explaining anything at all.

4 On “Showing Up”: Making Things Out Versus Making Them Up I have not only allowed but emphasized that anything’s showing up inexorably implicates a situated “taking” of things – a perspective, e.g., from which any occurrence makes sense as the kind of thing it is. I have, however, resisted the idea that this situated taking represents an exhaustively semantic phenomenon, suggesting that it makes a difference if instead it be thought a chiefly personal idea. Nevertheless, it seems reasonable to think this doesn’t really move the ball very much; for all that we may have rendered the “semantic” more complicated by taking it as comprised by the personal, don’t we still just have the basically idealist notion that of course things only ever show up for us? And haven’t we, to that extent, jettisoned the idea of any mind-independent reality? Can I concede as much and still resist characterizations of Madhyamaka as anti-realist? To be sure, there is of course a sense in which Madhyamaka does, on my reading, jettison mind-independence as criterial for realism, though it’s important to appreciate the context for that. It is the tradition’s cardinal without-self doctrine (anātmavāda) that’s always finally at issue for Buddhist philosophers. Abhidharma elaborates this as meaning persons are not individuated by selves just because everything about being a person admits of reduction to essentially impersonal existents. I take it Madhyamaka’s guiding thought is that that doesn’t make for a consistently Buddhist view, since it leaves Abhidharma’s impersonal dharmas with just the kind of reality rightly denied of selves. While Abhidharma remains a useful first step in undermining self-grasping, it must thus be understood that on a consistently Buddhist view, dharmas, too, turn out not to withstand rigorous analysis. This is just the typically Mahāyāna thought that not only persons but also dharmas are ultimately “selfless.” It must, then, be kept in view that it’s regarding the Buddhist analysis of personality that Madhyamaka makes sense as rejecting Abhidharma’s criteria of existence; that is what’s at stake in arguing that reductionism is not coherently intelligible without reference to the very level of description it purports to explain. Insofar, then, as mind is itself an eminently personal phenomenon, Madhyamaka does indeed argue that mind-independence cannot be criterial for reality. That point looks rather different, however, in this context; it is particularly regarding the analysis of personality that personal considerations turn out to be ineliminable. So understood, this reading does not entail the subjectivist conclusion that any “convention” is as good as another; all the value and other judgments we regularly make in ordinary discourse remain as intelligible as ever. Indeed, Candrakīrti’s point is that it’s only

The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits…

275

if we suppose that all this requires unimpeachable metaphysical backing that practical reason becomes unintelligible. To that extent, the claim that reality is not essentially mind-independent is reasonably thought problematic only if one persists in demanding just the criterion Candrakīrti rejects. One can, of course, argue that his arguments aren’t convincing – but it would beg the question simply to assert that his position is problematic because the world goes missing. Indeed, it is arguably a necessary presupposition of conventionally true discourse that we take there to be a mind-independent reality – a presupposition we understand perfectly well, albeit that “perfectly well” cannot mean “in a philosophically technical way that can withstand ultimate analysis.” That there is a mind-independent reality must, of course, be recognized as just conventionally true – but insofar as the upshot of Madhyamaka just is that it’s only as conventionally true that anything at all is intelligible,39 it makes no sense to suppose we settle for less in thinking this “just” conventionally true. That the world is not essentially mind-independent, then, does not mean there is no difference at all between mind and world; indeed, to think otherwise would be precisely to allow that mind is essentially real  – that what we consider reality is essentially a product of conceptual activity. Surely, though, the most consistently Madhyamaka view is that conceptual activity is no more essentially real than Abhidharma’s ontological primitives. That any explanation is intelligible only relative to some taking of things cannot, therefore, mean there is nothing at all apart from the “taking.” To suppose that’s the case is to attribute to Madhyamaka a confusion nicely characterized by Clifford Geertz – the confusion, “endemic in the West since Plato at least, of the imagined with the imaginary, the fictional with the false, making things out with making things up” (1988: 140). Clearly, a consistently Madhyamaka view of the matter would be that how we “take” things is itself empty of any essence – and that is among other things to say it depends on what is there to be taken. The significance of this is too often overlooked, which makes tempting a veritably idealist understanding of Madhyamaka’s principal claim  – as, for example, David Burton’s contention that for Madhyamaka, “if the mind’s activity of conceptual construction did not occur, there would be no entities” (1999: 68). That reading is the basis for Burton’s argument that Madhyamaka entails nihilism, notwithstanding its claim to eschew that. Madhyamaka, on Burton’s reading, thus forgets that “[a]ll construction – no matter how complex – is finally based on an unconstructed reality” (1999: 92). Thus retaining precisely the idea that Madhyamaka rejects (i.e., that construction and the constructed must be essentially distinct), Burton thus explains why Madhyamaka entails nihilism: “If there is nothing unconstructed out of which and by whom/which conceptually constructed entities can be constructed, then it is impossible that these conceptually constructed entities themselves can

39

 Consider again Mūlamadhyamakakārikā 24.14 (note 12, above).

276

D. Arnold

exist.”40 But that is just to assert the very point that Madhyamaka rejects. Candrakīrti, we saw, argues precisely that the conventional makes sense only because it requires no metaphysical grounding in essentially distinct entities – and that’s a good thing, since his point is that none can be given. Against question-begging assertions that the world clearly goes missing for Madhyamaka, I want to conclude by sketching an additional case for thinking Madhyamaka is not aptly characterized as a kind of anti-realism – a case, indeed, for recognizing that Madhyamaka all along keeps the world in view. I’ve said, in this regard, that on a consistently Madhyamaka view, it cannot be that that mind and world are essentially distinct – which means not only that the intelligibility of the world is always relative to some taking of things, but also that how we take things itself depends on what is there to be taken. I now want to emphasize the significance of that by first making the philological point that I am warranted in translating prajñapyate as “shows up.” “Taking up what can be taken” (upādānam upādāya), I’ve translated Candrakīrti as saying, things like chariots “show up” (prajñapyate). Compared with standard translations of the expression at issue, that likely seems strange. Typically, prajñapyate and other variations on the underlying verbal form (including prajñapti) are translated with basically semantic terms; Candrakīrti’s Madhyamakāvatāra 6.158 is, then, usually understood as saying that chariots are imputed or designated in dependence on their parts.41 All modern translations of the Madhyamakāvatāra have been done from the Tibetan translation (which was until recently all that was extant), which renders the Sanskrit prajñapyate as ‘dogs pa, commonly defined in Tibetan dictionaries as meaning “impute, designate, label, name, refer to.”42 I suspect it’s partly owing to an unexamined tendency to translate variations on the word prajñapyate along such lines that many interpreters have been led to suppose that Madhyamaka loses all grip on the world. Philologically speaking, it’s not obvious that such translations are warranted. The nominal form prajñapti and the verbal form prajñapyate both derive from the causative stem of the verbal root pra-√jñā, “to know,” “understand,” etc.; the causative sense is thus “to show” or “point out.”43 According to Edgerton’s Buddhist Hybrid Sanskrit Dictionary (1953, s.v. prajñapti), the nominal form likewise denotes an act of “making known”; the Pali Text Society’s dictionary similarly informs that  Burton (1999: 4–5). Burton’s critique parallels that of Sthiramati, a proponent of Buddhism’s idealist Yogācāra school of thought. Sthiramati denounced Madhyamaka for pressing its analysis not just against all existents, but also against consciousness (vijñāna). The resultant Madhyamaka position – that “even consciousness, just like the objects thereof, exists only conventionally, not ultimately” – is untenable, Sthiramati thus thinks, since it “entails that the conventional doesn’t exist, either.” That, he says with another use of the familiar word upādāna, is because “the conventional does not make sense without something taken up” (na hi saṃvṛtir nirupādānā yujyate; translated from Buescher 2007: 42). 41  Huntington (1989: 176) translates “designated”; Padmakara (2002: 90) translates “imputed.” 42   So, the entry from the Rangjung Yeshe dictionary (http://rywiki.tsadra.org/index.php/ Main_Page). 43  So Apte (1992), s.v. prajñā. 40

The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits…

277

the Pāli equivalent paññatti primarily has the senses of “making known” and “manifestation.”44 It’s clear, then, that the lexically basic sense of these words is simply to denote events of making known or becoming manifest. It is therefore eminently reasonable (and arguably preferable) to understand Candrakīrti as saying at Madhyamakāvatāra 6.158 that notwithstanding his having shown the incoherence of any philosophically technical analysis of them, chariots nevertheless “show up” for us. Indeed, he specifically says in the following verse (6.159) that a chariot counts as a “whole” or an “appropriator” of its parts insofar as chariot is the subject (kartṛ) of the verb – that is, presumably, subject of the passive verb prajñapyate in 6.158. To that extent, Candrakīrti thus takes it that a chariot has, as it were, some “agency” in its becoming manifest.45 This suggests that apart from whatever conceptual “taking” is in play, something really shows up for us – something that thus provides (even if not essentially) some constraint on our conceptual activity. That this is as Candrakīrti intends becomes clear when Candrakīrti contrasts conventionally real existents with things that do not make sense as thus “showing up.” Among the contexts in which he does so is with regard to a verse in which Nāgārjuna himself riffs on the question of how to understand (in this case) “selves” relative to the impersonal constituents “appropriated” (upādāna) as such. Concisely stating the same point Candrakīrti made regarding chariots, Nāgārjuna thus says at Mūlamadhyamakakārikā 28.7 that “a self is not other than what is appropriated (upādāna), nor is it the same as what is appropriated, nor does it exist without anything appropriated. Even so, we are not entitled to the conclusion that it doesn’t exist.”46 Like Candrakīrti, Nāgārjuna thus argues that since it’s only insofar as a “self” shows up that anything can be individuated as parts thereof, it cannot coherently be thought that the parts alone are real; hence, we would be mistaken to conclude from his analysis of the relation that selves just do not exist. Again using the same personalist formulation already familiar from his Madhyamakāvatāra, Candrakīrti’s commentary thus unpacks the upshot of Nāgārjuna’s verse: How could what shows up relative to the aggregates (skandhān upādāya prajñapyate) “not exist”? Being non-existent, a barren woman’s son does not show up relative to the aggregates; but when there is something appropriated (upādāna), how could it make sense that there be no appropriator (upādātṛ)? It therefore makes no sense that it be non-existent, and so the conclusion that the self does not exist makes no sense, either.47

 Rhys Davids and Stede (1999), s.v. paññatti. (This entry also includes “description, designation, name, idea, notion, concept.”) 45  Matilal notes in this regard that the grammarian Pāṇini disregarded any distinction “between agents (kartṛ) which are sentient beings and those which are not” (1990: 42). 46  Compare Siderits and Katsura (2015: 322). This is among the verses Vetter (1982: 177) has specified as having a pudgalavādin resonance; Vetter (1992) also extensively considers Mūlamadhyamakakārikā chapter 27. 47  Translated from La Vallée Poussin (1970b: 578). 44

278

D. Arnold

Candrakīrti thus emphasizes, in effect, that what shows up for us is not just up to us; it depends, as well, on what is actually existent.48 Again, “actually” does not here mean essentially; whatever is actually existent will of course admit of innumerable descriptions, and nothing, therefore, could show up “as it is in itself.” Nevertheless, it makes sense to say its existence is actual, that it exists in fact; for while of course any fact of the matter can be only conventionally so, it is no less a fact for that. Indeed, only as conventionally real can anything be a fact. What Madhyamaka shows is just that mind-independent facts do not just give themselves to us as they are in themselves, as any encounter with them inexorably involves, in Geertz’s felicitous terms, some making things out; we remain, however, no less able to distinguish making things out from making them up. To that extent, one can persist in thinking that Madhyamaka leaves us with no purchase on the world – that Candrakīrti leaves us in a “dismal slough” of relativism,49 in a world that’s just “concepts” all the way down – only insofar as one retains the very presupposition Madhyamaka targets: that reasoning is arbitrary and facts unreal unless grounded in ultimately real existents. To the extent, however, that Madhyamaka cogently shows that criterion to be incoherent, it begs the question to invoke any such criterion in evaluating Madhyamaka. I’m thus inclined to resist Mark’s view that Madhyamaka accepts the svabhāva criterion, and that Madhyamaka’s point is therefore to argue only that nothing can exemplify it. That way of characterizing Madhyamaka’s principal move makes it tempting, I think, to suppose there remains some work for the svabhāva criterion to do – tempting to suppose, with Mark, that even given Madhyamaka’s view, realism still makes sense only given that criterion. Against that, I take it Madhyamaka’s point is to deny precisely that presupposition; to that extent, there seems to me little reason to worry that Madhyamaka asks us to relinquish any grip on reality. The difference between these orientations, then, is the difference between saying that existents fail to exemplify a criterion that nonetheless remains intelligible (given which, the status of the world immediately becomes dubious) – and saying, as I would, that because there is a world of existents, the failure of the same criterion to be exemplified therein can only mean the criterion is incoherent. Madhyamaka’s point, I thus take it, is precisely that the world can be as it is only because the svabhāva criterion cannot obtain. That, it seems to me, is just as Nāgārjuna argues in Mūlamadhyamakakārikā chapter 24, the thrust of which is to turn the tables on the interlocutor who thinks Madhyamaka tantamount to nihilism: “If all this is not

 Consider, too, Madhyamakāvatāra 6.25–26, where Candrakīrti says that what counts as true in ordinary discourse is “what everyone takes as apprehendable by any of the six senses (so long as these are unimpaired)”; other things  – everything, e.g., “variously imagined by non-Buddhists (whose heads are nodding with the sleep of ignorance),” and everything “imagined in such cases as magical illusions and mirages” – are “falsely constructed.” 49  See The Cowherds (2016: 43–54), et passim. 48

The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits…

279

empty, there is neither origination nor cessation, so it’s entailed for you that the four noble truths do not exist.”50 Taking my bearings from that move, I find it odd to think Madhyamaka accepts Abhidharma’s criterion of existent; for in that case, Madhyamaka could only conclude the world is not really as it is. To be sure, there are good Buddhist reasons to think that is just as Madhyamaka would want to show; after all, the Buddhist diagnosis of our existential predicament is that we are deeply habituated to mistakenly understanding ourselves and the world – of course a Buddhist would thus want to show that the world is not really as it “is”! Among the wagers of my reading of Madhyamaka, though, is that the converse intuition – my thought that Madhyamaka chiefly targets a misbegotten criterion of reality precisely so as to save the intelligibility of reality – also makes sense of the Buddhist diagnosis. On this reading, too, we are benightedly habituated to mistakenly understanding reality; it’s just that the mistake consists in our erroneously projecting onto existents a way of being that cannot possibly be instantiated. In particular, we habitually suppose that being a person – being, that is, a sentient organism whose suffering results from our finding ourselves with no choice but to act in a profoundly imperfect world that dwarfs our efforts – must really consist in being a self. To cling to that idea is to think there must be something that one finally is – that for all the flux and indeterminacy of our fleeting lives, ultimately we are lastingly significant existents. But in yearning for that status, we desire something that makes no sense: that our personality be completely present at once. Insofar as the most salient fact about persons is their finding themselves always already compelled to act – always already facing an indeterminate future that never stops calling for decision – there is nothing it could look like for the career of a person to be completely present. That there are persons, then, is not something that Buddhists ought to have a stake in rejecting. What we require release from is our attachment to thinking that being a person could matter only if it meant being a wholly present self. That attachment is pernicious because it is bound to be frustrated, since being such an existent makes no sense of the career of a person; we would, then, be well rid of the idea that personality means nothing unless it means being a self. But ridding ourselves of that would still leave a world that is, just as before, the only place wherein the careers of persons can play out. It seems to me that’s as good an account as any of what Nāgārjuna meant in saying there is, ultimately, no difference between nirvāṇa and saṃsāra. And I see no problem with saying it is ultimately true that that is so.

 Cf. Siderits and Katsura (2015: 279).

50

280

D. Arnold

References Apte, Vaman Shivaram. 1992. The Practical Sanskrit-English Dictionary. Revised and enlarged edition, ed. P.K. Gode and C.G. Karve. Kyoto: Rinsen. (Reprint; revised and enlarged edition first published in Poona, 1957.). Arnold, Dan. 2005. Buddhists, Brahmins, and Belief: Epistemology in South in South Asian Philosophy of Religion. New York: Columbia University Press. ———. 2006. On How it Can Be Ultimately True That There Is No Ultimate Truth: Thoughts on Mark Siderits’s Personal Identity and Buddhist Philosophy. APA Newsletter on Asian and Asian-American Philosophers and Philosophies 6 (1): 12–16. ———. 2012a. Brains, Buddhas, and Believing: The Problem of Intentionality in Classical Buddhist and Cognitive-Scientific Philosophy of Mind. New York: Columbia University Press. ———. 2012b. The Deceptive Simplicity of Nāgārjuna’s Arguments against Motion: Another Look at Mūlamadhyamakakārikā Chapter 2. Journal of Indian Philosophy 40/5 (2012): 553–591. ———. 2019. The Sense Madhyamaka Makes as a Buddhist Position: Reflections on a ‘Performativist Account of the Language of Self’. Journal of Indian Philosophy 47: 697–726. https://doi.org/10.1007/s10781-­019-­09390-­5. Baker, Lynne Rudder. 1987. Saving Belief: A Critique of Physicalism. Princeton: Princeton University Press. Buescher, Hartmut. 2007. Sthiramati’s Triṃśikāvijñaptibhāṣya: Critical Editions of the Sanskrit Text and Its Tibetan Translation. Vienna: Verlag der Österrichischen Akademie der Wissenschaften. Burton, David. 1999. Emptiness Appraised: A Critical Study of Nāgārjuna’s Philosophy. London: Curzon. Carpenter, Amber. 2015. Persons Keeping Their Karma Together: The Reasons for the Pudgalavāda in Early Buddhism. In The Moon Points Back, ed. Koji Tanaka et  al., 1–44. New York: Oxford University Press. Cousins, L.S. 1994. Person and Self. In Buddhism into the Year 2000: International Conference Proceedings, 15–31. Bangkok: Dhammakaya Foundation. The Cowherds. 2016. Moonpaths: Ethics and Emptiness. New York: Oxford University Press. Dreyfus, Georges. 1997. Recognizing Reality: Dharmakīrti’s Philosophy and its Tibetan Interpreters. Albany: SUNY Press. Duerlinger, James, trans. 2009. Vasubandhu’s Abhidharmakośa: The Critique of the Pudgalavādins’ Theory of Persons. In Buddhist Philosophy: Essential Readings, ed. William Edelglass and Jay Garfield, 286–296. New York: Oxford University Press. Edgerton. 1953. Buddhist Hybrid Sanskrit Grammar and Dictionary. Vol. 2: Dictionary. New Haven: Yale University Press. Ganeri, Janeri. 2007. The Concealed Art of the Soul: Theories of the Self and Practices of Truth in Indian Ethics and Epistemology. Oxford: Oxford University Press. Garfield, Jay. 2006. Reductionism and Fictionalism: Comments on Siderits’s Personal Identity and Buddhist Philosophy. APA Newsletter on Asian and Asian-American Philosophers and Philosophies 6 (1): 1–7. ———, ed. 2019. Wilfrid Sellars and Buddhist Philosophy. New York: Routledge. Geertz, Clifford. 1988. Works and Lives: The Anthropologist as Author. Stanford: Stanford University Press. Harrison, Paul. 2006. Vajracchedikā Prajñāpāramitā: A New English Translation of the Sanskrit Text Based on Two Manuscripts from Greater Gandhāra. In Buddhist Manuscripts in the Schøyen Collection, ed. Jens Braarvig, vol. III, 133–159. Leiden: Brill. Huntington, C.W., and Geshe Namgyal Wangchen. 1989. The Emptiness of Emptiness: An Introduction to Early Indian Mādhyamika. Honolulu: University of Hawaii Press. Johnston, Mark. 1992. Reasons and Reductionism. The Philosophical Review 101 (3): 589–618. Kachru, Sonam. 2015. Minds and Worlds: A Philosophical Commentary on the Twenty Verses of Vasubandhu. Ph.D. dissertation, University of Chicago.

The Real According to Madhyamaka, Or: Thoughts on Whether Mark Siderits…

281

Kapstein, Matthew. 1987. Self and Personal Identity in Indian Buddhist Scholasticism: A Philosophical Investigation. Ph.D. dissertation, Brown University. ———. 2001. Reason’s Traces: Identity and Interpretation in Indian & Tibetan Buddhist Thought. Boston: Wisdom Publications. Kripke, Saul. 1982. Wittgenstein on Rules and Private Language: An Elementary Exposition. Cambridge, MA: Harvard University Press. La Vallée Poussin, Louis de, ed. 1970a. Madhyamakāvatāra par Candrakīrti: Traduction tibétaine. Bibliotheca Buddhica 9. Osnabrück: Biblio Verlag. (Reprint.) ———, ed. 1970b. Mūlamadhyamakakārikās (Mādhyamikasūtras) de Nāgārjuna, avec la Prasannapadā Commentaire de Candrakīrti. Bibliotheca Buddhica. Vol. 4. Osnabrück: Biblio Verlag. (Reprint.) Li, Xuezhu, ed. 2015. Madhyamakāvatāra-kārikā Chapter 6. Journal of Indian Philosophy 43: 1–30. MacDonald, Anne. 2009. Knowing Nothing: Candrakīrti and Yogic Perception. In Yogic Perception, Meditation and Altered States of Consciousness, ed. Eli Franco, 133–168. Vienna: Verlag der Österrichischen Akademie der Wissenschaften. ———. 2015. The Quest for an English-speaking Nāgārjuna. Indo-Iranian Journal 58: 357–375. Matilal, B.K. 1990. The Word and the World: India’s Contribution to the Study of Language. Delhi: Oxford University Press. Matsumoto, Shirō. 1990. The Mādhyamika Philosophy of Tsong-kha-pa. Memoirs of the Research Department of the Toyo Bunko 48: 17–47. Moran, Richard. 2001. Authority and Estrangement: An Essay on Self-Knowledge. Princeton: Princeton University Press. Padmakara Translation Group. 2002. Introduction to the Middle Way: Chandrakirti’s Madhyamakavatara with Commentary by Jamgön Mipham. Boston: Shambhala. Pradhan, Prahlad. 1975. Abhidharmakośabhāṣyam of Vasubandhu. Patna: K. P. Jayaswal Research Institute. Rhys Davids, T.W., and William Stede, eds. 1999. The Pali Text Society’s Pali-English Dictionary. Oxford: Pali Text Society. Salvini, Mattia. 2011. Upādāyaprajñaptiḥ and the Meaning of Absolutives: Grammar and Syntax in the Interpretation of Madhyamaka. Journal of Indian Philosophy 39: 229–244. Schayer, Stanislaw. 1929–30. Feur und Brennstoff: Ein Kapitel aus dem Mādhyamika-Śāstra des Nāgārjuna mit der Vṛtti des Candrakīrti. Rocznik Orjentalistyczny 7: 26–52. Sellars, Wilfrid. 1963. Science, Perception and Reality. Atascadero: Ridgeview Publishing. Siderits, Mark. 2001. Buddhism and Techno-Physicalism: Is the Eightfold Path a Program? Philosophy East and West 51 (3): 307–314. ———. 2003. Personal Identity and Buddhist Philosophy: Empty Persons. Burlington: Ashgate. (Second Edition, 2015.). ———. 2006. Replies to Garfield, Taber, and Arnold. APA Newsletter on Asian and Asian-­ American Philosophers and Philosophies 6 (1): 16–21. ———. 2007. Buddhism as Philosophy: An Introduction. Indianapolis: Hackett. ———. 2009. Is Reductionism Expressible? In Pointing at the Moon: Buddhism, Logic, Analytic Philosophy, ed. Mario D’Amato et al., 57–69. New York: Oxford University Press. ———. 2016. Two Truths, the Inexpressible and Analysis: What (I think) Mādhyamikas Should Say. Indogaku Chibettogaku Kenkyū [*Journal of Indian and Tibetan Studies] 20: 131–139. Siderits, Mark and Shoryu Katsura, trans. 2015. Nāgārjuna’s Middle Way: The Mūlamadhyamakakārikā. Boston: Wisdom Publications. Siderits, Mark, Tom Tillemans, and Arindam Chakrabarti, eds. 2011. Apoha: Buddhist Nominalism and Human Cognition. New York: Columbia University Press. Thompson, Evan. 2014. Waking, Dreaming, Being. New York: Columbia University Press. Vetter, Tilmann. 1982. Zum Problem der Person in Nāgārjunas Mūla-Madhyamaka-Kārikās. In Offenbarung als Heilserfahrung im Christentum, Hinduismus und Buddhismus, ed. Walter Strolz and Shizuteru Ueda, 167–185. Freiberg: Herder.

282

D. Arnold

———. 1992. On the Authenticity of the Ratnāvalī. Asiatische Studien/Études Asiatiques XLVI/1: 492–506. Walser, Joseph. 2005. Nāgārjuna in Context: Mahāyāna Buddhism and Early Indian Culture. New York: Columbia University Press. Williams, Paul, with Anthony Tribe and Alexander Wynne. 2012. Buddhist Thought: A Complete Introduction to the Indian Tradition. New York: Routledge. Wogihara, Unrai, ed. 1936. Sphuṭārthā Abhidharmakośavyākhyā by Yaśomitra, Part II (IV-VIII Kośasthānam). Tokyo: The Publishing Association of Abhidharmakośavyākhyā.

Anti-Realism and Realism About the Past: A Present for Mark Siderits Arindam Chakrabarti

“If the present and the future exist dependent on the past, then present and future would be at the past time. If present and future do not exist there (at that time), then how would present and future exist dependent on that? …No existent whatsoever exists; how, then, will there be time?”—Nāgārjuna MMK 19. 1,2, 6 (translation by Siderits and Katsura)1

1.  Nyāya realists drew an important distinction between absence and non-existence. Perished particulars, such as Aristotle, emperor Ashoka, or the Bamiyan Buddha statues in Afghanistan are absent now but they are not non-existent in the sense of having become unreal or fictional. By dying, my grandmother did not become nonexistent like Snow White, though she suffered post-cessation absence. In order to be really dead, one could aptly remark, she has to be real. Can we therefore be realists about now deceased individuals or past event-entities or should we listen to Vasubandhu and adopt a subtle anti-realism about wholly past entities and events. Can both Michael Dummett and his death (in 2011) be deemed unreal now? Inspired by Buddhist philosophers from Nāgārjuna to Śāntarakṣita, Mark Siderits has been committed to a nuanced fictionalist version of anti-realism not only about universals, permanent selves, and the external world but even about consciousness. In this paper I try, first, to reconstruct such an anti-realism about the past using Vasubandhu’s critique of the omni-existentist (sarvāstivādin) early Buddhists, and then respond to it by bringing in Bhāsarvajña the Kashmir-Naiyāyika and Michael Dummett into dialogue with Vasubandhu and his realist Buddhist predecessors. 2.  All men are mortal. Michael Dummett was a man. So he was mortal. It sounds more idiomatic to say that he was a logician rather than that he is a logician because,  Nāgārjuna’s Middle Way, translation and commentary by Mark Siderits and Shoryu Katsura, Wisdom, Boston 2013. 1

A. Chakrabarti (*) University of Hawai’i, Department of Philosophy, Honolulu, HI, USA © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_14

283

284

A. Chakrabarti

he not only died in 2011, he is dead at present. In the early nineteen eighties, he supervised me and wrote extensive comments on my (still unpublished) paper on singular death-sentences, sentences of the form “M.D. is dead” (which is not the example I used in that paper which I originally wrote for him as part of my dissertation on negative existentials). The question, without entering into the metaphysical debate about survival or immortality of the soul, is whether the embodied person M.D. is now non-existent or just absent currrently but present at a different time? Can we existentially generalize from the death-sentence, “M.D. is now dead”, that there exists at least one person who is now dead? What do I mean by “realism about the past and future”? Based on Dummett’s conviction that the theory of meaning is the heart of metaphysics, a Dummettian formulation would be to say that realism about the past and future is the view that our statements about the past and future should be understood in terms of the conditions of their being either determinately true or determinately false independently of current evidence or our memories or expectations, tendencies or intentions, even if we have no currently available ways of verifying or justifying such statements or their negations. I am trying to defend a moderate realism about the past (and the future) without agreeing with any of the four kinds of omni-existentists, and fully respecting the common-sense view that what has ceased to exist is simply not there except as an identifiable absentee to a post-destruction absence. 3.  Absence, we understand, should not be confused with nonexistence. In order to be an absentee in my class, a student has to be really existing and present somewhere else. Harry Potter cannot be absent from my class, because he is nonexistent. We cannot count the present (2019) Tsar of Russia (unless a journalist ironically refers to Putin as a Tsar) as an absentee from a world-economy summit, because, though the adjective “present” is part of his defining description, he does not exist: he is fictional. Does that show, by the way, that the word “existent” is not a synonym of “present” (as Bede Rundle2 has recently argued)? We could put it this way: just as being present is not enough for really existing—witness the present Tsar of Russia— being absent is not enough for nonexistence. Genuine non-existence of the sort enjoyed by square circles and Sherlock Holmes takes much more than mere absence here and now. In fact, I wish to argue that it is not non-existence but existence which is necessary for being absent, in so far as only the real can be missing from a place at a time. Now the question is: after an object (which really existed in the past) ceases to exist, what does it suffer: absence or nonexistence? Is the perished past item—person, event, process or thing or phase—absent or nonexistent? Could the perished particular be totally absent and yet, because of its absence, somehow existent? That is ultimately the main question raised in this essay. Whether or not our use of names and descriptions can pick out utterly non-­ existent, merely dreamt-of, imaginary or future particulars, they would be of little  Why Is There Something Rather Than Nothing? By Bede Rundle, Oxford University Press, Oxford 2004. 2

Anti-Realism and Realism About the Past: A Present for Mark Siderits

285

use unless they were able to pick out absent particulars. For our singular terms to secure their real reference, we do not need to go around carrying the objects and people we speak of in a sack, so that we can present them before an audience every time we speak of them. All telling does not need to be a “show-and-tell” session. To translate a Sanskrit semantic idiom, every singular reference to a cow does not have to grab the cow by its horn. After a person dies, we do not routinely apologize in the obituary that we are unable to report exactly who it is that passed away because her name no longer has a Bedeutung! If “empty” means lacking any reference, then Arthur Prior was mistaken in calling the name of a dead-person a “now-empty” term, for my mother’s death does not rob her name, or any definite description denoting her (e.g.“Arindam’s mother”), of its reference. If “Arindam’s mother” is not an empty term, then it is not a now-empty term. Had it been now-empty then the sentence “Arindam’s mother is deceased” could not be now-true (even in the most un-free logic). Reference may or may not presuppose existence, but it could not presuppose actual living or presence of the referent concurrent with the use of the singular term with that reference. Among the absent entities that we can still speak of with the help of their (recalled or recorded) names must be people who died recently or long ago, buildings which have been burnt down, or cities which have been wiped out of the face of the earth, and animal-species which are extinct. While it is clear that such perished particulars can definitely be objects of linguistic reference, memorial and inferential knowledge, it is not clear at all what sort of existence they continue to enjoy even when they are extinct, and how a present use of a proper name can manage to pick out an entity entirely confined to the past. This question is not as silly a product of grammatical illusion as the question: “What is the kind of being that Pegasus must have if he is being depicted, imagined, or spoken of?” or “Where does the flame go when it goes out?” or “Does a dog dream when it is put to sleep?”. Neither does it have anything to do with the question of (the soul’s) survival of death. If the soul of my dead grandfather is still around, this soul, especially because it would have survived his death, would not be a perished particular, for to survive is not to perish. Hence the soul issue is not relevant to our subject-matter. If dying is not ceasing to exist then we are not speaking of the ontological status of the dead here. For the purposes of this paper, then, let us mean by “dying” simply ‘ceasing to exist’. If we assume, on the other hand, that “Mr.N is dead” simply tells us that the unfortunate Mr.N. has lost all ontological status, then whom this misfortune has befallen will be logically a much more interesting question than the ancient issue of why the post-death not-being is worse than the pre-birth not-being. Statements made now in the past tense about this deceased person would not be able to give us an inference ticket for a present-tensed or tenseless existential generalization. If, supposing the Mr. N had attained Buddhist Nirvana, from the statement: “Mr. N was cheerful”, under this cessation-without-­ remnant scenario, shall we be able to infer: “There exists an x such that x was cheerful”, without sounding as if some man who is still around has simply lost his cheer? 4.  These questions draw us into deep puzzles regarding existence and tense. From very ancient times—at least from the beginning of the Christian era—the question

286

A. Chakrabarti

whether events in all three times are equally real had divided (Vaibhāşika) Buddhist metaphysicians in India into different camps. Even though staunchly adhering to the Buddha’s basic doctrine of impermanence of all things, some early Buddhist realists called themselves “omni-existentists” (sarvāstivādin) precisely because they believed that the impermanent event-factors (dharma-s) of past and future are as real as the ephemeral process-things of the present. If in the past, depending upon its own causal conditions, a process arose and soon after generating its effect, perished, then as situated in that past time, it is (this “is” being a tenseless marker of being as such) real. Thus the past as past, the present as present, and the future as future, are all equally real. This should not be mistaken either for an eternalism (“what is, is there for all times”) about all real events, or the view that all the past events are “still(= till now)” continuing. By calling the short-lived past processes “real” these realist Buddhists could not have meant that they persist at present, for that would flatly go against the theory of universal impermanence. Thus Vasubandhu’s commentary presents the Omni-existentist remark defensively: “Who said that the past and the future exist like the present? (Common sense may ask—) How, otherwise, do they exist? (To this, we answer—) Well, as past, and as future!” (ADKB3 V, 26–27). Thus they could stick to the doctrine of universal flux, and even momentariness of all beings while accepting the reality of the past and the future, precisely because they rejected what is currently called “Presentism”. Presentism is the view that to be real is to be existing now. Indeed, come to think of it, Presentism could be summarized as an ontological threat to past and future entities of the form “Either now or never”! Eternalism, on the other hand, could adopt the slogan: “To be, is to be for ever”. Early Buddhists could not have accepted Presentism because it privileges a certain speaker-relative ego-centricity, as if there is a special ontological advantage to being broadly contemporaneous with this token utterance of “I” or “Now”. And of course, the very spirit of “dependent arising” (pratītya-samutpāda)—the cornerstone of Buddhist metaphysics—goes against Eternalism. Three arguments are given by Early Buddhist believers in the existence of event-­ things –dharma-s (“quanta” –according to Andrei Paribok of St, Petersburgh University Russia) in all three parts of time, for the position that the past (and the future—though in this paper from now on I shall only discuss past event-things) entities—however impermanent—do exist after all, as past entities (ADKB, Vth Compendium Section, verse 25). The first one is from Karmic fruition (phalāt). Since our actions have consequences in the long run, the current consequences we are suffering and enjoying must be connected by causal relation to the past good and bad actions. Since we cannot find the causes of current suffering in any current reality, and they could not be uncaused, these Vaibhasika philosophers postulate the existence of past events and actions as causally responsible, hence real.  Abhidharmakośabhāṣyam (ADKB) by Vasubandhu (Author), − 4 Volume Set (English, French and Sanskrit Edition) (Sanskrit), L.  De La Vallee Poussin (French Translator) (Author), Leo M. Pruden (English Translator) (Author) Asian Humanities Press 1991. 3

Anti-Realism and Realism About the Past: A Present for Mark Siderits

287

Then there is the hermeneutic argument, appealing to what the Buddha said and how it must be interpreted. It runs as follows: According to the authentic records of his discourses, the Buddha repeatedly asserts that he has heard of noble spiritual aspirants (ārya-śrāvaka) having to cultivate indifference to, or non-reliance (anapekṣā) on, those visible tangible forms which have gone by, telling themselves not to get attached to the ceased past with nostalgic investment (and not to look forward to the not-yet-come future with too much yearning). Had past events been as unreal as impossible fictions like a donkey’s horns then the Buddha would not waste his breath cautioning people against getting attached to them. Past events could not thus be instructed to be objects of dispassion and detachment unless they were possible objects of passion and attachment. Only what is real can be objects of passionate attachment and hence objects one should practice detachment from. Therefore, past events, even when they have ceased, must be as real as present events, hence worth cautioning against. A third argument is given from the need to have an intentional object of our memory or knowledge of the past. The spirit of this argument is rather straightforwardly realist. “A real cognition arises, only when there is a real object. Had the past and future not been there, then our remembrance or mnemic consciousness of the past, or our anticipation of the future would be cognition-without-an-object. But without an object there would be no genuine cognition (sad-viṣayatvāt)”. The past particulars, though perished, are shown to be read on the basis of these above arguments. 5.  Four types of omni-existentists are listed by Vasubandhu. Their difference of opinion centers the question which naturally arises when their common tenet is asserted: “All entities in the past, present, and future exist”. The question is “What, then, accounts for the experienced difference in tense, between the past, the present and the future?” Those who take this felt tense-difference to be a difference in mode of being (bhāva) are (1) Way of Being—Differentialists (e.g. Bhadanta Dharmatrāta). Those who take it to be a difference in salience and non-salience of defining criteria (lakṣaṇa) are (2) Criterial Differentialists (e.g. Bhadanta Ghosaka). Those who take tense distinctions to be a difference in relative position or situation (avasthā) are Situation-Differentialists (Bhadanta Vasumitra). Those who take such distinctions to be a matter of reciprocal relations of the Otherwise-to-Otherwise(anyathā-­ anyathā) are Mutual Relation Differentialists (Bhadanta Buddhadeva). When the present pot perishes and becomes (is recalled as) a past pot, the substance of the pot does not change, but only its mode of being or a property alters. It sheds its currency and assumes pastness, just as the same lump of gold could shed its bangle shape (configuration) or mode and adopt a necklace mode (configuration) of being. Futurity, currency, and pastness are successive passing modes of being which makes no difference to the core of the changing entity itself. Change is a tense-modal alteration of the same-staying entity. This is the view of the first group of philosophers who have been called Mode-Differentialists (bhāvānyathā-vādin).

288

A. Chakrabarti

As could be expected, this view is rejected by Vasubandhu because it appears to be indistinguishable from a Sāmkhya-type Eternalism with respect to the essential substance of the entity in question. Even before it originates, an entity possessing futurity would then be real, and even after its cessation, it would have to be there to bear the mode of being called “pastness”. How would that be consistent with genuine coming into being and passing away? Everything would be ontologically beginningless and endless, which would go against the Buddha’s fundamental wisdom! The second view considers a thing to be inseparable from all its properties (somewhat like David Armstrong’s recent position that each thing has all its properties necessarily). So throughout the career of an entity, it possesses all its characteristics which make it what it is. Yet, not all these properties are saliently available or attained at the same time during its career. Some times some properties are attained while others are lapsed or unattained. At different times different properties are manifest while others are submerged or unmanifest. Thus, according to the Criterial Differentialists, in each mutable entity all three tenses—pastness, presentness and futureness are all available at the same time but only one of these features is saliently attained at any one time. When its pastness is fully manifest, although it has futurity and presentness, these criterial features are not recognizably salient in it at that time, and as a result, it is not known as or called “present” or “future”. What is remarkable is that Kamalaśīla’s refutation of this view anticipates McTaggart’s incoherence objection to the A-series: The incompatible properties of pastness and presentness and futureness could not be all available in one entity., Kamalaśīla insists, because that would lead to an intolerable “cross-­ mixture”(samkara). Yet, Criterial Differentialists (lakṣaṇānyathātva-vādins) who take the event-things and their tense-properties to be indistinguishable end up intermixing these very incompatible properties by insisting that what is future is also present and past for the same process-thing has all three tense-marking temporal properties. How could pastness and futureness, being incompatible properties, be co-located in one process? The third and fourth views seem to grapple with this problem by offering a pointof-view-relative solution. The third view is illustrated by the computing example of the same marble placed in the slot of primary unit is called one, placed in the slot of hundred is called a hundred etc. The fourth view is explained in terms of the relational nature of the incompatible properties of motherhood and daughterhood. A cannot be at once the mother and daughter of B. But B, who is a daughter to A, could be a mother to C. Just as the same woman (present) becomes a mother with respect to her child (future), but herself as a daughter bestows motherhood on an older woman (past), the same present event, while being “earlier than” the future event, renders a previous event its own “anterior” or past. The attempt here seems to be to reduce the A-series of past, present and future to the B-series of earlier and later. Against this fourth view once again the serious charge of tense-mixing (addhvā-­ sāmkarya) is brought. If the distinction between past, present and future times and the events happening at these different times has to be admitted, as the omni-­ existentist professes to do, then these tense-segments cannot be conflated with one

Anti-Realism and Realism About the Past: A Present for Mark Siderits

289

another. Yet, the fourth view, trying to do justice to the mutually relative nature of the (what nowadays we call A-series-) epithet “past” and “future” etc., had to assert that in the past, the present moment was the future, so, what is present (now) is future-in-the-past. Why this “intermixture” (sāmkarya) is so intolerable would not be quite so clear until one unearths Vasubandhu’s presuppositions that the only way to maintain realism about time is to make “past” and “future” absolutely incompatible like “past” and “not past” whereas the same event seems to assume both these characterizations, whence the suspicion that pastness and futurity must be observer/ speaker/mind-dependent rather than objective. Though all of them were struggling with the puzzles of tense and existence, it is not easy to figure out from these cryptic descriptions which of these views were trying to focus on the fixed B-series of earlier and later, and which ones were focusing on the dynamic A-series of a perpetually “growing block” of ceased events, a moving now, and an unrolling coil of yet to come. In what seems to be the most pellucid contemporary exposition of Vasubandhu’s anti-realism concerning the three times, Jonathan Gold formulates the problem as follows: If a thing exists in the past, present, and future, that seems equivalent to saying that it always exists, which is again equivalent to saying that it is eternal. It would appear to be straightforward, then, that based upon the doctrine that conditioned things are impermanent, Buddhists should accept the unreality or nonexistence, of past and future objects (Gold 2015, 25–584).

We see in these passages that Vasubandhu is reporting an ongoing dispute where one party takes the “is” ascribed to the past (“That smashed pot is past”) as a tenseless copula as in “There is a prior not-being of the lamp which is lit, there is a posterior not-being of the lamp after it is extinguished” where it would be useless quibbling to detect an inconsistency between “is” and “not-being”. “Is” (asti) here, the text (ADKB V 25–26) says, is not at all a verb, let alone a present-tense conjugated form of the verb “to be”. It is an immodifiable constant (nipāta) such as “again” or “and”. Just as “again” does not change its meaning and a repetition (which is what “again” signifies) does not become any less of a repetition, when we pass from “She does it again” to “She did it again”, or to “She will do it again”, “is” does not signify any fainter variety of existence when we pass from “This is now” to “That is past”. Thus we can say without any contradiction: “The past pot is a pot though it is not there (=does not exist) now”. The other party seems to appeal to the continuing causal efficacy of the past which we suffer and experience in the present and argues for the substantial existence of the perished events, in so far as, through their consequences, past actions ‘live in their subsequent fruits’. It looks as if the first sense of existence ascribed to perished particulars, more like the Quinean ‘being the value of a variable’, is captured by the tenseless existential quantifier, whereas the second sense makes “exists” a substantial  Paving the Great Way, by Jonathan C. Gold, Columbia University Press, New York 2015.

4

290

A. Chakrabarti

property earned by virtue of having brought about real effects. But once earned, could this thick existence-as-effectiveness be lost? 6.  Now, apart from the existence which is captured by the existential quantifier, at the end of his classic essay The Thought: A logical inquiry, Frege speaks of a thicker existence which is standardly translated as reality or actuality (Wirklichkeit): The world of the real is a world in which this acts on that, changes it and again experiences reactions itself and is changed by them. All this is a process in time. We will hardly recognize what is timeless and unchangeable as real (Strawson 1967, 37).

While existence, captured by the existential quantifier, is not a first-level predicate at all, reality (Wirklichkeit) is a first-level property of those individuals which can participate in the temporal process of causal interaction with each other. Numbers, thoughts, value-ranges, the null-set, are all existent entities which lack actual reality, because they do not act. But, just because Frege treats actual reality to be a first level property, would he treat it as a losable property? Peter Geach thought he would. Singular sentences reporting the demise of a person, according to Geach, are to be taken as reports of loss of actuality by the deceased. But it seems clear to me that Geach is wrong, for by perishing something does not shift from one ontological category to another. By dying at the age of 71, my mother did not lose her actuality (Wirklichkeit) and join the rank of abstract entities, numbers and thoughts. To make my opposite intuition vivid—and what is a philosophical argument except analytical vivification of a logical intuition?—let me rehearse Frege’s ontology of the three realms: A. The realm of actual objective things (e.g. live people, stones, trees, events like September 11.) B. The realm of actual subjective things (e.g. ideas, mental images, pains, etc.) C. The realm of non-actual objective things (e.g. numbers, value-ranges, thoughts, senses of words) If Geach were right, then the Twin Towers of World Trade Center of New York City moved from Category A to Category C, on 11th September 2001. But, surely, Frege could not have meant that all dead people and cities and buildings which no longer exist currently count as abstract objects! What sense does it make to say about something that it was once concrete and causally efficacious, but now has become as abstract as the binomial theorem? Frege himself never equates being actual with being currently alive or available to perception. Actuality, like the property of being a physical object, seems to be a first level property but an unlosable one. The riddle, if it is a genuine riddle, about what exactly is lost by an item when it ceases to exist is not solved by the Fregean distinction between existence and actuality. When we speak of a concrete particular going clean out of existence, we are not talking about a mere transformation like wax melting. We are trying to capture the transit of the thing from being there to no longer being there, and not trying to take literally sentimental remarks such as “My dog has now become nothing more than

Anti-Realism and Realism About the Past: A Present for Mark Siderits

291

a photograph, a mere memory”. A subjective or merely dreamt-of cup is a very different cup of tea than a real but broken cup. “I cannot say of the cup I dreamt of last night that after waking up it…ceased to exist. It never did exist”.5 If I saw a cup or a house last week and this week the cup is pulverized or the house is burnt to ashes, just as it would be a mistake to think that the cup or the house—still there—has just transformed itself from an objective actual material object to a subjective imaginary mental object, it would be equally mistaken to think that by perishing the cup or the house has turned into a non-actual abstract entity. First, that would mean that an item can stay the same while exiting one ontological category and entering another. Second, it would preserve the existence of the cup when the account was supposed to show us how the cup no longer exists. Both would be fatal flaws of the account. Michael Dummett was thus justified in rejecting Geach’s view that actuality is that tensed property of being around now which the deceased people and perished material objects lack. But his reason for rejecting the view is not exactly my argument put forward above: If “Mr.N is dead” meant that Mr. N has lost actual reality, that is, he is no longer an actually real person, then by dying Mr.N would shift from the category of alive humans to the category of subjective mental things or objective abstract things such as propositions and value-ranges, which would be utterly implausible. This was my reason for disagreeing with Geach. Dummett’s reason is different. If Geach were right, then, Dummett fears, “the verb ‘exists’ would simply be equivocal: after all, what difference of sense could be greater than one involving a difference of logical type, that between a quantifier and a first level predicate?” (FPL p386). A similar worry can attach to the Buddhist omni-existentist insistence that “asti” is treated as an underivable unitary technical term for a minimal reality (compatible with momentariness) having nothing to do with “asti” as the present tensed singular conjugation of the verb to exist (as). The word becomes suspiciously equivocal. How can “there is” be a second level special quantifier-predicate of predicates while “there was” remain a first level predicate of perishable particulars admitting of tense-variations? 7 .  It is in the context of demarcating perception from memory that Bhāsaravajňa6—a robust realist with a present-centric view of existence—discusses the status of ceased past items. One defining feature of perception is supposed to be: being generated by its object (artha-ja-tva). Since the past object does not exist at the time of the cognition (with respect to which it is past), it cannot cause our awareness of it (hence by definition perception is only of the present, but for the nagging puzzle about the fact that as a cause, the object has to be “antecedent” to the perception it causes!), hence memory—our cognition of the past—is never perceptual. But then even as an object of recall, what sort of an entity is this no-longer existent past particular?  Leonard Linsky “On ceasing to Exist” Mind April 1960.  Nyāyabhūṣaṇam by Bhāsarvajna, Shaddarshana Prakashana, Varanasi 1968.

5 6

292

A. Chakrabarti

It should not be confused with its own post-cessation absence. The past event or item has ceased to be, but its cessation or posterior absence is precisely what prevails now. The absentee cannot be equated with the absence. Hence the past is not a kind of not-being or absence. Nor is it a presence or a being. Therefore, some suggest that a past entity (along with future entities) belongs to a third or middle category: neither being nor not-being, just as certain actions are neither virtuous nor unvirtuous in the sense of being vicious, they are neutral. But Bhāsaravajňa rejects this suggestion. Since there is no other appropriate use of words outside of positive and the negative, and double negation brings us back to the positive, if the perished past is not a being now, it must be a not-being. If it cannot be a non-being it must be a kind of positive entity, a kind of reality. What if someone complains that since the entirely past entity has no existence (=presence), it cannot be a real being? Well, that would be a fallacious argument since it is such real things as a pot which is said to be past, present and future, so how could it be devoid of existence? But then if the pot has existence in all three times, then would it not become eternal? Bhāsaravajňa’s answer to this is very peculiar. “Even the pot—the substratum— is not there at all times, so how could it possess the property of existence at all times? Where would the property of existence be without a property-bearer?” He, then, responds to this question as follows: The same pot—like an actor on stage— assumes different persona or roles, changing costume as it were, sometimes behaving like a yet-to-be, sometimes as now-happening and sometimes as its-all-over or bygone. But then there seems to be a regress: the very fleeting roles assumed are themselves sometimes assumed, sometimes not yet assumed and sometimes given up or relinquished. Is it the same role or character, like the pot, taking up the further role-playing character as to-be-assumed, currently assumed and already-assumed and dropped? We can understand this easily with respect to absence of something in a place. If something or someone is removed or gone from a certain place, we do not take that to mean complete annihilation of the object or person. We say that it is not in that place and from the point of view of that place it is said to be absent, though it may be present somewhere else. But with time it becomes extremely puzzling. Just as there is nothing special about here, the particular place where we are speaking from, intuitively there is nothing special about now, the present time either. Yet, things which were very much real at one point but are absent now, are not only said to be absent but said to have ceased to exist. Lack of here-ness could be mere local absence, but lack of now-ness seems to amount to not being there at all. What, then, should be the ontological status of objects which existed in the past but do not exist now? Are they absent or non-existent? 8. Without becoming committed eternalists, presentists, or omni-existentists, Nyâya realists about the past would clearly answer this question by categorizing perished particulars as absent but not non-existent. Of course, the metaphysical question of the reality or unreality of the past is intricately connected to the epistemological question of the status of memory as a variety of knowledge (of the real past). Bertrand Russell notoriously upheld the view that in correctly remembering

Anti-Realism and Realism About the Past: A Present for Mark Siderits

293

the past we have direct acquaintance with the past. But then, whatever is an object of direct Russellian acquaintance, by definition, needs to be given and present to the perceiver-knower. The relation of acquaintance is defined as the converse of the relation of presentation. If one has Russellian acquaintance with a past particular or event, then that past object would have to be present which sounds flatly contradictory. Nyâya direct realists avoided such a conundrum by refusing to accord the status of knowledge to memory. Though memory can act as an occasioning causal factor and connector in certain non-ordinary sense-perceptions such as synesthetic seeing of tastes, no direct perceptual access to the past is conceptually permissible, in Nyâya. If we mix the Russellian epistemological idiom with Nyâya realism about the past, the following question, therefore, remains unanswered: Is my episodic recall, let’s say, of how Michael Dummett laughed on a particular afternoon, forty some years ago, only knowledge by description since it cannot be knowledge by acquaintance?

Buddhist Shipping Containers Koji Tanaka

1 Siderits on Priest At the end of his review of The Fifth Corner of Four: An Essay on Buddhist Metaphysics and the Catuṣkoṭi by Graham Priest, Mark Siderits (2019) remarks: [I]t might be said that just such decentering [i.e., the decentering that can be induced by looking at another tradition] occurred when Buddhist thought came to China from India, yet we still think of East Asian Buddhist thought as Buddhist. Perhaps something similar might be said about Priest’s ‘reconstruction’ of the tetralemma in Asian Buddhist philosophy.

Priest’s The Fifth Corner of Four is a book about the argumentative method, catuṣkoṭi or tetralemma, developed and used by classical Indian philosophers including Buddhists. The Catuṣkoṭi and various other devices went through some transformation by Chinese Buddhists when they were ‘exported to China in Buddhist shipping containers’ (Siderits 2019). The book follows the development of this method by the Buddha in certain sūtras, Indian Madhyamaka, Chinese Huayan and finally Chan Buddhism. Siderits takes issue with Priest’s attribution of dialetheism (the view that there are true contradictions) to Buddhists who make use of the catuṣkoṭi. He claims that ‘Nāgārjuna et  al. did not say many of the things Priest takes them to have said’. He gives evidence to substantiate his claim. While he criticises Priest’s philology, Siderits nevertheless suggests that Priest’s work is ‘of considerable interest’ for two reasons. First, ‘when two independent traditions use similar methods to work on similar issues, it is always possible that one may have hit on approaches that the other missed’. Second, ‘the decentering that can be induced by looking at another tradition may trigger fresh insights, even if those

K. Tanaka (*) School of Philosophy, Research School of Social Sciences, Australian National University, Canberra, ACT, Australia e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_15

295

296

K. Tanaka

insights are not ones that were actually developed by the tradition marked “other”’. He then ends his review with the above quoted remark. Now, Siderits seems to be implying that Priest’s work offers new approaches to the issues that concerned traditional Buddhist philosophers and triggers fresh insights even though it is philologically problematic. What are those approaches and insights? The main task of this paper is to unpack some of Siderits’ thought on those. He explains why Priest’s take on the catuṣkoṭi as it was used and conceived by classical Indian philosophers is problematic. However, he is rather cryptic about the positive contributions that he thinks Priest’s work brings out. But I am not going to conduct another book review. Instead, by analysing some of Priest’s work, I will suggest a decentering that must have taken place in China. In particular, I will suggest that the two threads of Siderits’ remark are tightly connected. That is, I will show that Priest has ‘sinologised’ Indian Buddhism with the use of modern logic. For anyone familiar with the forms of Buddhist philosophy developed in India, various forms of Chinese Buddhism can often look strange. Some ideas that were debated and defended by Indian Buddhists do not appear anywhere in Chinese Buddhism and some assumptions have been quietly added to Buddhism by Chinese Buddhists. By showing that Priest’s work brings out clearly the ‘innovation’ that Chinese Buddhists gave to Buddhism as it was ‘exported to China in Buddhist shipping containers’, I will show that the decentering that must have taken place in China is exactly what Priest makes use of.

2 On Nāgārjuna’s Ontological Paradox Many of Priest’s thoughts contained in The Fifth Corner of Four can be traced to his early work with Jay Garfield. What is crucial in my reconstruction of Siderits’s suggestion is already present and more explicit in the early work. So, I will follow Priest’s early work rather than his latest work.1 Garfield and Priest (2003) argue that Nāgārjuna’s doctrine of emptiness, when it is analysed to its logical end, can be seen as entailing that reality is contradictory. They do not claim that Nāgārjuna presents an argument or arguments for the contradictory nature of reality. Nor do they claim that he (explicitly) accepts that reality is contradictory. Their claim is that the doctrine of emptiness entails that reality is contradictory and this can be shown by rational reconstruction of Nāgārjuna’s thought. Garfield and Priest understand the doctrine of emptiness to be that nothing exists with svabhāva. The term svabhāva has several connotations. But when it is understood as essential property or essence that ultimately gives an object its numerical

 I will largely follow Tanaka (2016) in what follows. For Siderits’ criticism of Priest’s work with Garfield, see Siderits (2013). For Garfield and Priest’s latest rejoinder, see Garfield and Priest (this volume). 1

Buddhist Shipping Containers

297

identity, the lack of svabhāva means that ultimately nothing exists. So, the doctrine of emptiness entails that ‘there is no ultimate reality’ (p. 10). According to Garfield and Priest, Nāgārjuna nevertheless ‘tells us about the nature of ultimate reality’ (p. 11) suggesting that there is ultimate reality for Nāgārjuna. Hence, so Garfield and Priest conclude, the doctrine of emptiness entails that there is and is not ultimate reality. This means that if something exists, it does and does not exist ultimately. They call it Nāgārjuna’s ontological paradox: what ultimately exists is paradoxical. They attribute to Nāgārjuna the claim that the ultimate reality is paradoxical.2 Garfield and Priest’s interpretation of Nāgārjuna is radical in the sense that the Madhyamaka tradition (as well as most Buddhist traditions) is usually thought of as rejecting contradictions (perhaps except some Chinese and Japanese traditions such as Chan/Zen). Nāgārjuna himself seems to be sensitive to the idea that his doctrine is contradictory and rejects such an idea.3 Garfield and Priest are well aware of this. They take on the ‘burden of proof’ to demonstrate that the doctrine of emptiness entails the ontological paradox. As part of demonstration, they show that the doctrine of emptiness can be characterised by a structure in terms of the inclosure schema. The schema has two parts. First, transcendence: something, which has a limit, transcends that limit. Second, closure: it is, nevertheless, within the limit. By showing that the inclosure schema entails a contradiction, they argue that what ultimately exists must be paradoxical for Nāgārjuna. As I will show in the next section, the inclosure schema entails the domain principle: for anything of a certain kind, there are all of those things. If the doctrine of emptiness can be characterised by the inclosure schema and the conclusion entailed by the schema can be attributed to Nāgārjuna, Garfield and Priest must accept the idea that Nāgārjuna is committed to the domain principle. However, the domain principle implies the existence of a whole under which all things can be unified. This means that Garfield and Priest’s analysis entails that Nāgārjuna is committed to the existence of the one that categorises all things as empty. In the Indian context, attributing conceptual recourses that commit Nāgārjuna to the existence of the one is questionable. The idea that there is an all-encompassing whole did develop in China (and transported to Korea and Japan). Garfield and Priest’s analysis of Nāgārjuna is, thus, a sinologification of Indian ideas which is part of the decentering that might have ‘occurred when Buddhist thought came to China from India’ (Siderits 2019). This is what I will show in the remainder of the paper.

 In this paper, I will deal directly with the ontological paradox and leave aside the issue about language. For a discussion of the semantic issues that come with the ontological paradox, see Tanaka (2016). 3  See, for instance, Vigrahavyāvaryanī 29. For a discussion, see Westerhoff (2009) pp. 183ff. 2

298

K. Tanaka

3 The Inclosure Schema and the Domain Principle My first task is to show that the inclosure schema entails the domain principle: for anything of a certain kind, there are all of those things. The inclosure schema concerns properties, φ and ψ, and a function, δ, that satisfy the following conditions: (1) Ω = {X : φ(X)} exists and ψ(Ω). (2) For all X ⊆ Ω such that ψ(X): (i) δ(X) ∉ X (Transcendence) (ii) δ(X) ∈ Ω (Closure) Let φ(X) to mean that X is empty, ψ(X) that X is a set of things with some common nature, and δ(X) is the nature of things in X. Then (1) says that the set of all things that are empty exists and that such a set is a set of things with some common nature. It is an ontological claim about what exists and a metaphysical claim about what it is like. (i) says that the nature of things in the set of empty things having some common nature is not part of the set of empty things, and (ii) says that it is part of the set of all things that are empty. That the inclosure schema interpreted as above entails an ontological paradox can be shown as follows. Assume that X ⊆ Ω and ψ(X) (i.e., X is a set of empty things with some common nature). δ(X) is that nature. But all things are empty. So δ(X) ∈ Ω (where Ω is the set of all empty things) (Closure). This means that δ(X) (i.e., the nature of all empty things) has no nature. So δ(X) ∉ X (Transcendence). Hence δ(X) ∈ X (the nature of all empty things is empty) and δ(X) ∉ X (the nature of all empty things is not empty since emptiness is the nature commonly shared by all things that are empty). This means that emptiness is and is not empty. This must be the case, Garfield and Priest argue, since all empty things share a common nature, namely, emptiness. Given that emptiness pertains to ultimate reality, what ultimately exists is empty since everything is ultimately empty but it is not empty since it has emptiness as its nature. Can we legitimately attribute this ontological paradox to Nāgārjuna? The schema is formulated set-theoretically. It is hard to know how Nāgārjuna would understand set-theoretic notions such as set membership. To be charitable to Garfield and Priest, let’s understand the schema in terms that Nāgārjuna might understand.4 He might understand mereological relations (part-whole relations). If we understand the schema mereologically, (1) says that any empty thing is part of the totality of all empty things. But this is the domain principle. So (1), understood mereologically, is an instance of the domain principle: for anything of a certain kind, there are all of those things. This shows that the inclosure schema entails the domain principle.

 I say ‘might’ because I am very sure that Nāgārjuna would not understand what Garfield and Priest are talking about in terms of the inclosure schema. The conceptual resources such as modern mathematics that are necessary to understand the inclosure schema were just not available to Nāgārjuna. 4

Buddhist Shipping Containers

299

Now the domain principle commits one to the existence of wholes that are constituted by myriad things and that unify these things to be parts of the wholes. The first half of (1) of the inclosure schema says that there exists a totality of things that can be uniformly characterised as being empty. Given that everything is empty, it effectively asserts the existence of the totality of empty things. This means that Garfield and Priest’s analysis commits the existence of the one that encompasses and unifies all myriad things as the Chinese would put it.

4 The Domain Principle in the Indian Context Before examining the domain principle in the context of Nāgārjuna, I should make clear what is not at issue. What is not at issue is the notion of totality or the possibility of talking about all things whether ultimately or conventionally. Nāgārjuna (and Mādhymikas) might be comfortable talking about all things. He might even admit the existence of the totality of all things. Indeed, the doctrine of emptiness contains a universal quantifier over all things: all things are empty. So, he must be able to talk about all things just like the Buddha must have been able to talk about all things by claiming that all things are duḥkha (suffering). What is at issue is the domain principle that, for each thing that is empty, it forms part of the whole consisting of all empty things. To put it using the terminology of particular/universal, the domain principle has it that for each particular thing that is empty, it forms part of the universal under which all particular empty things fall. In other words, what the domain principle, understood mereologically, commits one to is the existence of the whole that unifies all things as empty. From the perspectives of Vasubandhu and Dignāga-Dharmakīrti, there cannot be such unifying wholes. They would, thus, reject the domain principle. For instance, Vasubandhu, in the Abhidharmakośa, argues for the position that there are only particulars. He employs a reductive analysis to come to this conclusion. He shows that any macro-object can be analysed in terms of its parts. Because of this, macro-­ objects do not have real or ultimate existence. For Vasubandhu, there cannot ultimately be wholes that unify things and classifies them as certain kinds.5 So, for him, it is not the case that for any particular thing, there are all of those things. Hence, the domain principle must be rejected by Vasubandhu. What about Nāgārjuna? Would he accept the domain principle? Given that he does not explicitly accept or reject the principle or anything similar (as far as I can

 See, for instance, Siderits (2007) Chap. 6.

5

300

K. Tanaka

see), it is hard to categorically demonstrate that he would accept or reject it.6 However, there is reason to think that Nāgārjuna would not accept it. To think that Nāgārjuna would accept the existence of unifying wholes, we would have to think of him as rejecting the kind of argument that Vasubandhu advances against the ultimate existence of wholes in the Abhidharmakośa. In particular, Nāgārjuna would be seen as rejecting the reductive analysis showing that macro-objects do not have real or ultimate existence. Far from rejecting such an analysis, at least in some parts of his arguments, Nāgārjuna seems to be applying it to the ‘parts’ and showing that the parts themselves cannot have real or ultimate existence on the very basis of reductive analyses. As Garfield (2015) writes: The anti-realism with respect to the macroscopic, composite entities of ordinary life espoused by early Abhidharma philosophy is extended in the Madhyamaka perspective to the dharmas [i.e., the ultimate parts] themselves. (p. 62)

An application of reductive analysis to its logical end showing the non-existence of ultimate parts is an integral part of Nāgārjuna’s argument. Based on such an application, he was able to show that nothing has ultimate existence and, thus, there is no ultimate reality. If this is right, it is hard to think of Nāgārjuna as rejecting Vasubandhu’s argument (or similar arguments). If he were to reject Vasubandhu’s argument, he would be undermining his own argument. There is, thus, a prima facie case against attributing the commitment to the existence of unifying wholes and, hence, the domain principle to Nāgārjuna. I cannot conclude from the above discussion that no Buddhists in India were committed to the domain principle. Such a general conclusion would require a thorough investigation of the history of Buddhism in India. For instance, we would have to examine the pudgalavādins who seem to have rejected Vasubandhu’s (or someone like him) reductive analysis.7 However, what I have shown is that some of the major figures in Indian Buddhist philosophy would not accept the domain principle. When we use the terminology of particular and universal, the domain principle says that for each particular, there is a universal under which it falls. So, the domain principle would entail the existence of universals. Given that the commitment or the lack of the commitment to the existence of universals was one of the distinctive issues that

 It is never clear to me how to evaluate historical counterfactuals such as: If Nāgārjuna were here now reading Garfield and Priest’s work, he would accept the inclosure schema. How can we examine the conceptual apparatus of Nāgārjuna if it is not the conceptual apparatus that he had when he was writing down his thought but the one that he would have had if he were in a position to understand Garfield and Priest’s work? Priest (2018), on the other hand, seems optimistic when he says: ‘Would the actors in the drama of this book have used those insights and tools had they been available to them? I see no reason why not.’ (p. 149.) I do not intend to settle this issue in this paper. However, I do wonder where Priest’s confidence comes from. 7  For an examination of the pudgalavādins, see Carpenter (2015). The analysis of the pudgalavādins by Coseru (this volume) does not settle the question of whether or not the pudgalavādins presupposed or (implicitly) made use of the domain principle. 6

Buddhist Shipping Containers

301

separated non-Buddhists from Buddhists in India, an analysis of Indian Buddhist thought that entails the existence of universals is, therefore, generally problematic.

5 The Domain Principle in the Chinese Context The thought that there is a whole of which everything, every myriad thing, is a part is ubiquitous in China.8 The Book of Changes describes the world to have began as primordial qi (here understood as vital energy) which is undifferentiated. Over time, this undifferentiated whole fractured into layers or regions. Eventually, discrete shapes and individual things emerged out of this process. In this way, ‘the things of the universe are one in a deep and distinctive sense’ (Ivanhoe 2017: 20–21). Early Daoists gave this metaphysical view a twist. Instead of undifferentiated, primordial qi, they considered the world to have come from wu (nothing) the state where there are no things (as opposed to the state of nothingness). This idea is expressed in various passages in the Daodejing. Neo-Confucians gave this traditional metaphysical view yet another twist. The Book of Changes and the Daodejing present the world as a single body and everything in it is its part. Neo-Confucians think of the world as a unified system consisting of li (principles or patterns). The world is conceived as a living organism with a unified structure and every part of it is interconnected in terms of this structure. Under the influence of Buddhism, in particular Huayan Buddhism with the image of the Net of Indra, they took this to mean that everything in the world contains the principles or patterns of the world. This allowed them to ‘develop a more robust and dramatic sense of oneness as a kind of identity between self and world’ (Ivanhoe 2017: 22).9 It is important to note that the principles or patterns that manifest in everything in the world are the principles or patterns of the world. And, to think of the principles or patterns of the world is to think of the world as a whole. Everything in the world is unified under the principles or patterns of the world and the world is the whole of which everything in it is a part.10 For neo-Confucians, thus, there is the whole which unifies all the myriad things. The same or similar metaphysical view can be found in various traditions of Chinese Buddhism. I think one of the most striking differences between the forms

 A proper historical study of the development of Chinese thought is too complex to be conducted in one place. In this paper, I will follow Ivanhoe (2000, 2002, 2017). See also Ivanhoe, Flanagan, Harrison, Schwitzgebel and Sarkissian (2018). 9  It is not clear that Huayan Buddhists think of the relationship between the world (whole) and everything in it (every part) to be a kind of identity. For an alternative analysis of Huayan metaphysics, see Jones (2015). Here I follow Ivanhoe and assume for the sake of this paper that neo-­ Confucians have interpreted Huayan Buddhists to be saying that the relation is a kind of identity 10  This seems to be how Zhu Xi understand li (principles or patterns). See Kim (2000) for a discussion. 8

302

K. Tanaka

of Buddhist philosophy (scholarsticism) we know existed in India and those that developed in China has to do with the attitudes they have towards the whole that unifies all the myriad things. This is not the place to examine or speculate about the direction of influence between Buddhism and more ‘indigenous’ Chinese schools of thought such as neo-Confucianism and I won’t offer any historical examination of who influenced who. What I will do below is briefly present the metaphysical views held by some of the Chinese Buddhists that implies the existence of the One, the whole under which all things are unified. Chinese Buddhists tend to subscribe to and make use of not the Two Truths that often plays out in Indian context including Madhyamaka but the Three Truths.11 What those three truths are depends on the ‘schools’ of Chinese Buddhism. For Tiantai, according to Ziporyn (2016), the first and the second truths represent two extremes and the third truth represents the centre (zhong) that sits between those extremes. This centre is an important element of their metaphysical view as it ‘denotes the inclusiveness of all opposites as well as the subsuming field that unifies them all. It is thus the ground of the being of all the entities it subsumes, the centre of gravity that brings them into the relationships with each other that determines their identities’ (p. 153). Instead of thinking of the interconnection between things in the world in terms of principles or patters (li) as neo-Confucians subscribe, Tiantai views the interconnection in terms of what Ziporyn calls ‘recontexualisation’. To be something is to be a part of the world. Thus, to understand the nature of something is not to see it as it is in itself but to ‘recontextualise’ it by the rest of the world (Ziporyn 2016: 163). For Tiantai, the unification of the world is a transcendental condition that makes every little fragment of the world ‘findable’ (Ziporyn 2016: 143). A very much the same metaphysical view can be found in Huayan. The metaphor that is often used to illustrate reality by Huayan metaphysicians is the Net of Indra. The Net of Indra is described as consisting of infinite number of jewels. Each jewel (thought of as a node of the web) is said to be connected to and reflect all the other jewels of the Net. Each jewel ‘interpenetrates’ (or ‘mutually depends on’) each other. The whole Net also ‘interpenetrates’ each of the parts. By ‘interpenetrating’ each part, the whole gives unity to them. But the whole is not identical with any of the parts. This means that the whole exists over and above the parts. However, the ‘interpenetraring’ relations form an infinite chain which consists of all of the parts but also the whole. The whole is, thus, a part of this chain as much as all of the (proper) parts are. So, the whole is not distinct from the parts either. The whole Net is, thus, not distinct from the parts nor not distinct from them. Nevertheless, the whole is real as much as the myriad things are.12 This is not to say that Chinese philosophers have largely accepted the domain principle. I am not aware of anyone who has formulated their thought in terms of the  Three Truths, or something similar, can already be found in Vasubandhu’s Yogācāra. See, in particular, Trisvabhāvanirdeśa. Again, I leave speculation about the influence of Yogācāra for Chinese Buddhism for another occasion. 12  This is exactly the view that the modern analysis of the Net of Indra by Priest (2014) provides. 11

Buddhist Shipping Containers

303

principle. Nevertheless, the intellectual resources available in China are amenable to accepting the domain principle. At least, it is not difficult to imagine that the domain principle would find wide acceptance in China if it can be presented with appropriate terminologies.

6 Buddhist Shipping Containers Siderits starts his book review of Priest (2018) by writing that ‘the catuṣkoṭi or tetralemma [is] a device used by some classical Indian philosophers and then exported to China in Buddhist shipping containers’. What happened in the Buddhist shipping containers? A lot of things must have happened not just to the catuṣkoṭi but also to many other devices used by philosophers (whether Buddhist or not) in India. Following the above discussion, I speculate that one thing that must have happened is the change of focus. As we have seen, Chinese philosophers focus on the interconnections between the myriad things that exist and that we can observe in the world. This is a change of focus from emptiness to interdependence. One of the emphasis the notion of emptiness brings is the non-existence of ultimate reality. When the focus shifts to interdependence, the emphasis moves to the nature of the relationship between existing things.13 But, in so doing, there is a tendency to affirm the existence of ultimate reality where the ultimate reality is now conceived to be that everything is interdependent. It is this decentering that allows Chinese Buddhists to commit to the existence of the One. From a Chinese perspective, then, the analysis that Priest (with Garfield) provides makes perfect sense. The inclosure schema that he (or they) make use of entails the domain principle that commits the existence of the totality under which everything is unified. As we saw, there are various devices in China that are readily available to make good on that commitment. So, what could we have discovered when we opened the Buddhist shipping containers? One thing we might have found there is a contradiction (or various contradictions). Priest’s sinologisation reveals that the thought that existed in India can be seen to entail contradictions when examined through Chinese lens with the help of modern logic. Dōgen complained about many of the thoughts that were entertained by Chinese Buddhists. For instance, he writes: [T]here are careless fellows who form groups; they cannot be set straight by the few true masters. They say that the statement, “The eastern mountains travel in water,” or Nanquan’s story of a sickle, is illogical; what they mean is that any words having to do with logical thought are not buddha ancestors’ Zen stories, and that only illogical stories are buddha

 Priest (2015) presents this shift as entailment: emptiness entails interdependence. The entailment he presents seems invalid, however. Thanks go to Alan Háyek for pointing this out. An examination of Priest (2015) goes beyond the scope of this paper and so I will leave it for another occasion. Also I won’t speculate whether or not any historical Buddhist tried to derive interdependence from emptiness and whether or not such derivation is valid. 13

304

K. Tanaka

ancestors’ expressions. ... “Ancient masters used expedient phrases, which are beyond understanding, to slash entangles vines”: People who say this have never seen a true master and they have no eye of understanding. [T]hese last two or three hundred years, there have been many groups of bald-headed rascals. What a pity! ... How sad that they do not know about the phrases of logical thought, or penetrating logical thought in the phrases and stories! (Sansuikyō, translation in Tanahashi (1985).)

Interpreting Dōgen is notoriously difficult and I do not claim to have an authoritative interpretation of the passage.14 However, prima facie at least, Dōgen is admonishing those who subscribe to the ‘illogical’ nature of the Buddha’s teaching. In the context where paraconsistent logic (logic which does not trivialise contradictions) is not available, illogicality is equivalent to contradiction. Thus, Dōgen is criticising those who have interpreted the true teaching of the Buddha to be contradictory. Instead of hitting with keisaku as might happen in a Zen meditation hall, Siderits sits in the middle of all kinds of activities taking place around him while making crucial contributions to them. He observes that certain decentering can take place when some intellectual tradition is shipped to somewhere else.15 When various Buddhist devices were shipped to China from India, a new approach to the understanding of emptiness opened up. And such an approach might have provided fresh insights that the ultimate reality, once it is reconceptualised in the presence of the One, may, perhaps, be contradictory. Siderits teaches us that decentering may be inevitable and also crucial in the development of intellectual traditions. We just need to be reflective about our engagement with historical material by paying attention to the decentering that can take place when the shipping containers arrive at the destination.

References Carpenter, A. 2015. Persons Keeping Their Karma Together. In The Moon Points Back, ed. K. Tanaka et al., 1–44. Oxford: Oxford University Press. Garfield, J. 2015. Engaging Buddhism. New York: Oxford University Press. Garfield, J., and G. Priest. 2003. Nāgārjuna and the Limits of Thought. Philosophy East and West 53: 1–21. Ivanhoe, P.J. 2000. Confucian Moral Self Cultivation (Second Edition). Indianapolis: Hackett. ———. 2002. Ethics in the Confucian Tradition: The Thought of Mengzi and Wang Yangming (Second Edition). Indianapolis: Hackett. ———. 2017. Oneness. New York: Oxford University Press. Ivanhoe, P.J., O. Flaganan, V. Harrison, E. Schwitzgebel, and H. Sarkissian, eds. 2018. Oneness Hypothesis. Cambridge: MIT Press.

 This translation maybe questionable as a literally accurate translation. Yet I think it is one that captures the poetic aspect of Dōgen’s writing and, for that reason, captures his ‘sentiment’ he expresses in his writing. It is, thus, appropriate to use this translation in this context. 15  Westerhoff (2020) observes something similar which he describes in analogy to the study of ancient technology. 14

Buddhist Shipping Containers

305

Jones, N. 2015. Buddhist Reductionism and Emptiness in Huayan Perspective. In The Moon Points Back, ed. K. Tanaka et al., 128–149. Oxford: Oxford University Press. Kim, Y.S. 2000. The Natural Philosophy of Chu Hsi. Philadelphia: American Philosophical Society. Priest, G. 2014. One. Oxford: Oxford University Press. ———. 2015. The Net of Indra. In The Moon Points Back, ed. K. Tanaka et al., 113–127. Oxford: Oxford University Press. ———. 2018. The Fifth Corner of Four: An Essay on Buddhist Metaphysics and the Catuṣkoṭi. Oxford: Oxford University Press. Siderits, M. 2007. Buddhism as Philosophy. Aldershot: Ashgate. ———. 2013. Does a Table Have Buddha-Nature? Philosophy East and West 63: 373–386. ———. 2019. Review of The Fifth Corner of Four: An Essay on Buddhist Metaphysics and the Catuṣkoṭi by Graham Priest. Notre Dame Philosophical Reviews. Tanahashi, K., trans. 1985. Moon in a Dewdrop. San Francisco: North Point Press. Tanaka, K. 2016. On Nāgārjuna’s Ontological and Semantic Paradox. Philosophy East and West 66: 1292–1306. Westerhoff, J. 2009. Nāgārjuna’s Madhyamaka: A Philosophical Introduction. Oxford: Oxford University Press. ———. 2020. ‘The Fifth Corner of Four: An Essay on Buddhist Metaphysics and the Catuṣkoṭi by Graham Priest’ (book review), Mind: 129: 965–974. Ziporyn, B. 2016. Emptiness and Omnipresence. Bloomington: Indiana University Press.

Finite and Infinite: On Not Making ‘Them’ Different Enough Rupert Read and Christian Greiffenhagen

Abbreviations LFM Lectures on the Foundations of Mathematics, Ludwig Wittgenstein PG Philosophical Grammar, Ludwig Wittgenstein PI Philosophical Investigations, Ludwig Wittgenstein PR Philosophical Remarks, Ludwig Wittgenstein ROM II Remarks on the Foundations of Mathematics volume 2, Ludwig Wittgenstein WVC Wittgenstein and the Vienna Circle

1 Introduction The infinite has always stirred the emotions of mankind more deeply than any other question; the infinite has stimulated and fertilized reason as few other ideas have; but also the infinite, more than any other notion, is in need of clarification. (Hilbert 1925, p.371)

We agree with Hilbert’s assessment that the concept of ‘infinite’ stands in need of clarification – however, our proposed ‘solution’ is almost diammetrically opposed to that of Hilbert.

R. Read (*) University of East Anglia, School of Politics, Philosophy, Language and Communication Studies, Norwich, UK e-mail: [email protected] C. Greiffenhagen The Chinese University of Hong Kong, Department of Sociology, Kowloon Tong, Hong Kong e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_16

307

308

R. Read and C. Greiffenhagen

2 On Not Making Different Enough There seems to be something inherently puzzling about the notion of ‘infinity’. In this paper we want to argue that most of the puzzles arise by making the ‘finite’ and the ‘infinite’ not different enough. In particular, the concept of ‘infinite’ is often taken to mean (‘simply’) ‘infinitely big’ which suggests that it is like or akin to something finite (a number or an object) but that its ‘size’ is just unimaginably or inconceivably great. We are suggesting that ‘finite’ and ‘infinite’ are of a different order and are therefore not (easily) comparable, at all. Even to say that they are ‘incommensurable’ might be too weak (or even misleading), as the notion of incommensurability suggests a certain commonality. For example, when Kuhn characterizes Newtonian and Einsteinian physics as ‘incommensurable’, he simultaneously emphasizes that they are both scientific theories (within the philosophy and history of science – although perhaps not for the scientists themselves). In contrast, the difference between finite and infinite is more like that between science and religion, or between molecules and thinking, or between space and smells; or between numerals and numbers… We think that there is a tendency (and temptation) to turn ‘the infinite’ into an object (which tendency we might term ‘thingification’, a more vividly allusive term than its more popular near-synonym, “reification”), an object that although it is ‘infinitely big’ is still a something. It is sometimes also said that our ‘finite minds’ cannot ‘grasp’ ‘the infinite’. We are thinking here of popular science texts which tend to say things like: “Our minds cannot ‘grasp’ the notion of ‘infinity’ or ‘seventeen dimensions’,” and thus encourage an unnecessary and unwarranted sense of boggle and of obeisance before the ‘intellectual authority’ of theoretical physicists. One nowadays-very-popular way of dealing with this phenomenon is to postulate the ‘cognitive closure’ of the human mind; to claim that there are thoughts that the finite human mind is incapable of thinking. This move is uncannily reminiscent of what for centuries was standard theological doctrine; it is amusing, to say the least, to see ruthlessly scientistic figures in this regard parroting the worst excesses of the long-standing theological denigration of the human.1 We are sceptical of the intelligibility of this move2; we think that the limit to thought is no limit at all. This doesn’t of course mean that the human mind is infinite; what it means is that we are not clear that any sense has yet been given to the contrast ‘finite’ vs. ‘infinite’, where the human mind is concerned. Saying “Our minds cannot grasp the notion of  A nice example here is provided by that would-be ‘paragon’ of atheism, A.J.  Ayer: “A being whose intellect was infinitely powerful would take no interest in logic and mathematics. For he would be able to see at a glance everything that his definitions implied, and, accordingly, could never learn anything from logical inference which he was not fully conscious of already. But our intellects are not of this order.” (1932, p.82) Compare also here Read’s argument against Michael Dummett, who charmingly admits the theological root of his philosophising around finite and infinite: see Read 2008a. 2  And our scepticism has of course Wittgensteinian as well as Nietzschean grounds: “It isn’t just impossible ‘for us men’ to run through the natural numbers one by one; it’s impossible, it means nothing.” (Wittgenstein, PR, p.146). For amplification, see for instance Read’s “What does ‘signify’ signify?”, Read 2001a and resolute readings of Wittgenstein’s philosophy. 1

Finite and Infinite: On Not Making ‘Them’ Different Enough

309

‘infinity’ or ‘seventeen dimensions’” is, so far as we are concerned, as yet no better than saying “Our minds cannot grasp the notion of its being a certain time on the Sun right now, e.g. Five o’clock; and yet, it is a certain time on the Sun right now, whether we know it or not.”3 The thingificational move is to suggest that the infinite is something just like the finite except much much (and so on…) bigger. In a way, then, the problem lies in thinking that the ‘the’ in “the finite” and “the infinite” is similar; the “the” in “the infinite” is taken to indicate some thing, as it does in the case of the finite. Of course, this could all simply be taken to signify ‘concepts’, which would still leave open the possibility that these two concepts could be very different. However, it seems to us that the ‘the’ is more often taken to point to entities, in which case it is easy to overlook the entirely fundamental differences between “the finite” and “the infinite”. It is easy to fall into the temptation to make ‘the finite’ and ‘the infinite’ comparable: for example, within mathematicians we might talk of ‘infinitely many numbers’, an expression which has the same form as ‘seven numbers’. It suggests that it is quite a small step to go from, say, a set of seven numbers to a set of infinitely many numbers. However, we would argue that this step (from the finite to the infinite) is enormous––incalculable!––and involves changing the sense of the expressions that were used to describe the finite case. In fact, the word “enormous” above is still ineluctably inadequate, systematically misleading. Any word that one uses hereabouts is inadequate to steer one clear of likely misunderstanding.4 The position here is much akin to that delicately exposed by Frege at the greatest moments in his philosophising, when he spoke of “hints” or of “pinches of salt”. The task is to find a way to philosophise therapeutically so as to minimise (not remove entirely – that is just not on the cards) the risk of confusing oneself and others, and to philosophise in a way which does not, as Frege does, leave one seeming to hint at something inexpressible, something ineffable which we then fantasise could be effed.5 That simply places the problem at one remove, and does not (dis-)solve it. For we then keep the finite and infinite properly deeply separate and different, but at the cost of bringing them closer together again at the level of the unsayable. We then find ourselves thinking things like: for God, the infinite and the finite are not so different after all. For Him, they can both be surveyed. So: our point in this paper might be happily presented rather as follows: even for God, the finite and the infinite are utterly discrepant (This is a grammatical, not an ontological, empirical, or religious remark). They are not two kinds of a thing, two versions of one substance, at all. Their grammars are through and through discreet. Take another example of where finitude and infinitude have been run too close together. Linguists (famously, Chomsky) have likened our ability to speak (i.e., the competence of language users) as being able to create an ‘infinite number of  This is a justly famous example from Wittgenstein; see e.g. PI para.s 350–1.  As we shall briefly surmise in the Conclusion, this is arguably why Buddhist thinking at its best — in some Zen, and in Nāgārjuna, for instance—does not trust any assertoric let alone theoretical thought. 5  See below for discussion of this in specific connection with a famous ‘locale’ for the discussion: Wittgenstein’s Tractatus. 3 4

310

R. Read and C. Greiffenhagen

sentences’.6 Here, this ‘in principle’ set of sentences (which is infinite) is compared with the ‘actual’ set of uttered sentences in anybody’s lifetime (which is finite). See the following examples: on mastering a finite vocabulary and a finitely stated set of rules, we are prepared to produce and to understand any of a potential infinitude of sentences (Davidson 1967: p.304) the enormous gulf between the finite mind and the infinite number of sentences that a competent speaker can understand (Niles 1992, p.195)

Another epochal example, ‘inherited’ again from Frege, is the following: philosophers discussing Wittgenstein’s Tractatus have (in its ‘ineffabilitist’ reading) summarised the book as: ‘the Tractatus shows something that cannot be said’ – which then makes it comparable to ‘something that can be said’. This mythic mistake in the reading of Wittgenstein is deeply parallel to the mistake of not making different enough that we are emphasizing here: and one nice feature of this case for our purposes is that a great deal has been written in recent years making clear the mistake, in the case of reading the Tractatus.7 In each of these cases, ‘the infinite’ (or ‘the unsayable’) is turned into something that can be compared to ‘the finite’ (or ‘the sayable’). In contrast, we want to argue that the two are of a different mode or order of being. One way in which this might be put is: comparing ‘finite’ and ‘infinite’ is committing a category mistake (but see below). Thus when it comes to similarities and differences it seems us that in these cases, it is easy to overlook the differences and only focus on the similarities (So: another example might be the way in which humans ‘tell the time’ and the way in which watches ‘tell the time’).

3 Ontic Versus Ontological We might try to express what we are trying to say using Heidegger’s ‘ontic—ontological’ distinction.8 Heidegger is pointing to the danger of making ‘existing’ and ‘the thing that exists’ not different enough. Heidegger took this danger to be writ  Cf. Coulter 1991  See for instance the essays collected in Part II of The New Wittgenstein (Crary and Read, 2000), especially that by Conant (and excluding that by Hacker, a leading exponent of the ‘ineffabilist’ reading of Wittgenstein that we are here challenging as inadequately radical, inadequately characterising the difference between saying and showing, by means of making the latter look like just a deviant or mysterious version of the former). See also the arguments made by Read and Deans (2000). These papers make clear that we resolutists are the true friends of showing. It is the ineffabilists who fail to take ‘showing’ sufficiently seriously – for they covertly render it as just another form of saying. 8  For discussion, see Read’s work critiquing Louis Sass, e.g. his “On approaching schizophrenia through Wittgenstein (2001b)”, and Sass’s “Heidegger, schizophrenia and the ontological difference”(1992), 109–132. Schizophrenics make the character of ‘the ontological difference’ abundantly clear, by virtue of their so palpably mistaking ontological matters for ontic ones, whereas philosophers (such as Descartes) make the same error, but so much less obviously. 6 7

Finite and Infinite: On Not Making ‘Them’ Different Enough

311

large throughout Western philosophy, and took the making clear of “the ontological difference” to be a primary task of true philosophical thinking. Examples: it is the danger of making the following not different enough (and in each case, this involves ‘thingification’): • Theology: does God exist? – does a stone/an airplane exist?9 • Winch/Evans-Pritchard: witchcraft/science10 • Frege: concepts/object11 Or consider Descartes, at the foundation of modern Western philosophy. He wanted to consider the ‘mind’ and the ‘body’ as very different entities. He wanted to say that the mind and the body are totally different. However, all he managed to say was that they are totally different things. For Descartes, the difference was only ontic – whereas we think that the difference between them is ontological.12 Wittgenstein makes this criticism, implicitly, of Descartes, at PI 339 (see below). This is in fact a very important and little understood dimension of Wittgenstein’s ‘private language’ considerations. It is fairly widely understood that Wittgenstein wished to reduce the felt sense of an ineluctable ‘gap’ between mind and body, between consciousness and physicality13; that he wished to enable us to have a sense of ourselves as whole and wholly unproblematically embodied creatures, through and through. But what it far less widely understood is that Wittgenstein was just as concerned to avoid unduly assimilating the ‘grammar’ of mind and the ‘grammar’ of body (to avoid, for instance, the deadly error of all forms of behaviourism). Wittgenstein thought that Cartesianism was not just too inclined to produce a sense of a gulf between mind and body, but also and equally too inclined to make them out to be the same kind of thing. Thus, Cartesianism not only in one critical respect over-estimates the difference between mind and body; in another critical respect it under-estimates it. For it misses, in Heidegger’s sense, the ontological difference between body and mind. Take for instance PI 339 [underlinings added], a brilliant passage: “One might say “Thinking is an incorporeal process”…if one were using this to distinguish the grammar of the word “think” from that of, say, the word “eat”. Only that makes the difference between the meanings look too slight. (It is like saying: numerals are actual, and numbers non-actual, objects.)”14  For discussion, see Wittgenstein’s Lectures and Conversations on Aesthetics, Psychology and Religion, especially Lecture I of the Religion lectures. 10  For explication, see Hutchinson, Read & Sharrock 2008. 11  See Frege’s great critique of Kerry, and discussions thereof by Conant, Read, etc.. When one tries to speak of the concept horse, one misses one’s target. When one tries to speak of infinity, one misses one’s target. Whenever we imagine infinity, we miss it. For ‘infinity’ is ontological, not ontic. 12  It is perhaps worth noting here that some recent authors, most notably radically therapeutic Wittgensteinians Gordon Baker and Katharine Morris, have argued that the point made here is true only of ‘Cartesianism’, not of Descartes’s own views. This interesting exegetical question is beyond the scope of our paper to investigate. 13  For detail of Wittgenstein’s therapeutic way of doing this, see e.g. Read 2008b which focuses quite heavily on the important passages around PI 420. Cf. also Read 2012, Chap. 8. 14  Compare also PI 304. 9

312

R. Read and C. Greiffenhagen

This remark of Wittgenstein’s might be a suitable epigraph for the present paper. It captures precisely the kind of thing we want to say about the way the terms ‘infinite’ and ‘finite’ are generally used, when they lead into philosophical confusion. (And what this indicates is, ironically, that it is we Wittgensteinians who are the true friends of mind.15 By contrast, substance dualists turn mind into just another kind of stuff.) Infinity is not just more of the same––it differs from finity in the same truly ontological fashion that mind differs from body. Reducing the difference between ‘mind’ and ‘body’ to (only) an ontic difference is the heart of Ryle’s critique in The Concept of Mind (1949) and the leading instance of import of Ryle’s idea of a ‘category mistake’. Ryle’s analogy with the University on the one hand, and the Colleges, Libraries, Departments, etc. on the other, is meant to explain this. However, even Ryle is in danger of reducing the differences to ontic ones, namely that both sides are ‘categories’ but not (just) the same.16 In all these cases if we do not make these two things different enough, then we are likely to run into problems (confusions) later on. The difficulty here is a deep philosophical difficulty; such difficulties do not get any deeper than this, we would submit. And their ramifications are extremely wide.17 Philosophical therapy involves as a central task overcoming the superficial similarities between declarative sentences (e.g. “This task is finite”, “This task is infinite”) whose grammars are deeply discrepant. (Something of relevance here is Wittgenstein’s later writing style, with its essential use of self-deconstructing thought-experiments, different voices, genuine questions, and so forth. Wittgenstein wrote in a style so as to reduce the chance that he would be interpreted, as his early work had been, as putting forward assertions.)

 This sense of mind, as too deeply different from the corporeal to be adequately represented by the term, “incorporeal process”, combined with the sense of mind found in thinking of the mind as through and through embodied, together make up our concept of mind, and indeed of consciousness. 16  See Palmer’s Concept and object1988, and the essays by Conant and Witherspoon in The New Wittgenstein (Crary & Read 2000) and in Wittgenstein in America (McCarthy & Stidd 2001) on Ryle’s ‘Carnapian’ tendency to use the concept of ‘category mistake’ in ineffaciously and unpersuasively seeking to present perspicuously ontological differences. The term ‘category mistake’ tends to lead one systematically to under-estimate the difference between the two terms/‘things’ one is comparing, the categories being ‘mistaken’ for each other. ‘Mythic categorial nonsense’ might be a more apposite term than ‘category mistake’ for averting this risk. But any term alone is not enough; the actual work of averting philosophical delusion in oneself and in others always remains to be done, after one has picked the terms least likely in any given context to confuse or delude one in the course of that therapeutic work. Here is an example of where the concept of a “category mistake” or “conceptual transgression” is leading a would-be Wittgensteinian philosopher astray: “the problem rests on the illicit use of finite-bound concepts for infinite series” (Shanker 1987, p.183). 17  For instance, the deeper-than-deep difference between finite and infinite is at the root of all the main confusions unearthed by Read (2008a, 2002). 15

Finite and Infinite: On Not Making ‘Them’ Different Enough

313

In cases like those we have discussed, it is not as if both A ‘exists’ and B ‘exists’ – and the differences lies in ‘A’ and ‘B’. It is, as we might put it, the ‘exists’ which has a different sense. The difference between ‘finite’ and ‘infinite’ is an ontological not (just) an ontic difference. It is not a difference between different ‘kinds of things’ but between different ‘modes of existing’. (This still tends inexorably to make the difference sound too slight… This is what searching for philosophical liberation is like. Two steps forward, one step back. The distinction ontic vs. ontological can be misread in this same fashion!: It can so easily sound as if two different kinds of things in the Universe are being compared…) The danger is to give in to the tendency to ‘thingify’ (or ‘reify’). For example, within linguistics it might seem sensible to proceed, following mathematics, by saying: Let’s let X: = set of all possible sentences of our language. Now clearly X is infinite. The puzzling question: How are we able to ‘speak’ X? Example: Sense can be made of the point by recasting it as follows: the infinite totality of sentences of any given speaker’s language can be so permuted, or mapped onto itself, that (a) the totality of the speaker’s dispositions to the verbal behavior remains invariant, and yet (b) the mapping is no mere correlation of sentences with equivalent sentences, in any plausible sense of equivalence however loose. (Quine 1960, p.27)

Here Quine moves from our ability to speak a language to a statement about ‘the infinite totality of sentences’, thereby ‘thingifying’ an ability into a set. In this respect, Quine and Chomsky, supposed implacable philosophic foes, are of a piece. The first move in the conjuring trick, the move that generates a pseudo-problem, is to ‘create’ a thing: X.18 If pressed to express our point positively, we think that the ‘product’  – ‘process’ distinction could, if taken very cautiously, best express the difference that we are tying to capture – at least for cases such as ‘finite’ – ‘infinite’ and ‘body’ – ‘mind’. But any such positive presentation, as explained above, is liable to be misunderstood – to be turned back into a categorial or ontic distinction only.

4 An Entity or An Ability? The tendency to speak of the ‘infinite’ leads one toward conceptualising infinity as an entity. In contrast, we would want to emphasize that ‘infinity’ is very often not an expression of a ‘thing’ but of an ‘ability’ (to be able to do something). Consider the example of learning how to drive a car. Although you might be taking your lessons in Liverpool, you are not learning how to drive a car in Liverpool (or even: only in those streets where the driving lessons took place) but you are learning how to drive a car pretty much anywhere (with possible exceptions, e.g., a very steep hill or a country where they drive on the other side of the road––but even then you can learn the latter at least without great difficulty). There is no ‘inference’ 18

 Compare PI 36.

314

R. Read and C. Greiffenhagen

or ‘extension’ step from the-streets-where-I-learned-to-drive to streets-I-have-­ never-been-to, where the former is supposed to be a finite set and the latter a (potentially?) infinite set. No, learning how to drive a car is learning how to drive ‘anywhere’. (Both of these are exhibited in the sign for infinity: ∞. On the one hand, by giving a symbol to ‘infinity’ it is easy to think that it therefore is an entity, a thing. On the other hand, the form of the symbols points to the continuous and dynamic nature of the concept: you can go on, and on (inside the sign).) Within philosophy the mismove we just analysed is a common mistake. For example, within linguistics it is easy to contrast two sets: the sets of ‘examples’ that a language-speaker has heard so far and the (imagined) set of possible sentences that a language-speaker could utter: again, the former is seen as finite, the latter as infinite.19 Firstly, the setup of this (pseudo-)problem commits the same mistake as in the example of learning how to drive a car, i.e., it mischaracterises what it is to acquire an ability or competence (here: speaking a language): there is no ‘extension’ or ‘inference’ involved. Secondly, the setup of this problem is a strong example of ‘theoreticism’, here: a focus solely on syntactic properties of a language, neglecting any semantic or pragmatic ‘restrictions’. Although there are an enormous number of meaningful sentences that a speaker might utter (and surely more sentences than any speaker will ever utter in her or his lifetime), that does not imply that the number is ‘unbounded’ (although it is not possible to give an exact bound). Linguistics overemphasize the possibility of creating new sentences. We do not of course deny this possibility, but want to argue for some caution on how to put the extent and nature of it. Without such caution, one ends up unnecessarily committing oneself to philosophical absurdities, in the course of genuine Linguistics work. A similar example of such over-exaggeration is the statement sometimes found within sociology and anthropology that there are ‘infinitely many contexts’ for any action or utterance. Again, the analyst (correctly) wants to emphasize that any action or utterance derives its meaning from (and gives meaning to) the context in which it takes place. However, in a next step, disregarding all practical constraints (and very often all common sense), this occasional difference is turned into an ‘in principle’ one, where there are infinitely many (possible) contexts.

5 Different Notions of ‘Infinity’ Wittgenstein remarks that “If you speak of the concept ‘infinity’, you must remember that this word has many different meanings and bear in mind which one we are going to speak of at this particular moment.” (Wittgenstein, PR, p.304)  It is worth noting that, actually, this infinitude of sentences only even appears to be generated if we stick to a syntactic (or possibly a pragmatic) version of ‘sentence’. If, by contrast, one defines ‘sentence’ semantically, as something that actually means something (rather than just a string of nothing), then eventually one comes back to sentences one already encountered. 19

Finite and Infinite: On Not Making ‘Them’ Different Enough

315

Let us then introduce a very crude, but hopefully helpful, initial tripartite distinction: (a) An ordinary common-sense use of ‘infinite’: “The sky seemed to stretch away to infinity”, or “The infinite ocean”, or “An infinite array of possibilities suddenly opened up before her.” (b) A mathematical use of ‘infinite’. (c) A philosophical or academic use of ‘infinite’: philosophy; theoretical linguistics; much mathematical ‘prose’. The common-sense notion usually means either ‘very, very big’ (‘inconceivably big’) or ‘you can go on and on’ (the latter being important, somewhat distinct, and sometimes closer to mathematical usage – see e.g. Wittgenstein’s discussion of this in LFM). “The infinite” here, crucially, opens one up to something verb-al: you can always add one more. This is something which (if you are alive) you can do. This ties one nicely back toward the ‘ability-concept’ version of infinity, already discussed above. The mathematical notion usually is ‘without bounds’ or ‘you can always do X’. There is a sense in which this, (b), is a subset of (a) (see PI 124). There need be no problem with (b); there need be no problem with Cantor’s technique, for instance. The problem comes only in the way it is conceptualised or handled – i.e. in (c). Two important questions: 1. How much does (b) or (c) exhibit (a)? 2. How much do developments in (b) have an impact on either (a) or (c)? With respect to the second question, mathematicians have in the last one hundred and fifty years greatly extended the notion of ‘infinite’. They have done so, partly, by extending concepts taken from the ‘finite’ (e.g., ‘size’, ‘number’, ‘addition’) to ‘infinity’. There is nothing wrong with that. However, how much have these developments actually contributed to our understanding of (a) or (b)? In our view, there is nothing ‘wrong’ with either the common-sense notion of infinity, nor with the mathematical treatment of infinity (within mathematics). However, problems do arise in philosophy (of language, mind, and mathematics).

6 Everyday or Philosophical Use? the number of swans we have seen, or are ever likely to see, is an infinitesimally small sub-­ set of the infinite set of swans (Collins 1985, p.26)

What are we to make of such a statement? It could simply be seen as an (everyday) metaphorical way of speaking (i.e., ‘infinite’ as ‘very, very big’). However, that is not what Collins is doing here (or at least: trying to). Collins is pursuing a philosophical (sceptical) project, where the infinite set of swans is used in the argument that we can never definitely know whether all swans are white (A first pass at a Wittgensteinian response, of course, would be to point out that ‘All swans are white’ is perhaps more a grammatical than an empirical statement). Note that, strictly

316

R. Read and C. Greiffenhagen

speaking, there is only a finite number of swans on earth (although it might be difficult to specify the exact number). In other words, the number of swans is indefinite, rather than infinite. Put perhaps Collins is thinking about all the swans that could possibly exist. The reasoning might be: if the earth is not destroyed and swans not extinguished, then there is an infinite number of swans (as year at least one new swan is born for an ‘infinite number’ of years). But note that we have now moved from swans as part of everyday life to swans as part of theoretical – speculative -- philosophy. This move is very similar to the one within linguistics, where we are told that there are infinitely many sentences because we can, for example, always add a ‘very’ to the sentence “The sky is very big.” (“The sky is very, very big.”, “The sky is very, very, very big.”, and so on). In both cases the example is initially supposed to elucidate something about our lives, but subsequently removed from it (and placed closer to the idealised world of mathematics).

7 Mathematical Infinity Mathematics proceeds by extension. For example, given the integer numbers Z, it is possible to construct the rational numbers Q as fractions of integer numbers (and from the rational numbers we can construct the real numbers R). Furthermore, we can ‘embed’ the original numbers in the newly constructed ones. For example, from Z to Q we can represent a number z from Z as z/1 as an element in Q. This is often expressed in the following manner:

NÍZÍQÍR

For example, the number 3 is an element of all four sets. Well, this is not quite true. A more exact way of putting this would be: we can embed the number 3 from the natural numbers N in the other sets in such a way that it behaves ‘just like before’. However, strictly speaking we have four different 3s, namely 3N, 3Z, 3Q, and 3R. Now this particular case does not often lead to any confusion. However, in our view the temptation is to neglect the changes that occur when making mathematical developments; i.e., to think that the original is straightforwardly ‘part of’ the newly development. Although the same ‘words’ might occur, this does not mean that the ‘sense’ is the same. The sense has also changed. This seems to us what might have happened in the context of ‘infinity’.

8 Infinity as ‘Not Finite’ The origin of ‘infinity’ within mathematics was to express ‘without limit’ or ‘no upper bound’. Hence: “infinitely many natural numbers” (there is no greatest natural number) or “infinitely many sentences” (there is no limit on the number of

Finite and Infinite: On Not Making ‘Them’ Different Enough

317

sentences that we can form). Infinity here is defined negatively: does not have a limit, upper bound, end, etc. And Wittgenstein is absolutely right to observe that [t]he word ‘infinite’ has a different syntax from a number word (WVC 228). Where the nonsense starts is with our habit of thinking of a large number as closer to infinity than a small one. // As I’ve said, the infinite doesn’t rival the finite. The infinite is that whose essence is to exclude nothing finite. (PR, p.157)

‘Infinite’, in this sense, is not a number. ‘Infinitely many’ does not specify ‘how many’ there are, but it is an expression about something that we can not do (find an upper bound). Alternatively, it is an expression of something that we can always do: given any finite list (of numbers or sentence) we can find another item (number or sentence) which has not been part of that list. Infinity in this sense is not a number but a property (or a process) – however, it sometimes ‘looks’ like a number.

9 ‘Taming’ Infinity Mathematically For centuries mathematicians puzzled over the concept of ‘infinity’, but did not bother too much about it. However, this changed with advances in set theory in the nineteenth century. It is probably not too much an exaggeration to say that all the mathematical problems of the so-called ‘foundation crisis’ were results of not knowing how to deal with the notion of infinity (e.g., Russell’s paradox ‘the set of all sets that do not contain themselves’). all the remarkable problems and discoveries of the Foundations of Mathematics, the paradoxes of the theory of aggregates, Russell’s theory of types, with its axiom of reducibility, Cantor’s arithmetic of transfinite numbers, with its insoluble problems such as the ‘continuum problem’, the problems connected with functions in extension and the multiplicative axiom – all these merely express in one or another the well-known difficulties which arise when we attempt to treat an infinite process as completed. (Watson 1938, p.450)20

In other words, new developments in set theory led to some (local) inconstancies and inconsistencies. As mathematicians are ‘superstitious’ of inconsistencies, as Wittgenstein remarked, this meant that mathematicians had to ‘tame’ infinity. In other words, they had to find ways of handling it mathematically in such a way that it would not lead to contradictions. Thus mathematicians might ask: • What properties does ‘infinity’ have? • How can we define what we mean by ‘infinite’? • Given such a definition, can we find ways of distinguishing between different ‘kinds’ of infinity?  Compare also this remark: “If we pay close attention, we find that the literature of mathematics is replete with absurdities and inanities, which can usually be blamed on the infinite” (Hilbert 1925, p.370). Hilbert’s approach, different from Watson’s, nevertheless uses much the same idea. 20

R. Read and C. Greiffenhagen

318

• And if we had different kinds of infinity could we ‘calculate’ with them? In other words: could we treat them as ‘numbers’ – albeit as very strange ones? That’s what happened in set theory in the nineteenth century. Mathematicians came up with a definition for ‘infinity’. For example: “A set is infinite if there exists a one-­to-one correspondence between that set and a proper subset of itself.” And that is why Cantor’s diagonalisation argument had such an impact, as it introduced two different kinds of infinity: ‘countable’ (those that have ‘the same’ infinite as the natural numbers, e.g., the rational numbers) and ‘uncountable’ (those that were not ‘the same’ in this regard as the natural numbers, e.g., the real numbers). Note that saying this does not imply that ‘infinity’ behaves like a number. All we have done is enriched our ‘range’ of a property (‘infinite’); rather than being a binary property (either infinite or not) we have now different ‘values’ or ‘kinds’ (‘not infinite’, ‘countable’, and ‘uncountable’). It’s a bit like moving from talking only about ‘blue’ (either blue or not) to talking about different shades of blue (e.g., ‘baby blue’, ‘navy blue’, ‘light blue’, ‘dark blue’, ‘sapphire’). However, if we have different kinds of infinity, what are the ways that we might differentiate between them? And then the next fundamental idea: As we are talking about a property of sets, why not extend our notion of ‘number of elements in the set’ (cardinality)? For finite sets we have:

a, b, c, d = 4

Let’s therefore define the cardinality for the natural numbers N as À0:

N = 0,1, 2,3, 4,  = À 0



And now we can ‘calculate’ with these new cardinal ‘numbers’. However, we might get strange results. For example:

À0 = À0 + À0

What is important to note is that we have extended our notions of ‘number’ and ‘addition’ (and even ‘equals’) here. There is nothing wrong with that – we just need to be aware that we are now dealing with a very specialised notion of ‘addition’ and ‘number’. “ω+2 is not >ω+1 in the way n+1 is >n” (Shanker 1987, p.172) An infinite class is not a class which contains more members than a finite one, in the ordinary sense of the word ‘more’. If we say that an infinite number is greater that a finite one, that doesn’t make the two comparable, because in that statement the word ‘greater’ hasn't the same meaning as it has say in the proposition 5>4! (PG 464).

Finite and Infinite: On Not Making ‘Them’ Different Enough

319

And hence the ‘continuum hypothesis’ is not a question which can be ‘calculated’ – although it can be posed in terms of an equation: Let P(N) denote the ‘power set’ of the natural numbers (the set of all the subsets of N), then we want to know:

P  N ? = ? R



This is not a ‘calculation’. Neither was Cantor’s proof that

N

1

R



In order to demonstrate this ‘equation’, Cantor introduced a new proof method (namely diagonalisation) which changed the notion of ‘infinity’. We might see Gödel’s famous proof as extending the notion of mathematical ‘infinity’. Similarly, we would expect a new proof method to ‘answer’ the continuum hypothesis. The dangerous, deceptive thing about the idea: ‘The real numbers cannot be arranged in a series, or again ‘The set … is not denumerable’ is that it makes the determination of a concept—concept formation—look like a fact of nature. (FOM, II, §19, p.131)

So what has been shown? It has been shown that, mathematically speaking, the concept of the natural numbers is different from that of the real numbers. However, falling into the temptation of ‘thingifying’, this is then seen as showing that the ‘thing’ N (the natural numbers) is ‘smaller’ than the ‘thing’ R (the real numbers): If it were said: “Consideration of the diagonal procedures shews you that the concept ‘real number’ has much less analogy with the concept ‘cardinal number’ than we, being misled by certain analogies, are inclined to believe”, that would have a good and honest sense. But just the opposite happens: one pretends to compare the ‘set’ of real numbers in magnitude with that of cardinal numbers. The difference in kind between the two conceptions is represented, by a skew form of expression, as difference of extension. I believe, and hope, that a future generation will laugh at this hocus pocus. (FOM, II, §22, p.132)

Again, the problematic step is in a sense to move from a process (we can/cannot do this) to a property (the cardinality of a set). In sum: mathematics has ‘tamed’ infinity’. In a sense, it has successfully ‘thingified’ or ‘onticised’ infinity: it has turned the ontological difference between ‘finite’ and ‘infinite’ into an ontic one. And that is okay too––certainly, as long as we stay within pure mathematics. However, within other fields (linguistics, theology, common sense) we must take great care to remember that the difference between ‘finite’ and ‘infinite’ is an ontological, not an ontic, one. The mathematical taming works so long as we recognise that the mathematics of infinity (e.g. adding infinity to infinity and ending up with infinity) is ‘a language-game’, a practice, not an insight into a new realm / a paradise that we can now encounter. That one can calculate with ‘infinities’ is fine  – and proves/enables nothing about comparing such infinities with finities. Within mathematics, infinity has been tamed into something ontic. But only within mathematics. Otherwise – elsewhere – infinity remains ontologically different from finity.

R. Read and C. Greiffenhagen

320

10 Treating Infinity as a Number Mathematicians have fond ways to treat infinity as a number – but they have not thereby found new numbers (of the old kind), but rather they have extended the concept of number. “Consideration of the diagonal procedures shews you that the concept ‘real number’ has much less analogy with the concept ‘cardinal number’ than we, being misled by certain analogies, are inclined to believe”, that would have a good and honest sense. But just the opposite happens: one pretends to compare the ‘set’ of real numbers in magnitude with that of cardinal numbers. The difference in kind between the two conceptions is represented, by a skew form of expression, as difference of extension. I believe, and hope, that a future generation will laugh at this hocus pocus. (FOM, II, §22, p.132)

Problems arise if this confusion is carried over into philosophy. For example: we have certainly narrowed the gap between three and infinity (Knuth 1976, p.1235)

Knuth is talking about technological developments in the last centuries which have had implications for the size of numbers that can be used in practical applications. It is possible to work with extremely large numbers (through the aid of computers). For example, it is possible to find greater prime numbers or expand the decimal extension of pi than ever before. So we might want to say that we can today deal with larger numbers than a century ago. However, in our view, the difference between any finite number and ‘infinity’ has not been altered in the slightest – we have not got one iota closer to ‘the infinite’ – because the difference is not one of ‘size’ but a conceptual difference. Where the nonsense starts is with our habit of thinking of a large number as closer to infinity than a small one. (Wittgenstein, PR, p.157)

This confusion seems to originate within mathematics. Even non-mathematical readers might have encountered something like the following:

    23423333122    2  1  0

In this symbolic expression it looks like −∞ is the greatest negative number – but this is only a metaphorical way of speaking (or a highly technical one, within mathematical set theory). This way of speaking carries problems when used within philosophy or sociology. In sum, one needs to question whether any intelligible use has anywhere been set out for the notion of ‘infinite totality’, for the idea of an ‘infinitely large’ number: ‘actual infinity’, in the numbers.21  In connection with the traditional concepts of ‘potential’ and ‘actual’ infinity, we might express our point as follows: The standard view is that ‘actual’ infinity would be (a real) infinity. But an ‘actual’ infinity is a would-be thing. And infinity is entirely mis-identified as some kind of a thing. We would thus be inclined to reverse the polarity; a broadly Wittgensteinian approach should see the notion of potential infinity as coming closer to what one wanted to index when one spoke of infinity. Infinity is, as it were: there can always (potentially) be more. This could be seen as a radicalisation of a thought already present in Aristotle, and obscured in the prose of Hilbert, and at times of Cantor. 21

Finite and Infinite: On Not Making ‘Them’ Different Enough

321

11 ‘Getting Inspiration’ and ‘Looking Back’ It seems clear that mathematics originates in what Husserl calls the Lebenswelt (the life-world). However, it seems equally clear that mathematics subsequently gets in a sense quite detached from the Lebenswelt – and it is important to keep this detachment (perhaps most explicitly and dramatically expressed in Hilbert’s view of mathematics as ‘mechanical manipulation of meaningless symbols’) in mind. In other words, although there might be ‘traces’ of the Lebenswelt in mathematics, it is equally (if not more) important to remember that mathematics has ‘moved on’ and left its origins behind. With respect to the concept of ‘infinity’ we might want to say that mathematical infinity took the common-sense notion of infinity as its starting point, but developed it drastically and dramatically. There is of course nothing wrong with doing this, as long as we stay within mathematics. However, as soon as we leave mathematics proper, it is easy to forget how far the ‘new’ mathematical concepts and the ‘original’ concept have drifted apart. It is this issue that we think is the heart of the confusion of philosophers who talk about ‘infinity’ from the perspective of mathematical set theory. When ‘looking back’ at the original concept these philosophers see the origins but overlook the separation. In other words, they make the difference not big enough. Modern mathematical ‘infinity’ has almost nothing to do with our intuitive concept of ‘infinity’. Mathematics has, as we point out above, ‘tamed’ infinity. It has changed the concept in such a way that it can work with it. This might lead philosophers to the following reasoning: ordinary people cannot grasp the infinite – but mathematicians can grasp it.22 However, that is to neglect that we are operating with two different notions of ‘infinity’ here. Mathematics clearly is a magnificent achievement, which in a certain sense is ‘above’ the Lebenswelt. However, speaking in this way might tempt one to think that we are now on a ‘higher plane’ from which we have a clearer perspective of what is on the ground, i.e., when looking back. However, in our view this view is mythically mistaken. Looking at the Lebenswelt only through the eyes of mathematics is mistaking the map for the territory. For Wittgenstein, the point of philosophy is not to climb up some ladder in order to get a clearer picture of what is ‘down there’, but it is to stay on the ground and resist the temptation to fly into the sky. Not a ‘view from above’ but a ‘view from the street’ should be aim of philosophy. Or, put different, the ‘view from above’ and  And some people really do think this: “Infinite totalities cannot depend on us for their existence; if they exist, they must exist independently of our states of mind, of our thoughts, and of the course of our experience. [...] mathematicians seem able to direct their minds toward, to think cogently about and accumulate compelling truths about, infinite mathematical entities. [...] there seems to be a parity of the following sort between the existence of infinite mathematical entities and states of affairs in the world: they exist independently of us, and this means at least that thoughts ‘about them’ are true or false independently of our states of mind and of what we happen to know” (Tragesser 1984, p.38). 22

322

R. Read and C. Greiffenhagen

the ‘view from the street’ is not ‘just’ a change in perspective, they constitute very different ideas about what a ‘view’ is.23

12 Conclusion: Making Different Without Mystifying The general problem that we have touched upon here is the difficulty of simultaneously emphasising differences while not falling into the temptation to mystify (or, alternatively: to de-mystify without levelling). ‘Infinity’ is nothing mysterious. The common-sense notion might be not as clear and definite as, say, the notion of ‘dog’, but we nevertheless do have an understanding of infinity: sometimes ‘very, very big’, sometimes ‘without limits’, sometimes ‘I can go on and on’. Neither is the notion of mathematical ‘infinity’ mysterious. It is difficult, in the sense that some considerable and tough training is involved – but that is also true for, say, circus acrobatics, top-notch jazz improvisation, or walking in very high heels. A parallel argument could perhaps we made with respect to ‘common sense’ and ‘science’. Many academics, especially in the social studies of science (SSS), sociology of scientific knowledge (SSK), or science and technology studies (STS), have focussed on de-mystifying science (i.e., pointing out that scientists are human too). Perhaps the paradigmatic example here is Latour and Woolgar’s (1979) Laboratory Life, where they attempt to show “the idiosyncratic, local, heterogeneous, contextual, and multifaceted character of scientific practices” (p.152). However, academics involved in de-mystifying science often give in to the temptation of saying that science is ‘just’ common sense, i.e., of levelling science and common sense, of making them too similar. Again, Latour and Woolgar: “we have suggested that the mysterious thought processes employed by scientists in their setting is not strikingly different from those techniques employed to muddle through in daily life encounters” (p.166). However, we would argue that they are (also) strikingly different: for one thing, most people from the street (including, we suspect, Latour and Woolgar) could simply not do what the scientists in the laboratory are doing, at least without changing their whole lives.24 The point that we are trying to make is that it is important to take mathematics and science seriously: they are ‘special’ (but so are good art, and spiritual religion). This does not mean that they are ‘higher’ or more ‘foundational’. In particular, they are not a better version of common sense. Mathematics can in a way be read as a branch of common-sense. And in another way as quite deeply different to common-­ sense. In neither sense is it a better version of common-sense. And its being in a way

 For fleshing out of these metaphors, see Hutchinson & Read 2008.  This has to do with the false dichotomy in Latour and Woolgar between ‘Science is superhuman’ and ‘Science is banal’. Kuhn’s account in this ballpark is more helpful: see Sharrock and Read’s Kuhn (2002) for discussion of the importance of science being hard, and involving lots of work, and of the neglect of this important fact in most philosophy of science, except in Kuhn and some of those he has influenced. 23 24

Finite and Infinite: On Not Making ‘Them’ Different Enough

323

deeply different to common sense need not make it mysterious  – or philosophical.25 With respect to infinity: mathematical infinity is a highly developed, original concept. However, it does not constitute a better understanding of our common-­ sense criteria of infinity. Because these are just different. In terms of our threefold distinction between ‘uses’ of infinity, earlier: one can have (a) and (b). Our advice is not to go from (b) unknowingly to (c) – to philosophy, to theory of mind, etc. Chomsky, Quine, Penrose etc. aren’t helped by their mathematical knowledge. In the midst of this, they fall into the trap of failing to take sufficiently seriously the difference between (what one means by – or, sometimes, merely wants to mean by) “finite” and “infinite”. The kind of difference that Wittgenstein meant above all to teach. To teach (of course) not as a thesis, but as a non-assertoric ‘liberatory’ (broadly soteriological) move. To free one from the intellectual captivity that is inherent in most philosophical uses of the finite vs infinite distinction, and (at the ‘meta’ level) to free one from the compulsion to make any systematic or dogmatic claim. And here finally we come directly to Siderits’s Nāgārjuna.26 For Nāgārjuna, like Wittgenstein, is in a small minority in the history of philosophy, a minority that we wish to champion, in refusing, as Siderits and O’Brien put it, to “[construct] a system or [defend] a thesis”.27 Nāgārjuna uses the concept of infinity to question standard metaphysical assumptions but without any such constructive or thesis-like intention; to the contrary. Our approach has been to utilise Wittgensteinian thinking to undermine hegemonic attractive assumptions about the nature of ‘infinity’. Our Conclusion includes the noting that Nāgārjuna’s critical thinking on infinity, certainly as interpreted by Siderits, is a great predecessor to Wittgenstein’s,28 a predecessor neglected in most Western philosophy.  On the separation of maths and philosophy, see our paper on Godel, forthcoming.  Siderits & O’Brien 1976. This paper is of particular interest (in the context of the festschrift for Siderits which the present paper is part) because of its being an early piece of Siderits’s that doesn’t (unlike well-known instances of his more recent thinking) mention Wittgenstein but that is specifically about the topic of our paper (in a manner that we find obviously Wittgensteinian): namely, infinity (as a topic within which to clarify the nature of philosophy). 27  The fuller quotation, from p.288 of Siderits and O’Brien (1976), reads: “Nāgārjuna’s chief task in MMK is to provide a philosophical rationale for the notion of sunyataor “emptiness,” which is the key term in the Prajfiparamita Sutras, the earliest Mahayana literature. …To this end Nāgārjuna constructs a dialectic which he considers capable of reducing the metaphysical theories of his opponents (chiefly Sarvastivada, Sramkhya,and Nyaya) either to contradiction or to a conclusion which is unacceptable to the opponent. Unlike Zeno, however, Nāgārjuna is not refuting the theories of his opponents simply as a negative proof of his own thesis: Nāgārjuna has no thesis to defend-at least not at the object-level of analysis where metaphysical theories compete with one another. Instead his dialectic constitutes a meta-level critique of all the metaphysical theses expounded by his contemporaries.” (Italics ours.) The anticipation of Wittgenstein methodologically here is striking. Cf. also p.297 of Siderits and O’Brien. 28  Cf. also p.298 of Siderits and O’Brien; their exegesis according to which for Nāgārjuna “metaphysics is a fundamentally misguided under- taking” is strikingly anticipative of Wittgenstein’s approach. (For a thorough comparison of Buddhist methodology in philosophy to Wittgensteinian methodology, see Read 2012b, Chap. 8.) 25 26

324

R. Read and C. Greiffenhagen

References Ayer, A.J. 1932. Language, Truth and Logic. London: Penguin. Collins, H.M. 1985. Changing Order: Replication and Induction in Scientific Practice. London Beverly Hills: Sage Publications. Coulter, J. 1991. Is the `New Sentence Problem’ a Genuine Problem? Theory & Psychology 1 (3): 317–336. Crary, A., and R. Read, eds. 2000. The New Wittgenstein. London: Routledge. Davidson, D. 1967. Truth and Meaning. Synthese 17: 304–323. Hilbert, D. 1925. On the Infinite. Hutchinson, P., and R.  Read. 2008. Toward a Perspicuous Presentation of “Perspicuous Presentation”. Philosophical Investigations 31 (2): 141–160. Hutchinson, P., R. Read, and W. Sharrock. 2008. There Is No Such Thing as a Social Science: In Defence of Peter Winch. Aldershot: Ashgate. Knuth, D.E. 1976. Mathematics and Computer Science: Coping with Finiteness. Science 194 (4271): 1235–1242. Latour, B., and S. Woolgar. 1979. Laboratory Life: The Construction of Scientific Facts. Princeton: Princeton University Press. McCarthy, T., and S. Stidd, eds. 2001. Wittgenstein in America. Oxford: Oxford University Press. Niles, I. 1992. Wittgenstein and Infinite Linguistic Competence. Midwest Studies in Philosophy 17 (1): 193–213. Palmer, A. 1988. Concept and Object: The Unity of the Proposition in Logic and Psychology. London/New York: Routledge. Quine, W.V.O. 1960. World and Object. Cambridge, MA: The MIT Press. Read, R. 2001a. What Does ‘Signify’ Signify?: A Response to Gillett. Philosophical Psychology 14 (4): 499–514. ———. 2001b. On Approaching Schizophrenia Through Wittgenstein. Philosophical Psychology 14 (4): 449–475. ———. 2002. Is ‘What Is time?’ a Good Question to Ask? Philosophy 77 (2): 193–209. ———. 2008a. Applying Wittgenstein. London/New York: Continuum. ———. 2008b. The ‘Hard’ Problem of Consciousness is Continually Reproduced and Made Harder by All Attempts to Solve It. Theory, Culture & Society 25 (2): 51–86. ———. 2012. A Wittgensteinian Way with Paradoxes. Plymouth: Lexington Books. Read, R., and R. Deans. 2003. “Nothing is Shown”: A ‘Resolute’ Response to Mounce, Emiliani, Koethe and Vilhauer. Philosophical Investigations 26 (3): 239–268. Ryle, G. 1949. The Concept of Mind. Chicago: University of Chicago Press. Sass, L. 1992. Heidegger, Schizophrenia and the Ontological Difference. Philosophical Psychology 5 (2): 109–132. Shanker, S. 1987. Wittgenstein and the Turning Point in the Philosophy of Mathematics. Albany: State of New York Press. Sharrock, W., and R. Read. 2002. Kuhn. Oxford: Polity. Siderits, M., and J.D. O’Brien. 1976. Zeno and Nāgārjuna on Motion. Philosophy East and West 26 (3): 281–299. Tragesser, R.S. 1984. Husserl and Realism in Logic and Mathematics. Cambridge: Cambridge University Press. Watson, A.G.D. 1938. Mathematics and Its Foundations. Mind 47 (188): 440–451.

Part III

Epistemology

Mark Siderits on Anumāna Shoryu Katsura

1 Introduction The following is a comment on “Deductive, Inductive, Both, or Neither?” in Siderits (2016: 120–137) from the viewpoint of a historian of Indian philosophy trained in philological discipline. Generally speaking, I accept an overall evaluation of anumāna by Siderits. According to him, anumāna is not an argument as he defines1 but an inference2; it is neither deductive nor inductive; nor is it a hybrid of both inductive and deductive, though he says: “there is an epistemic process that might justly be called inductive in nature. When we go through this process of induction, we bring together evidence that we take to justify our projecting from some observation—e.g. that there is smoke on the mountain—to a conclusion we wish to draw—that there is fire on the mountain”3 and “it (i.e. anumāna) instantiates a deductively valid argument pattern, modus ponens.”4 By ‘a deductively valid argument pattern’ he means the following formulation: “There is fire on the mountain; because there is smoke on the mountain; wherever smoke is, fire is.”5  Siderits (2016: 122): “By an argument I shall here mean the sort of thing that is studied in formal logic. An argument will then be a kind of formal structure: it will consist of a set of propositions (which are a type of abstract object) with certain logical relationships among them.” 2  loc. cit. “an inference, which is a particular epistemic event carried out by some particular cognizer at a determinate place and time. More specifically it is a cognitive process wherein the cognizer aims at the production of a veridical cognition through the occurrence of other cognitions.” 3  Siderits (2016: 132). 4  loc. cit. 5  Siderits (2016: 126). 1

S. Katsura (*) Ryukoku University, Research Center for World Buddhist Cultures, Kyoto, Japan e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_17

327

328

S. Katsura

However, I cannot accept the specific aim of his article; namely, he says at the beginning: “Here I shall focus on the Nyāya treatment of anumāna before and after Dignāga. I shall claim that there is no such transition to be found there, that later authors are simply spelling out what was already implicit in the earliest stratum of Indian thought about inferential reasoning. I shall try to support this claim by investigating the nature of the argumentation involved in the anumāna.”6

By the phrase ‘such transition’ he means “a transition in Indian logic from a conception of inference as aiming at mere probability to a conception that sees it as establishing certainty.”7 In other words, he wants to argue that Dignāga (c. 480–540) did not really change the nature of Indian logic by introducing the theory of pervasion (vyāpti) because even before Dignāga Nyāya logicians implicitly assumed the idea of pervasion of the reason (hetu) by the target property (sādhya-dharma). This is a challenge to my understanding of the history of Indian logic because I once tried to demonstrate, by collecting evidences from the classical Indian texts on debate (vāda) and epistemology & logic (pramāṇa) from the Caraka-saṃhitā up to Dharmakīrti’s, that it is Dignāga in his Pramāṇasamuccaya-vṛtti Chapters 2 and 4 who first established the concept of pervasion of the inferential mark (liṅga) by its target (liṅgin) or that of the proving property (sādhana-dharma) by the target property.8 Therefore, I feel responsible to examine his arguments carefully and to prove that Nyāya logicians before Dignāga did not possess the idea of pervasion. Nevertheless, it is to be noted at the outset that I do not think Dignāga’s introduction of the theory of pervasion drastically changed the ‘inductive’ nature of Indian logic, though the theory became widely accepted by the post-Dignāga Indian logicians. Siderits begins his analysis of the form of anumāna in the following manner9: “I turn now to the form of the anumāna, and specifically the role that a statement of pervasion (vyāpti) plays in its analysis. In the oldest stratum of Nyāya discussions of the topic, the anumāna is described as having the form: p is S Because p is H Like in the sapakṣa Unlike in the vipakṣa.  (1) For instance: There is fire on the mountain Because there is smoke on the mountain Like in the kitchen Unlike on the lake.”   (2)

I do not understand what he means by the expression ‘the oldest stratum of Nyāya discussions of the topic.’ The Nyāyasūtra (c. 150), the earliest manual of Nyāya

 Siderits (2016: 121).  Siderits (2016: 120). 8  Katsura (1986a) and its English summary (1986b). 9  Siderits (2016: 124). 6 7

Mark Siderits on Anumāna

329

school, enumerates the five members of a proof formulation, namely, ‘assertion’ (pratijñā), ‘reason’ (hetu), ‘exemplification/corroboration’ (udāharaṇa), ‘application’ (upanaya) and ‘conclusion’ (nigamana) with their respective definitions; however, it does not present any proof or refutation in that or in any other formal way, though it contains many debates on various topics such as ātman, momentariness and God. As an example of the early proof formulations, I extract the following proof (sthāpanā) from the Caraka-saṃhitā (c. 100): [assertion] puruṣa (i.e. ātman) is eternal (nityaḥ puruṣaḥ). [reason] Because it is unproduced (akṛtakatvāt). [example (dṛṣṭānta)] Like the space (yathākāśaṃ). [application] And as the space is unproduced as well as eternal, so is puruṣa (yathā cākṛtakam ākāśaṃ tac ca nityaṃ tathā puruṣaḥ). [conclusion] Therefore, it is eternal (tasmān nityaḥ).10  (3)

Regarding the Carakasaṃhitā’s formulation of anumāna, Siderits says: “The formulation of anumāna found in the early Carakasaṃhitā does not include citation of an instance of vipakṣa. This, I would claim, is because here inference is understood in the context of public debate, where it is the job of the opponent to raise possible counterexamples to the alleged pervasion, and the job of the proponent to then disarm them. Once the anumāna comes to be conceptualized as a means whereby an individual cognizer can come to have reliable indirect beliefs, this process of seeking possible counter-examples was internalized, as in the earliest Nyāya formulations.”11

The Carakasaṃhitā actually gives the following counter-proof (pratiṣṭhāpanā) that is directly opposed to the proof (3): [assertion] A puruṣa (i.e. a man) is non-eternal (anityaḥ puruṣaḥ). [reason] Because he is perceived by the senses (aindriyakatvāt). [example] Like a pot (yathā ghaṭaḥ). [application] And as a pot is perceived by the senses as well as non-eternal, so is this [man] (yathā ghaṭa aindriyakaḥ sa cānityas tathā cāyam). [conclusion] Therefore, he is non-eternal (tasmād anityaḥ).12  (4)

Since the counter-proof (4) does not share the same reason (i.e. ‘the property of being unproduced’) with the proof (3), the example of the counter-proof (4) ‘a pot’, being produced, cannot be the counter-example to the alleged pervasion of the proof (3), viz. “Whatever is unproduced is eternal”. Therefore, the opponent in the Carakasaṃhitā is not doing the job of raising possible counter-examples. As I shall show immediately, “an instance in vipakṣa” or a dissimilar example (vaidharmya-­ dṛṣṭānta) in Indian proof formulation has nothing to do with the search for the possible counter-examples to the pervasion.

 CS p. 267.  Siderits (2016: 124, fn. 83). 12  CS p. 267. 10 11

330

S. Katsura

2 Formulation of the Example Statement It is to be noted that in the above formulations (3) and (4) there is no equivalent for Siderits’ example statement: “Like in the sapakṣa; unlike in the vipakṣa”. As far as I know, there exists no statement of the example that reads: “Like in the sapakṣa; unlike in the vipakṣa.” The technical terms sapakṣa and vipakṣa did not exist among non-Buddhist philosophers probably before Uddyotakara (c. 550).13 It is Dignāga who first introduced those two terms in order to explain the three conditions (trairūpya) of a proper inferential mark or reason (liṅga/hetu).14 Siderits’ formulation “Like in the kitchen; unlike on the lake” may sound O.K. but it is again Dignāga who first insisted on the necessity of stating both the similar and the dissimilar examples (sādharmya- and vaidharmya-dṛṣṭānta) in a single proof.15 It is true that the Nyāyasūtra recognized the two ways of proving the target property, namely, through similarity (sādharmya) and dissimilarity (vaidharmya) between the reason and the example,16 but neither the Nyāyasūtra nor the Nyāyabhāṣya uses the technical terms ‘the similar and the dissimilar examples’. The author of the Nyāyabhāṣya, Vātsyāyana/Pakṣilasvāmin (c. fifth century), who seems to have been active just before Dignāga, gave the following pair of the five-membered proofs17: [assertion] Sound is non-eternal (anityaḥ śabdaḥ). [reason] Because it is of the nature of arising (utpattidharmakatvād). [exemplification (udāharaṇa)] The substance such as a plate that is of the nature of arising  is non-eternal (utpattidharmakaṃ sthālyādi dravyam anityam). [application] And sound is of the nature of arising in the similar way (tathā cotpattidharmakaḥ śabdaḥ). [conclusion] Therefore, being of the nature of arising, sound is non-eternal (tasmād utpattidharmakatvād anityaḥ śabdaḥ).  (5)

 In the Nyāyabhāṣya there is no occurrence of the terms sapakṣa and asapakṣa/vipakṣa; the term pakṣa is used only in the sense of ‘assertion/thesis’ in the pairs of pakṣa & pratipakṣa (a counterthesis) and svapakṣa (own thesis) & parapakṣa (the opponent’s thesis). 14  In trairūpya formulae before Dignāga sapakṣa and asapakṣa/vipakṣa are represented by *sajātīya (同類) and *vijātīya (異類), etc. See Katsura (1985). For Dignāga’s understanding of pakṣa, sapakṣa and vipakṣa, please see Katsura (2004b). 15  Katsura (2004a: 159ff). 16  Nyāyasūtra 1.1.34–37. See Dasti & Phillips (2017: 182–183). 17  NBh pp.  315–316 tatra sādharmyokte tāvad dhetau vākyam anityaḥ śabda iti pratijñā/ utpattidharmakatvād iti hetuḥ/ utpattidharmakaṃ sthālyādi dravyam anityam ity udāharaṇam/ tathā cotpattidharmakaḥ śabda ity upanayaḥ/ tasmād utpattidharmakatvād anityaḥ śabda iti nigamanam/ vaidharmyokte ‘py anityaḥ śabda utpattidharmakatvāt, anutpattidharmakam ātmādi dravyaṃ nityaṃ dṛṣṭam, na ca tathānutpattidharmakaḥ śabdaḥ tasmād utpattidharmakatvād anityaḥ śabda iti/ 13

Mark Siderits on Anumāna

331

[assertion] Sound is non-eternal (anityaḥ śabdaḥ). [reason] Because it is of the nature of arising (utpattidharmakatvād). [exemplification (udāharaṇa)] The substance such as ātman that is of the nature of non-arising is observed to be eternal (anutpattidharmakam ātmādi dravyaṃ nityaṃ dṛṣṭam). [application] But sound is not of the nature of non-arising in the similar way (na ca tathānutpattidharmakaḥ śabdaḥ). [conclusion] Therefore, being of the nature of arising, sound is non-eternal (tasmād utpattidharmakatvād anityaḥ śabdaḥ).  (6)

The above formulation (5) may be called a proof by similarity (sādharmya) and (6) a proof by dissimilarity (vaidharmya) of the reason with the example. When we compare the third member of (5) and (6) with that of the Carakasaṃhitā’s formulations (3) and (4), we notice that Vātsyāyana attributes both the proving and the target property (sādhana- and sādhya-dharma) to the actual examples such as a plate and ātman.18 He simply points to the mere coexistence (sāhacarya/anvaya) of the two properties in the examples. However, I do not see any hint to or a germ of the idea of pervasion in the third member of (5) and (6). It is most interesting that the third member of (6) contains the term dṛṣṭa (observed) as Dignāga’s statement of ‘example’ (dṛṣṭānta). It indicates that the statement of ‘exemplification’ is derived from observation and reveals ‘inductive’ aspect of Indian proof. Now, in order to justify the absence of an explicit statement of pervasion in his formulations (1) and (2), Siderits says: “There is, in other words, no explicit statement of pervasion included among the members of the anumāna. I claim, though, that the citing of examples drawn from the class of sapakṣa and vipakṣa is intended to serve as evidence for the existence of the appropriate relation of pervasion, so that a statement of pervasion may be seen as an implicit premiss in the overall structure. For consider the choice of lake as example of vipakṣa. The point is not that this is an instance that is devoid of both fire and smoke. The point is rather that the mist that rises from a lake at dawn might be mistaken for smoke, but is not smoke, so that the lake does not serve as a counter-example to the pervasion of smoke by fire. In other words, the lake is cited as evidence that a principled search for counter-examples has been made and has come up empty.”19

He regards the citation of the dissimilar example in the proof as evidence for the unsuccessful search for counter-examples to the pervasion; hence, he seems to consider that Naiyāyikas even before Dignāga implicitly assumed the idea of pervasion. I am sorry to say that Siderits has no historical ground. I have already shown that Naiyāyikas before Dignāga had no idea of sapakṣa and vipakṣa. There is no evidence that they presented both the similar and the dissimilar examples in a single proof formulation as in Siderits’ formulation (2). Vātsuyāyana’s formulations (5) and (6) are two independent ones and the ‘exemplification’ (udāharaṇa) of (6) does not show any sign of the search for the counter-examples to the pervasion, for the

 It is to be noted that Ernst Prets’ version of sthāpanā of CS attributes the two properties to a single example like NBh and it reads: nityaḥ puruṣaḥ, akṛtakatvāt, akṛtakam ākāśaṃ tac ca nityam, yathā cākṛtakam ākāśaṃ tathā puruṣaḥ, tasmān nityaḥ. See Prets (2010: 81). 19  Siderits (2016: 124). 18

332

S. Katsura

‘eternal’ substance ātman is never suspected to be ‘of the nature of arising’ (utpattidharmakatva) at least by non-Buddhist scholars. Among Naiyāyikas Uddyotakara, who was aware of the kind of the inference based on both the affirmative and negative concomitances (anvaya-vyatireka) between the reason (hetu) and the target property (sādhya-dharma),20 is expected to formulate the two examples in a single proof but he did not actually give such a formulation. As I shall discuss immediately, it is the very late Nyāya text such as the Tarkabhāṣā that has the concepts of sapakṣa and vipakṣa and presents the two examples in a single proof.21 Siderits gives as a typical instance of anumāna in the oldest Nyāya stratum “There is fire on the mountain because there is smoke on the mountain like in the kitchen (and) unlike on the lake”. The inference (anumāna) of fire from smoke is recorded as early as in the Caraka-saṃhitā22 but we cannot find among the old Nyāya texts the formulation with the two examples as mentioned by Siderits; it is again in the Tarkabhāṣā we find such a formulation. He says: “the mist that rises from a lake at dawn might be mistaken for smoke, but is not smoke, so that the lake does not serve as a counter-example to the pervasion of smoke by fire”; but as far as I checked, the idea of ‘mist’ being mistaken for smoke occurs only in the Buddhist text Nyāyapraveśa of Śaṅkarasvāmin (c. sixth century)23 but not in the Nyāya texts.

3 The Tarkabhāṣā on anumāna What Siderits considers to be anumāna in the oldest stratum of the Nyāya tradition, I believe, seems to come from the very late manual of Nyāya school such as the Tarkabhāṣā and the Tarkasaṃgraha. In order to demonstrate my point I would like to quote some passages from the Tarkabhāṣā. First of all, the late Naiyāyikas admit two kinds of anumāna, namely, anumāna for oneself (svārthānumāna) and anumāna for others (parārthānumāna)24; it is again Dignāga who first made such a distinction that became widely accepted by Indian philosophers as standard. Now I quote the interpretation of svārthānumāna from the Tarkabhāṣā: “svārthānumāna is the cause of one’s own cognition. To be more precise – Having for oneself grasped ‘pervasion’ (vyāpti) between smoke and fire by the qualified perception in the kitchen and elsewhere, one goes near a mountain and has a doubt if there is fire on the mountain; upon seeing a line of smoke that comes from the mountain to reach a cloud with-

 The other two kinds of inference are based on the affirmative or the negative concomitance only (kevalānvayin and kevalavyatirekin). 21  TBh p.  82: vādiprativādinoḥ saṃpratipattiviṣayo ‘rtho dṛṣṭāntaḥ | sa ca dvividhaḥ | ekaḥ sādharmyadṛṣṭāntaḥ | dhūmavattvahetor mahānasa iti | vaidharmyadṛṣṭāto yathā tasyaiva hetor mahāhrada iti | (N.B. I occasionally emend the Bombay edition with the help of other editions.) 22  CS p. 71: vahnir gūḍho dhūmena. 23  It occurs in the example of sandigdhāsiddha-hetu. NPra p. 142: bāṣpādibhāvena saṃdihyamāno bhūtasaṃghāto’gnisiddhāv upadiśyamānaḥ saṃdigdhāsiddhaḥ/ 24  TBh p. 39: anumānaṃ dvividham – svārthaṃ parārthaṃ ca | 20

Mark Siderits on Anumāna

333

out being disconnected from its root on the mountain, mental impression (saṃskāra) that is awakened by the perception of smoke makes one recall the pervasion that where there is smoke there is fire. After that one realizes that there is smoke on the mountain and from that one realizes that there is fire on the mountain. This is svārthānumāna.”25

[S]vārthānumāna represents the process of an inferential cognition that is initiated by the grasping of pervasion by perception in actual instances such as kitchen. It is not clear what Keśavamiśra means by ‘the qualified perception’ (viśiṣṭa-pratyakṣa); although he does not refer to ‘repeated observation’ (bhūyo-darśana) of smoke and fire in one place as Annaṃbhaṭṭa does in the Tarkasaṃgraha,26 he must have had a similar idea in this context. In any case, it is clear that the theory of pervasion is grounded on actual experiences or observations. Now, regarding parārthānumāna, Keśavamiśra says: “parārthānumāna, however, is a statement consisting of the five members that is composed by someone, who has inferred fire from smoke, in order to make the other realize [the existence of fire on the mountain]. For example, [Assertion] This mountain has fire. [Reason] Because it possesses the property of having smoke. [Exemplification] That which has smoke has fire, as e.g. kitchen. [Application] And this mountain is like that (i.e. has smoke). [Conclusion] Therefore, it is like that (i.e. has fire).   (7) The other too realizes the existence of fire [on the mountain] from the inferential mark (i.e. smoke) possessed of the five conditions27 [of a proper inferential mark] that

 TBh p. 39: svārthaṃ svapratipattihetuḥ | tathāhi, svayam eva mahānasādau viśiṣṭena pratyakṣeṇa dhūmāgnyor vyāptiṃ gṛhītvā parvatasamīpaṃ gatas tadgate cāgnau saṃdihānaḥ parvatavartinīm avicchinnamūlām abhraṃlihāṃ dhūmalekhāṃ paśyan dhūmadarśanodbuddha saṃskāro vyāptiṃ smarati yatra dhūmas tatrāgnir iti | tato ‘trāpi dhūmo ‘stīti pratipadyate tasmād atra parvate ‘gnir astīti svayam evaṃ pratipadyate tat svārthānumānam | Cf. TSaṃ p.  27: tatra svārthaṃ svānumitihetuḥ | tathā hi – svayam eva bhūyodarśanena ‘yatra dhūmas tatrāgniḥ’ iti mahānasādau vyāptiṃ gṛhītvā parvatasamīpaṃ gatas tadgate cāgnau saṃdihānaḥ parvate dhūmaṃ paśyan, vyāptiṃ smarati  – ‘yatra dhūmas tatrāgniḥ’ iti | tadanantaraṃ ‘vahnivyāpyadhūmavān ayaṃ parvataḥ’ iti jñānam utpadyate | ayam eva liṅgaparāmarśa ity ucyate | tasmāt ‘parvato vahnimān’ iti jñānam anumitir utpadyate | tad etat svārthānumānam | (Of the two kinds of anumāna, svārthānumāna is the cause of one’s own inferential knowldege. To be more precise – Having for oneself grasped ‘pervasion’ (vyāpti), “Where there is smoke, there is fire,” in the kitchen and elsewhere by repeated observation (bhūyodarśanena), one goes near a mountain and has a doubt if there is fire on that mountain; seeing smoke on the mountain, he recalls the pervasion “Where there is smoke, there is fire.” Immediately after that, there arises knowledge “This mountain has smoke that is pervaded by [the domain of] fire. This knowledge is called ‘reflection on an inferential mark’ (liṅga-parāmarśa). From that knowledge there arises an inferential knowledge “The mountain has fire.” This is svārthānumāna.) 26  Please see the above note. 27  TBh p.  42: pakṣadharmatvam, sapakṣe sattvam, vipakṣād vyāvṛttiḥ, abādhitaviṣayatvam, asatpratipakṣatvam. 25

334

S. Katsura

he has ­realized because of the above statement. Therefore, it is parārthānumāna (inference for others).”28

At last we come to what Siderits thinks a typical instance of anumāna in the oldest Nyāya stratum: “There is fire on the mountain because there is smoke on the mountain like in the kitchen (and) unlike on the lake”. Although the dissimilar example “Unlike on the lake” is not explicitly stated in the above formulation, it is implicitly assumed as I quote below. Keśavamiśra continues to explain parārthānumāna as follows: “In the above proof formulation, what is to be proved (sādhya) is the property of having fire on the mountain. The reason is the property of having smoke and it (i.e. the reason) has both the affirmative and the negative concomitances (anvaya-vyatirekin) because of the pervasion through the affirmative and the negative concomitances. To be more precise – ‘where there is the property of having smoke, there is the property of having fire like in the kitchen’ is the affirmative pervasion (anvaya-vyāpti) because we observe the existence of the affirmative concomitance of smoke and fire in the kitchen. Similarly, ‘where there is no fire, nor is there smoke like on the lake’ is the pervasion through the negative concomitance (vyatirekeṇa vyāptiḥ) because we observe the existence of the negative concomitance between fire and smoke on the lake.”29

Keśavamiśra is aware of the fact that the reason in the proof formulation (7), i.e., the property of having smoke, is supported by the two kinds of pervasion (viz. the affirmative and the negative pervasions) as well as by the two kinds of examples (viz. the similar and the dissimilar examples). He gives the same negative example “[Un]like on the lake” as Siderits did. In this connection, Keśavamiśra states that the order of smoke and fire in the negative pervasion is reversed from that in the affirmative pervasion. Although he quotes a verse from the Ślokavārttika of Kumārila (ca. sixth to seventh century) as his support, it is again Dignāga who first insisted to formulate the dissimilar example in that manner.30 Keśavamiśra says: “The negative pervasion is [to be presented] in this manner: the absence of the pervaded (vyāpya, e.g. smoke) in the affirmative pervasion is the pervader (vyāpaka) in this [negative

 TBh p.  40: yat tu kaścid dhūmād agnim anumāya paraṃ bodhayituṃ pañcāvayavaṃ vākyaṃ prayuṅkte tat parārthānumānam | tad yathā | parvato ‘yam agnimān | dhūmavattvāt | yo dhūmavān so ‘gnimān | yathā mahānasaḥ | tathā cāyaṃ | tasmāt tatheti | anena vākyena pratipāditāt pañcarūpopapannāl liṅgāt paro ‘py agniṃ pratipadyate | tenaitat parārthānumānam |  Cf. TSaṃ p.  27: yat tu svayaṃ dhūmād agnim anumāya parapratipattyarthaṃ pañcāvayavavākyaṃ prayuṅkte tat parārthānumānam | yathā ‘parvato vahnimān, dhūmavattvāt | yo yo dhūmavān sa sa vahnimān, yathā mahānasaḥ | tathā cāyam | tasmāt tathā’ iti | anena pratipāditāl liṅgāt paro ‘py agniṃ pratipadyate || 29  TBh p.  40: atra parvate ‘gnimattvaṃ sādhyam | dhūmavattvaṃ hetuḥ | sa cānvayavyatirekī | anvayena vyatirekeṇa ca vyāpteḥ | tathāhi yatra dhūmavattvaṃ tatrāgnimattvaṃ yathā mahānasa ity anvayavyāptiḥ | mahānase dhūmāgnyor anvayasadbhāvadarṡanāt | evaṃ yatrāgnir nāsti tatra dhūmo ‘pi nāsti yathā mahāhrada iti vyatirekeṇa vyāptiḥ | mahāhrade ‘gnidhūmavyatirekasya sadbhāvadarśanāt | 30  Pramāṇasamuccaya 4.2: sādhyenānugamo hetoḥ sādhyābhāve ca nāstitā |  khyāpyate yatra dṛṣṭāntaḥ sa sādharmyetaro dvidhā ||2|| Katsura (2004a) and (2016: 1243). 28

Mark Siderits on Anumāna

335

pervasion] and the absence of the pervader (e.g. fire) in the affirmative pervasion is the pervaded in this [negative pervasion]. Thus it is said: The pervaded-pervader relationship between the two absences (viz. the absence of the reason and that of the target property) is known to be reversed from the way it is admitted between the two things (viz. the reason and the target property). In the affirmative [pervasion] it is admitted that the reason is the pervaded and the target property is the pervader; in the other case (i.e. in the negative pervasion) the absence of the target property is the pervaded and the absence of the reason is the pervader. The pervaded is mentioned first and the pervader comes after that. The pervasion, being manifested in that way, becomes really clear.”31

Now Keśavamiśra explains the reason why the negative pervasion was not mentioned in the formulation (7) as follows: “Thus, in that way, there is the pervasion through both the affirmative and the negative concomitances in the case of such reason as the property of possessing smoke. In the statement [of the proof (7)] the affirmative pervasion alone is mentioned because the aim [of the proof] is achieved by the affirmative pervasion alone. Also because the affirmative concomitance is not oblique, it [alone] is mentioned. For, when the target is proved by the straight way, it is inappropriate to prove it by the oblique way. It is not because there is no negative pervasion [that the affirmative pervasion alone is mentioned in the formulation (7)]. Thus, in that way, the reason, i.e. the property of possessing smoke, has both the affirmative and the negative [pervasions].”32

Keśavamiśra seems to consider that it is not necessary to mention the negative pervasion in the case of smoke-fire inference because the affirmative pervasion alone can achieve the aim in the straight way. Here again I find the similarity with Dignāga’s approach to the example formulation; although he insists on stating both the similar and the dissimilar examples, he admits that it is not necessary to state the latter, if the former is well known.33 Finally Keśavamiśra presents the three-membered proof formulation that could have been extracted from any text of Buddhist logic.

 TBh p.  40: vyatirekavyāpter ayaṃ kramaḥ | anvayavyāptau yad vyāpyaṃ tadabhāvo ‘tra vyāpako, ‘nvayavyāptau yac ca vyāpakaṃ tadabhāvo ‘tra vyāpya iti | tad uktam – 31

vyāpyavyāpakabhāvo hi bhāvayor yādṛg iṣyate | tayor abhāvayos tasmād viparītaḥ pratīyate || (ŚV anumāna vv. 121 cd-122ab) anvaye sādhanaṃ vyāpyaṃ sādhyaṃ vyāpakam iṣyate | sādhyābhāvo ‘nyathā vyāpyo vyāpkaḥ sādhanātyayaḥ || vyāpyasya vacanaṃ pūrvaṃ vyāpakasya tataḥ param | evaṃ sphuṭīkṛtā vyāptiḥ sphuṭībhavati tattvataḥ || Regarding the order of the pervader and the pervaded, please see ŚV anumāna vv. 108  cd110ab and Katsura (1982). 32  TBh, p.  41: tad evaṃ dhūmavattve hetāv anvayena vyatirekeṇa ca vyāptir asti | yat tu vākye kevalānvayavyāptir eva pradarśyate tad ekenāpi caritārthatvāt | atrāpy anvayasyāvakratvāt pradarśanam | ṛjumārgeṇa siddhyato ‘rthasya vakreṇa sādhanāyogāt | na tu vyatirekavyāpter abhāvāt | tad evaṃ dhūmavattvaṃ hetur anvayavyatirekī | 33  Katsura (2004a: 168–169)

336

S. Katsura

“Similarly, other reasons such as the property of being produced, when non-eternity, etc., are to be proved, should be regarded as the reasons with both the affirmative and the negative concomitances. For example,    [Assertion] Sound is non-eternal.    [Reason] Because it is produced.    [Exemplification] Where there is the property of being produced, there is the property of being non-eternal like a pot; where there is no property of being non-eternal, there is no property of being produced like the sky.”34  (8)

So far I have shown that what Siderits thinks to be the form of the anumāna in the oldest stratum of Nyāya tradition possibly comes from the later manuals of Nyāya school such as the Trakabhāṣā that appear to be under the great influence of Dignāga’s theory of anumāna. As far as I have searched the available Nyāya texts before Dignāga, I failed to find the formulation of pervasion in the example statement. I do not think Siderits has any solid ground for claiming that, before Dignāga explicitly introduced the pervasion in the example statement, Naiyāyikas had an idea of pervasion, even though they did not state it. I hope that I had fed back a philologist’s expertise to the philosopher in order to change his grasp of the development of a certain logical concept in India.35 In the rest of this paper, I will briefly touch Buddhist concepts of anumāna.

4 Dignāga on sapakṣa, asapakṣa/vipakṣa and pakṣa With the help of the commentary by Jinendrabuddhi, we understand that Dignāga divided the world into two realms, viz., the observed or known (dṛṣṭa/vidita) realm and the unobserved or unknown (adṛṣṭa/avidita) realm, and that he further divided the former into two domains, viz., the similar domain (sapakṣa) and the dissimilar domain (asapakṣa/vipakṣa).36 The similarity and dissimilarity are determined from the perspective of the target property (sādhya-dharma) of a proof; for example, in the proof formulation (8) mentioned above, since ‘the property of being non-­eternal’ (anityatva) is the target property, the similar domain consists of non-eternal (anitya) things and the dissimilar domain consists of eternal (nitya) things. In this connection, it is to be noted that the similar domain (sapakṣa) and the dissimilar domain (asapakṣa/vipakṣa) should not be confused with another pair of the terms, namely, the similar example (sādharmya-dṛṣṭānta) and the dissimilar

 TBh p.  41: evam anye ‘py anityatvādau sādhye kṛtakatvādayo hetavo ‘nvayavyatirekiṇo draṣṭavyāḥ | yathā śabdo ‘nityaḥ, kṛtakatvād, ghaṭavad yatra kṛtakatvaṃ tatrānityatvaṃ, yatrānityatvābhāvas tatra kṛtakatvābhāvaḥ yathā gaganam | Supplied from The Tarkabhāṣā of Keśavamiśra, ed. by A.B. Gajendragadkar and R.D. Karmarkar, 1934, Poona, p. 29, fn. 1. 35  For more detailed analysis of Indian proof formulations, please see Katsura (forthcoming). 36  Dignāga uses the term asapakṣa in contrast with sapakṣa but his successors start using vipakṣa in place of asapaksa. Please see (Katsura 2004b). 34

Mark Siderits on Anumāna

337

example (vaidharmya-dxrṣṭānta) because the similar domain is characterized by the presence of the target property and the dissimilar domain is characterized by the absence of the target property, while the similar example such as ‘a pot’ possesses both the target property and the proving property (sādhana-dharma) such as ‘the property of being produced’ and the dissimilar example possesses neither the target nor the proving property. Therefore, it is misleading to translate sapakṣa and vipakṣa by ‘similar instances’ and ‘dissimilar instances’ respectively. Now, according to Dignāga, the subject or topic of the assertion (pakṣa) such as ‘sound’ belongs to the unobserved or unknown realm because it is questioned and has not been determined if it is non-eternal or not.37 In this framework of pakṣa (the unknown realm), and sapakṣa and asapakṣa/ vipakṣa (the known realm), one infers some state of affairs, which belongs to the so-far unknown realm (pakṣa), from the observations of the relationship between the two properties in the known realm (sapakṣa + asapakṣa/vipakṣa). For example, in the above formulation (8), suppose that we observe the presence of the property of being produced (i.e. sādhana-dharma) only (eva) in the non-eternal things (i.e. sapakṣa) and that we never (naiva) observe its presence in the eternal things (i.e. asapakṣa/vipakṣa), then, says Dignāga, we can assume that whatever possesses the property of being produced also possesses the property of being non-eternal – at least in the observed/known world or as far as we have observed/known. If we apply that general rule to the subject of the assertion ‘sound’ (i.e. pakṣa) in the unknown realm and if we can ascertain that sound is produced, then we can infer that sound too is non-eternal. This is essentially ‘reasoning by analogy’ or ‘inductive reasoning’ from the observed/known realm (of sapakṣa and asapakṣa/vipakṣa) to the pakṣa in the unobserved/unknown realm.  At the later stage of Nyāya tradition, Keśavamiśra (thirteenth century) defines pakṣa, sapakṣa and vipakṣa in the similar way as Dignāga. TBh p. 44: ke punaḥ pakṣa-sapakṣa-vipakṣāḥ | ucyante | saṃdigdhasādhyadharmā dharmī pakṣaḥ | yathā dhūmānumāne parvataḥ pakṣaḥ | niścitasādhyadharmā dharmī sapakṣaḥ | yathā tatraiva dhūmānumāne mahānasaḥ | vipakṣas tu niścitasādhyābhāvavān dharmī | yathā tatraiva mahāhrada iti | (But what are pakṣa, sapakṣa and vipakṣa? Answer: pakṣa is a property-possessor (dharmin) of which the possession of the target property is doubtful, as e.g. in the inference [of fire] from smoke [on the top of a mountain] the mountain is pakṣa; sapakṣa is a property-possessor of which the possession of the target property is ascertained, as e.g. the kitchen stove in the same inference from smoke; and vipakṣa is a property-possessor of which the possession of the absence of the target property is ascertained, as e.g. the lake in the same inference.)  Regarding the two kinds of examples, the commentator of TBh, Cinnaṃbhaṭṭa (p. 256) gives the following definitions that accord with Dignāga’s understanding: sādhyasādhanayor yatra sāhacaryaṃ dṛśyate sa sādharmyadṛṣṭāntaḥ | tadabhāvayoḥ sāhacaryaṃ yatra sa vaidharmyadṛṣṭāntaḥ | (The similar example is that in which the co-presence of the target and proving [properties] is observed, and the dissimilar example is that in which the co-presence of the absence of the both [properties is observed]. It is to be noted that even at this late stage the examples are regarded as the objects of ‘observation’, which is suggested by the word ‘dṛśyate’; Dignāga too inserts the word ‘dṛṣṭa’ sometimes in his formulation of the examples.  Cf. TSaṃ, p.  31: saṃdigdhasādhyavān pakṣaḥ | yathā dhūmavattve heatu parvataḥ || niścitasādhyavān sapakṣaḥ | yathā tatraiva mahānasaḥ || niścitasādhyābhāvavān vipakṣaḥ | yathā tatraiva mahāhradaḥ || 37

338

S. Katsura

5 Reasoning from anvaya and vyatireka In this connection I would like to refer to George Cardona’s idea of ‘Indian principle of induction’ who says: “Indian thinkers have used a mode of reasoning that involves the related presence (anvaya ‘continued presence’) and absence (vyatireka) of entities as follows: (1) a. When X occurs, Y occurs. b. When X is absent, Y is absent. (2) a. When X occurs, Y is absent. b. When X is absent, Y occurs.

If (1 a, b) hold in all instances for X and Y, so that these are shown consistently to occur together, one is entitled to say that a particular relation obtains between the two. Either (1a) or (1b) alone will not justify this, and a claim made on the basis of either can be falsified by showing that (2a) or (2b) holds. One relation that can be established by (1) is that X is a cause of Y. A special instance of the cause-effect relation involves the use of given speech units and the understanding by a hearer of given meanings. If (1 a, b) hold, the speech unit in question is considered the cause of one’s comprehending a meaning, which is attributed to that speech element.”38 According to Cardona, if anvaya-vyatireka relation is always observed between two items without any exception or deviation, then we can establish a particular relation such as the cause-effect relation between the two items. For example, when the word ‘go’ is uttered, there occurs the understanding of a quadruped animal ‘cow’ but not of any other animal, and when the word ‘go’ is not uttered, there does not occur the understanding of a cow; in this way, we can establish that the word ‘go’ is the cause of the understanding of its meaning ‘cow’. Cardona considers that it is Indian principle of induction to establish a certain relation such as the cause-effect relation by means of reasoning from anvaya and vyatireka. Cardona’s formulations of anvaya and vyatireka, namely, “When X occurs, Y occurs.” and “When X is absent, Y is absent”, reminds us of the formulations of idaṃpratyayatā (literally means ‘the nature of being conditioned by this’, i.e. the principle of causality/causation) from the Buddhist tradition. When the twelve-­ membered dependent origination (pratītyasamutpāda) is taught in the early Buddhist scripture such as the Majjhima Nikāya 115, it is often forwarded by the following phrases: “When this exists, that comes to be; with the arising of this, that arises. When this does not exist, that does not come to be; with the cessation of this, that ceases.”39 This set of phrases is later called idaṃpratyayatā or ‘agreement/convention regarding dharmas’ (dharma-saṃketa). It is most likely that even before Indian grammarians utilized the reasoning from anvaya and vyatireka in order to

 Cardona (1980). For more discussion on the related topics, please see Katsura (1979) and (2000b). 39  Translation by Bhikkhu Ñānamoli and Bhikkhu Bodhi. The Middle Length Discourses of the Buddha, A New Tranlsation of the Majjhima Nikāya, Wisdom Publications, 1995, p. 927. 38

Mark Siderits on Anumāna

339

analyze linguistic items into components and to attribute particular meanings to these components, Indian Buddhists had utilized the same principle in order to discover the cause-effect relation among dharmas such as the relation between ignorance (avidyā) and formations/dispositions (saṃskāra). As an example of Buddhist reasoning from anvaya and vyatireka I quote a passage from the ninth chapter of the Abhidharmakośa-bhāṣya of Vasubandhu. In order to show that there is no anumāna to prove the existence of ātman, he gives one example of what he thinks to be a proper anumāna, which is in fact reasoning from anvaya and vyatireka: “And there is anumāna. For example, [there is anumāna that proves the existence] of the five sense faculties. In this connection this is anumāna; it is observed (dṛṣṭa) that, suppose the cause X is present, when the other cause Y is absent, the effect Z is absent, but when Y is present, Z is present, as e.g. a sprout [that does not appear if there is no seed, which may be buried under the earth and invisible, even if there are the earth, water, etc.]. It is observed that suppose there are causes such as an visible object, which is reached by the light, and one’s own attention, there is no grasping of the object for blind or deaf people and there is grasping of the object for people who are neither blind nor deaf. Therefore, in this case also the absence and the presence of the other cause is [respectively] determined. And that ‘other cause’ is the sense faculty. This is anumāna.”40

In this connection there is anvaya-vyatireka relation between the cause Y (such as the seed or the sense organs) and the effect Z (such as the sprout or the grasping of the object) and by means of the reasoning from anvaya and vyatireka Vasubandhu proves the existence of the imperceptible sense faculties. Now, let’s go back to Dignāga. The standard formulation of the second and the third conditions of a proper reason runs as follows: [hetoḥ] sapakṣe sattvam (the reason’s presence in sapakṣa) and vipakṣe ‘sattvam (the reason’s absence in vipakṣa).41 Since sapakṣa is the domain where the target property occurs and vipakṣa is the domain where the target property is absent, the second and the third conditions may be reformulated like “where the target property occurs, the reason occurs” and “where the target property is absent, the reason is absent.” In other words, they take the form of anvaya (‘continued presence’ or the affirmative concomitance) and vyatireka (‘continued absence’ or the negative concomitance). Thus, it is clear that Indian principle of induction is applied to discover the proper reason, for the proper reason is the cause of the cognition of the target property with reference to a certain object. In this connection I would like to refer to Dignḡa’s treatment of anvaya and vyatireka when he discusses how a word expresses its meaning in the Pramāṇasamuccaya-­ vṛtti chapter 5.  AKBh, p.  479: anumānaṃ ca | tadyathā pañcānām indriyāṇām | tatredam anumānam | sati kāraṇe kāraṇāntarasyābhāve kāryasyābhāvo dṛṣṭo bhāve ca punar bhāvas tadyathāṅkurasya |  saty eva cābhāsaprāpte viṣaye manaskāre ca kāraṇe viṣayagrahaṇasyābhāvo dṛṣṭaḥ punaś ca bhāvo ’ndhabadhirādīnām anandhābadhirādīnāṃ ca | atas tatrāpi kāraṇāntarasyābhāvo bhāvaś ca niścīyate | yac ca tatkāraṇāntaraṃ tad indriyam ity etad anumānam | 41  For the formulations of the three conditions (trairūpya) of a proper reason and the interpretation of the Pramāṇasamuccaya 2.5 cd, please see Katsura (1983, 1985, 2000a) and Lasic (2009). 40

340

S. Katsura

“Since a word is not observed to apply to the referent of other words but it is observed to apply to a member of its own referent, it is easy to connect [a word to its referent/ meaning] and it does not deviate [from its referent]. (PS 5.34) For anvaya and vyatireka are the means to the word’s expressing its referent. And they are respectively defined as ‘its application to the similar domain’ (tulye vṛttiḥ) and ‘its non-­ application to the dissimilar domain’ (atulye ‘vṛttiḥ). Of the two, however, it is not necessarily possible to state the application to all instances whatever in the similar domain because it is impossible to state all the referents, as they are infinite in number. On the other hand, it is possible to state the non-application to the dissimilar domain, even though [its components are] infinite in number, on the basis of mere non-observation (adarśana-mātreṇa). And for this very reason, as [a word is] not observed to apply to other referents than its own relata (sambandhin), anumāna that excludes those [other referents] (tadvyavacchedānumāna) is called ‘that which expresses its own referent’ (svārthābhidhāna). If, however, anumāna is done by means of anvaya, with reference to one and the same entity (i.e. a tree), the word ‘vṛkṣa’ should not give rise to doubt appearing as ‘Is it Śiṃśapā or not?’ and so on. Yet, as there is doubt about its being Śiṃśapā or not, there should also be doubt appearing as ‘Is it made of the earth element (pārthiva)?’ ‘Is it substance (dravya)?’ and so on. However, the word ‘vṛkṣa’ is not observed to apply to what is not made of the earth element; hence, anumāna is done only by means of vyatireka.”42

In Cardona’s analysis of anvaya and vyatireka, the cause-effect relation holds between a word and the understanding of its referent/meaning. In the case of Dignāga, however, the relation stands between the referent/meaning and its linguistic expression/word. In other words, X in Cardona’s formulation becomes a referent/meaning of a word and Y is that word. So, it may run like this: “When a certain referent/meaning X is intended, a word Y is uttered; and when X is not intended, Y is not uttered.” It is most interesting to note that Dignāga is well aware of the fact that it is impossible to check a word’s application to the similar domain as well as non-application to the dissimilar domain, which seems to indicate that he had to face ‘the problem of induction’. He tries to solve the problem by stating that it is possible to establish the vyatireka relation as long as a counter-example is not observed in the dissimilar domain (adarśana-mātreṇa). In other words, as long as we do not observe that any other word than ‘go’, such as ‘aśva’ or ‘hastin’, is 42

 English translation is made from the following Sanskrit text given by Pind (2015 Part 1: 40–42) adṛṣṭer anyaśabdārthe svārthasyāṃśe ‘pi darśanāt |

śruteḥ sambandhasaukaryaṃ na cāsti vyabhicāritā || PS 5.34 || anvayavyatirekau hi śabdasyārthābhidhāne dvāram, tau ca tulyātulyayor vṛtyavṛttī. tatra tu tulye nāvaśyaṃ sarvatra vṛttir ākhyeyā kvacid, ānantye ‘rthasyākhyānāsambhavāt. atulye tu saty apy ānantye śakyam adarśanamātreṇāvṛtter ākhyānam. ata eva ca svasambandhibhyo ‘nyatrādarśanāt tadvyavacchedānumānaṃ svārthābhidhānam ity ucyate. anvayadvāreṇa cānumāne vṛkṣaśabdād ekasmin vastuni śiṃśapādyābhāsaḥ saṃśayo na syāt. tatsaṃśayavat pārthivatvadravyatvādyābhāso ‘pi saṃśayaḥ syāt. yatas tu , ato vyatirekamukhenaivānumānam.   I greatly benefitted from Pind’s English translation and Appendix 10 (Pind 2015 Part 2: 116–125, 239ff). For the former studies of the same passage, please see Frauwallner (1959) and Katsura (1979) and (1992b).

Mark Siderits on Anumāna

341

applied to a cow, we can safely say that the word ‘go’ is applied to a cow. Since Dignāga regards a verbal knowledge (śābda) is a kind of anumāna, what is mentioned above is equally applied to anumāna and I shall discuss it immediately below.

6 Pervasion (vyāpti) Now Dignāga adds the restrictive particle eva to the both second and the third conditions of a proper reason as ‘sapakṣa eva sattvam’ and ‘vipakṣe ‘sattvam eva’, which come to mean “only where the target property occurs, the reason occurs” and “where the target property is absent, the reason is certainly absent”. They can be easily put into the form of ‘pervasion’ (vyāpti), namely, “wherever the reason occurs, the target property occurs,” (anvaya-vyāpti) and “wherever the target property is absent, the reason is absent” (vyatireka-vyāpti). Those statements of pervasion may sound unrestricted and be suspected to have universal application but that is not the case because, as I mentioned above, they are restricted to ‘the observed and known (dṛṣṭa/vidita) realm’. However, in that restricted realm the statement of pervasion can and should be understood unrestrictedly. Therefore, in that restricted realm of known world anumāna like “Sound is non-eternal because it is produced; whatever is produced is non-eternal like a pot” can be regarded as a deductively valid argument.43 As I mentioned above Dignāga seems to hold that the inference from the known realm to the dubious and unknown realm is justified only by means of non-­ observation of the reason in the dissimilar domain (adarśanamātreṇa); in other words, as long as we do not discover the counter example in the dissimilar domain,  In this ‘restricted’ sense I seem to share the same interpretation of anumāna with Siderits (2016: 125): So after Dignāga introduced pervasion as a technical term and formulated criteria for its obtaining, later Naiyāyikas could readily agree with his analysis and make the implicit premisses explicit: 43

p is S Because p is H Wherever H is, S is, like in the sapakṣa Unlike in the vipakṣa. What is not entirely clear here is whether the statement of pervasion is meant to be unrestricted (having universal application), or restricted to just cases apart from the subject of dispute (p). Now if the statement of pervasion is to be understood unrestrictedly, then the argument consisting of these statements is deductively valid. The fact that it also includes statements citing some of the evidence in support of the pervasion (the statements concerning the sapakṣa and vipakṣa) in no way detracts from the deductive validity of the argument. No matter what statements we conjoin to the conjunction of ‘p is H’ and ‘For all x, if xH then xS’, it will remain impossible for the resulting conjunction to be true while ‘p  is S’ is false. Of course typically an argument from the statements concerning instances of sapakṣa and vipakṣa to the statement of pervasion is not deductively valid. But that was not the question. The question was rather whether the argument consisting of all the statements occurring in the formulation of an anumāna is deductively valid. For this formulation the answer is Yes.

342

S. Katsura

we can infer about the unknown world. Now Dharmakīrti strongly criticized the theory of adarśanamātra in the Pramāṇavārttika-svavṛtti, work written in his earlier career. Although he is supposed to have criticized Īśvarasena, disciple of Dignāga, who claimed that nonexistence can be proved by the third pramāṇa, i.e. ‘mere non-perception’, Dharmakīrti must have been aware of the weakness of Dignāga’s theory of anumāna. He introduced the idea of ‘natural connection’ (svabhāva-pratibandha) as the ground for establishing pervasion.44 Dharmakīrti admits only two kinds of relation between the reason/inferential mark and the target property, namely, the cause-effect relation and the relation of identity, which make up the natural connection that justifies the pervasion. If X (e.g. fire) is established as the cause of Y (e.g. smoke), then the effect Y is the inferential mark of X, and if the possessor of X (e.g. the property of being produced) is identified as the possessor of Y (e.g. the property of being non-eternal), in other words, if the domain of X is included in the domain of Y, then X is the inferential mark of Y. In this way Dharmakīrti tries to give the ontological and metaphysical ground for pervasion. He wanted to get rid of the hypothetical nature of Dignāga’s anumāna but I do not think that the introduction of the concept of ‘natural connection’ in anumāna really got rid of ‘inductive’ aspect of Indian anumāna. Dharmakīrti would formulate the proof (8) mentioned above in the following manner: [pervasion] Whatever is produced is non-eternal like a pot. [pakṣadharmatā] And sound is produced. [Conclusion: Sound is non-eternal.]    (9)

On the surface level, the above formulation looks very much like Aristotelian syllogism of Barbara type and may suggest that Dharmakīrti’s anumāna is ‘deductive’ in nature but the presence of an example ‘pot’ in the statement of pervasion reveals the reminiscence of ‘inductive’ aspect of Indian anumāna. This will lead to the problem of bahirvyāpti and antarvyāpti debate among the later Buddhist logicians such as Ratnākaraśānti and Jñānaśrīmitra, which is beyond the scope of the present paper.45

References Cardona, George. 1980. On Reasoning from anvaya and vyatireka in Early Advaita. In Studies in Indian Philosophy, A Memorial Volume in Honour of Pandit Sukhlaji Sanghvi, LD Institute of Indology, 1981 ed. D. Malvania and N.J. Shah, Ahmedabad. Dasti, Matthew, and Stephen Phillps. 2017. The Nyāya-sūtra, Selections with Early Commentaries. Indianapolis/Cambridge: Hackett Publishing Company, Inc. Frauwallner, Erich. 1959. Dignāga, sein Werk und seine Entwicklung. In Wiener Zeitschrift für die Kunde Süd- und Ost-Asiens, Band III.

 Please see Katsura (1986c) and (1992a) and for the more recent contribution to the topic, please see Kataoka (2014). 45  For the most up-to-date information of the topic, please see Ono (2004) and Sakai (2019). 44

Mark Siderits on Anumāna

343

Kataoka, Kei. 2014. Falsifiability in Indian Philosophy. South Asian Classical Studies, No. 9. Katsura, Shoryu. 1979. The apoha Theory of Dignāga. Journal of Indian and Buddhist Studies 28-1. ———. 1982. Kumārila’s Theory of Inference Compared with that of Dignāga (in Japanese). Journal of Indian and Buddhist Studies 31-1. ———. 1983. Dignāga on trairūpya. Journal of Indian and Buddhist Studies 32-1. ———. 1985. On trairūpya Formulae. In Dr. Kumoi Festschrift – Buddhism and Its Relation to Other Religions. Kyoto: Heirakuji Shoten. ———. 1986a. On the Origin and Development of the Concept of vyāpti in Indian Logic – from the Carakasaṁhitā up to Dharmakīrti  – (in Japanese), The Hiroshima University Studies Faculty of Letters, Vol. 45, Special Supplement 1. ———. 1986b. On the Origin and Development of the Concept of vyāpti in Indian Logic. Tetsugaku 38. ———. 1986c. Svabhāvapratibandha Revisited. Journal of Indian and Buddhist Studies 35-1. ———. 1992a. Pramāṇavārttika IV.202–206  – svabhāvapratibandha reconsidered. Journal of Indian and Buddhist Studies 40–2. ———. 1992b. Dignāga and Dharmakīrti on adarśanamātra and anupalabdhi, Festschrift for Prof. Dr. J. May, Asiatische Studien XLVI-1. ———. 2000a. Dignāga on trairūpya Reconsidered. Festschrift for Dr. Hiromasa Tosaki, Culture and Logic in India, Kyushu University Press. ———. 2000b. How did the Buddhists prove something? – The Nature of Buddhist Logic – The Numata Ehan Lecture in Buddhism 1996, the University of Calgary. ———. 2004a. The Role of dṛṣṭānta in Dignāga’s Logic, S. Katsura and E. Steinkellner, (eds.), The Role of the Example (dṛṣṭānta) in Classical Indian Logic, Wiener Studien zur Tibetologie und Buddhismuskunde No. 58, Wien. ———. 2004b. Pakṣa, sapakṣa and asapakṣa in Dignāga’s logic. Horin 11. ———. 2016. A Report on the Study of Sanskrit Manuscript of the Pramāṇasamuccayaṭīkā Chapter 4, Recovering the Example Section of the Nyāyamukha. Journal of Indian and Buddhist Studies 64-3. ———. forthcoming. A Brief History of Proof Formulation (prayoga) in Indian Logic up to Dharmakīrti, B.K. Matilal Commemoration Volume. Lasic, Horst. 2009. Dignāga’s Description of a Logical Mark in Pramāṇasamuccaya 2.5cd. International Journal of Buddhist Thought & Culture 13:7–22. Ono, Motoi. 2004. Antarvyāpti and bahirvyāpti in the School of Buddhist Logic – Focusing on Prajñākaragupta, Festschrift in Celebration of Professor Eshō Mikogami’s Longevity, Nagata Bunshōdō, pp. 457–492. Pind, Olle Holten. 2015. Dignāga’s Philosophy of Language, Pramāṇasamuccayavṛtti V on anyāpoha, ed. E.  Steinkellner, Part 1: Text and Part 2: Translation and Annotation. Wien: Österreichischen Akademie der Wissenschaften. Prets, Ernst. 2010. On the Proof Passage of the Carakasaṃhitā: Editions, Manuscripts and Commentaries. In Logic in Earliest Classical India, ed. Brendann S.  Gillon. Delhi: Motilal Banarsidass Publishers. 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, No. 31. Siderits, Mark. 2016. Studies in Buddhist Philosophy, ed. Jan Westerhoff. Oxford: Oxford University Press.

344

S. Katsura

Abbreviations AKBh: Abhidharmakośa-bhāṣya of Vasubandhu, ed. P. Pradhan, K.P. Jayaswal Research Institute, Patna, 1967. CS: The Caraka-saṃhitā of Agniveśa, ed. Vaidya Jadavaji Trikamji Āchārya, Bombay, 1941. NBh Nyāyabhāṣya of Vātsyāyana/Pakṣilasvāmin, ed. Taranatha Nyaya-Tarkatirtha, Calcutta Sanskrit Seires, 18 & 19, Calcutta, 1936–44. NPra Nyāyapraveśa of Śaṅkarasvāmin, M. Tachikawa, ed. 1971. A Sixth-century Manual of Indian Logic. Journal of Indian Philosophy 1:111–145. ŚV Ślokavārttika of Kumārila with the Commentary Nyāyaratnākara of Pārathasārathi Miśra, ed. Svāmī Dvārikādāsa Śāstrī, Tara Publications, Varanasi, 1978. TBh Tarkabhāṣā of Keśava Miśra with the commentary, Tarkabhāṣāprakāśikā of Cinnaṃbhaṭṭa, ed. Devadatta Ramkrishna Bhandarkar and Pandit Kedarnāth, Bhandarkar Oriental Research Institute, Bombay, 1979. TSaṃ Tarkasaṃgraha of Annaṃbhaṭṭa, ed. Nārāyaṇa Rāma Āchārya, 14th ed., Nirṇayasāgara Press, Bombay, 1963.

Buddhist Reductionism, Fictionalism, and Expressibility Laura P. Guerrero

1 Introduction While committed to the view that Buddhist Reductionism offers the best account of the Abhidharma distinction of the two truths, Siderits (2009) argues that Buddhist Reductionism has the surprising consequence of making itself inexpressible. This inexpressibility follows from the semantic insulation between conventional and ultimate discourses that Siderits argues is required in order to preserve classical logic for both types of discourse, avoiding contradiction and bivalence failure. I argue that inexpressibility is a problematic consequence that threatens to constitute a reductio of Buddhist Reductionism. However, there is an alternative fictionalist formulation of Buddhist Reductionism, similar to the one offered by Goodman (2005) that still captures the sense in which conventional discourse supervenes on ultimate discourse while avoiding the concerns that motivated the semantic insulation that rendered mixed vocabulary discourse inexpressible. This alternative fictionalist formulation of Buddhist Reductionism, based on the work of Kendall Walton, treats statements of ‘reduction’ as meta-linguistic generation principles that serve as prescriptions about what is to be imagined when engaging in conventional discourse. Since they do not involve statements of identity or about the composition of conventional entities in terms of ultimate entities they do not give rise to the kind of sorites issues that threaten classical logic. Metalinguistic truths about what is conventionally true can be expressed from an ultimate perspective, and by making bivalence itself a generative principle of conventional discourse, classical logic can be preserved at both the conventional and ultimate levels of discourse. Thus, I argue that Buddhist Reductionism can be, when properly formulated, both expressible and ultimately true. L. P. Guerrero (*) College of William and Mary, Department of Philosophy, Williamsburg, VA, USA e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_18

345

346

L. P. Guerrero

2 Two Truths in Abhidharma Traditions Buddhist thinkers and traditions understand the distinction between the two truths, and their relation, in a wide variety of ways. In the context of Abhidharma Buddhist philosophies, the distinction came to be understood, broadly, as a distinction between two kinds of objects. The Sarvāstivāda Abhidharma compendium the Mahāvibhāṣā, for example, distinguishes between various ways in which an entity can be said to exist. The text differentiates between entities that have real (dravya) existence and entities that have only provisional (prajñapti) existence. Cox (2004) explains that for the Mahāvibhāṣa, Dharmas that are determined by intrinsic nature (svabhāva) exist as real entities (dravyatas), in contrast to composite objects that lack such a unique, defining intrinsic nature. Such composite objects exist only provisionally (prajñaptitas) in dependence upon their constitutive fundamental dharmas (569–570).

According to the Mahāvibhāṣā, entities that exist in some fundamental sense form a distinct class of entities that have real existence. Entities that depend on those fundamental entities form another class of entities that exist merely “provisionally,” i.e. conventionally. Similarly, in the Sarvāstivāda-Sautrāntika Abhidharma Buddhist text the Abhidharmakośabhāṣya, Vasubandhu also differentiates between ultimately existing entities and entities that only exist conventionally in dependence on the ultimate reals. He then uses this distinction to explain a distinction between two kinds of truth: ultimate truth and conventional truth. He says in his commentary: If the idea of a thing disappears when this thing is broken into pieces, then this thing has relative existence (saṃvṛtisat); for example, a jug: the idea of a jug disappears when it is reduced to pieces. If the idea of a thing disappears when this thing is dissipated, or broken to pieces, by the mind, then this thing should be regarded as having relative existence; for example, water. If we grasp and remember the dharmas, such as color, etc., in the water, then the idea of water will disappear. These things,—jugs, clothes, etc., water, fire, etc.,— are given their different names from the relative point of view or conforming to conventional usage. Thus if one says, from the relative point of view, ‘There is a jug, there is water,’ one is speaking truly, and one is not speaking falsely. Consequently this is relatively true. That which differs is absolute truth. (Pruden 1990, 910–911)

Here, like in the Mahāvibhāṣā, Vasubandhu describes a kind of relative, conventional existence (saṃvṛtisat) that is conventional in virtue of the fact that its existence can be analyzed into more fundamental components, the dharmas such as color, etc., which are fundamental insofar as they cannot be so analyzed into anything more fundamental. An ontological distinction between conventional and ultimately existing objects is also drawn in Theravāda texts. In the canonical The Questions of King Milinda (Milindapañha), for example, ordinary entities such as chariots and people are described as existing as conventional designations made on the basis of more fundamental elements. Quoting another canonical text, the Milindapañha tells us that “just as it is by the condition precedent of the co-existence of its various parts that the word ‘chariot’ is used, just so it is that when the skandhas are there we talk of a

Buddhist Reductionism, Fictionalism, and Expressibility

347

‘being’” (Davids 1890, 45).1 In this Theravāda text, conventional objects such as chariots and persons are described as having a different kind of existence from those entities on which they depend for their being. While ultimately there are no chariots or persons, there is some sense in which such things do exist. They are described as existing conventionally, as “coming under the generally understood term, the designation in common use” (44). All these texts draw a distinction between conventionally real and ultimately real objects, grounding the former in some sense on the latter. However, these texts do not say much about the details of this grounding relation or its ontological implications. It was left to subsequent thinkers to makes sense of the implications of the views presented. Mark Siderits is one of the thinkers that has since come to reflect seriously on the philosophical and ontological implications of the views presented in texts such as these.

3 Buddhist Reductionism and Sider-Style Fictionalism Siderits characterizes the Abhidharma traditions that are reflected in the texts like those cited above as metaphysical and semantic realists. According to him, these traditions maintain that there is some mind-independent way the world is and that truth, strictly speaking, is determined by correspondence to that world. However, Siderits, like many other Buddhist scholars, also characterizes these traditions as local anti-realists about conventional objects, which he characterizes as mereological wholes.2 This mereological nihilism entails that virtually all of the ordinary objects of our experience turn out to be unreal. On this view the distinction between conventional and ultimate reality is used by these traditions to differentiate ordinary, but unreal, entities from the ultimately existing entities on which they depend. As we saw in the Abhidharmakośabhāṣya, the related distinction between conventional and ultimate truth is used to differentiate the distinct but related norms of acceptance for claims made about unreal, conventional entities and about ultimate entities. Siderits (2015) argues that for Abhidharma traditions of Buddhism the relationship between conventional and ultimate truth should be understood as a form of reductionism he calls ‘Buddhist Reductionism.’ Siderits calls the view (somewhat misleadingly, some think) ‘reductionist’ in order to highlight the fact that entities and claims at the conventional level depend on what is real and true at the ultimate level. Siderits argues that Abhidharma Buddhist conventional discourse is “reduced” to discourse about what is ultimately real. Reduction in this case involves the  The skandhas are the psycho-physical elements that are understood to serve as the foundation for the existence of persons. 2  See for example Bronkhorst (2009), Collins (1982), Ronkin (2005), and Walser (2005). For realist interpretations of conventional objects see Benovsky (2017a, b), Duerlinger (2003), Ganeri (2012), McDaniel (2019), and Williams (2000). 1

348

L. P. Guerrero

supervenience of conventional discourse, characterized in Siderits (2009) as a kind of fictionalism, on ultimate discourse. By reducing conventional discourse to ultimate discourse in this way, Siderits hopes to explain how discourse about unreal conventional objects can still be efficacious and how claims made within that discourse can still be assessed for truth and falsity (in some sense). Buddhist Reductionism is taken by him to explain how Abhidharma Buddhists can keep meaningful discourse about mereological wholes without having to admit them into their final ontology. Several scholars, including Siderits (2006, 2009), have drawn the obvious comparison between Buddhist conventional discourse and fictional discourse.3 Just as fictional discourse allows us to talk about unreal fictional entities in an ontologically uncommittal way, so does Buddhist conventional discourse allow us to talk about unreal conventional objects in an ontologically uncommittal way. Most fictionalist accounts share the concern of explaining the norms of acceptance for fictional claims (such that we might describe such claims as being fictionally true or fictionally false) in such a way as to avoid ontological commitment to the entities of such discourse. For Buddhist Reductionism, Siderits (2009) suggests adopting a model of fictionalism offered by Sider (1999). Sider (1999) offers an account of what he calls ‘quasi-truth.’ His idea is that we can explain why we accept discourse about unreal things by “discovering underlying truths” that supports it (325). These underlying truths, he says, “suffice for the claim in question to be, if not true, then at least quasi-­ true” (331). The account will allow, as he says, that “even if we deny that the problematic sentences are true, we can still accord them some positive status” (331). Siderits (2009: 61) characterizes Sider’s quasi-truth in the following way: Suppose S is among the widely accepted statements generated by acceptance of folk theory T. And suppose that T involves ontological commitments sufficiently problematic that S must be deemed false. Then S is quasi-true if there is some true statement P such that, had T been true, P would have been true and would have entailed S. P then counts as an underlying truth, since the quasi-truth of S supervenes on the truth of P.

Sider says of his view that “a sentence is quasi true if the world is similar enough to the way it would have to be for the sentence to be genuinely true” (1999, 332). The idea is that when a statement is quasi-true, the nature of the world is such as to provide a kind of “quasi-supervenience base” for it that is not sufficient for the statement’s truth (as it would be, for example, if the folk theory were true) but close enough to grant the statement some approximation to the truth (ibid: 332). A conventional claim is not, on this view, a translation or paraphrase of some ultimately true claim; it is not another way of saying something ultimately true. Rather, the conventional claim is accepted because there is an underlying ultimate truth that, while not equivalent, supports the claim’s acceptance.

 See for example Crittenden (1981), D’Amato (2013), Garfield (2006), Goodman (2005), and Sauchelli (2016). 3

Buddhist Reductionism, Fictionalism, and Expressibility

349

Take for example an ordinary statement such as “This pot has peas in it,” (Siderits’ (2009: 61) example). The folk theory presupposed by the acceptance of this claim is mereological realism. Both pots and peas are ordinarily thought to be objects that have parts. Siderits has argued on behalf of Abhidharma Buddhists that the ontological commitment to mereological wholes is sufficiently problematic that we cannot accept mereological realism as a theory (Siderits 2015). Thus there are neither pots nor peas in reality. There is, however, a true statement along the following lines: ‘There are aggregates of dharmas in configurations α and β.’ Had the folk theory of mereological realism been correct, the pot and peas would be real and identical to just such aggregates of dharmas and so the true statement would entail the truth of the ordinary statement, ‘This pot has peas in it.’ On Sider’s account this counterfactual entailment means that the ordinary statement is quasi-true and that the true statement is the underlying truth on which the quasi-truth of the conventional statement depends. Siderits’ Sider-inspired proposal is thus that conventional statements are quasi-­ true because if the theory that mereological wholes are real had been true, then it would be the case that conventional entities would be real because of the way the dharmas that would compose them actually are. The world composed of dharmas in causally connected sequences is close enough to what it would have to be like for sentences about mereological wholes to be true, and so statements involving wholes such as pots and peas are afforded the “alethic compliment” of being quasi-true (Siderits 2009, 61). Following Sider, the formulation for conventional truth would go something like this: (CTS) A statement C is conventionally true if there is some ultimately true statement U such that had the folk theory of objects been true, U would have been true and would have entailed C.

This notion of conventional truth defines it in terms of its approximation to the ultimate truth, something that is made clear in Sider’s characterization of quasi-true statements as being ‘almost’ true and the state of affairs they describe as being “similar enough” to how things actually are. We can see in Sider’s account the important connection between conventional, quasi-true claims and real world facts that Siderits is, rightly, keen to ensure any Abhidharma account of the conventional retains. Unlike anti-realist discourse that is ripe for elimination, Siderits argues that conventional discourse has a utility and degree of objectivity in light of its approximation to what is in fact true. Siderits (2009, 62) says of Sider’s account that it seems to capture much of what Buddhist reductionists have in mind when they speak of conventional truth. They held, for instance, that for every statement that is conventionally true, there is some ultimate truth that helps to explain why the acceptance of the statement in question would lead to successful practice.

The underlying idea here is that the convenience and utility of conventional ways of speaking are directly tied to the fact that such discourse supervenes on what is in fact true.

350

L. P. Guerrero

4 Inexpressibility of Sider-Style Fictionalism The Sider-style formulation of Buddhist Reductionism, however, gives rise to various problems on the grounds of which it should be rejected. One important difficulty is that the account saddles the Buddhist Reductionist with the burden of giving an account of counterfactuals, something Siderits (1991: 118; 2009: 62, note 8) himself has noted is generally problematic in the Buddhist context. Classical Indian philosophers generally did not accept that counterfactuals, for example those employed in reductio-type tarka and prasaṅga inferences, could be assessed for truth because in principle there is no means by which they could be known. Knowledge in this Indian context involves cognitive processes that put the cognitive agent in touch with what is known and can thus only be had with respect to things that actually exist. We cannot know what is not actual. This is part of the reason why the Sarvāstivādin Ābhidharmikas, for example, argue for the existence of past and future dharmas: in order to explain how we can have cognitions of, and knowledge about, the future and the past. Part of the problem the Buddhists have, then, with explaining ‘truths’ about unreal conventional objects is precisely the problem of explaining how cognitions that do not have any actual content, that do not get in touch with anything real, could still constitute a kind of knowledge event. In this context, Siderits’ appeal to another problematic type of cognition featuring a counterfactual is not a promising way to understand or explain what conventional reality and truth amount to. The Sider-style formulation of Buddhist Reductionism seems to constitute an attempt to solve one problem with an even bigger problem. The main problem with the Sider-style approach, however, is that Siderits argues that the view implies its own inexpressibility. This inexpressibility follows from Siderits’ argument that semantic insulation is necessary to address the bivalence failure and contradictions that would otherwise arise for discourse that allows for a mixed vocabulary.4 Since Buddhist Reductionism understood in terms of CTS is articulated using a mixed vocabulary, this means that Buddhist Reductionism is inexpressible. Siderits is concerned that were one to allow ultimately true statements to contain expressions that refer to conventionally real entities, then the difficult questions about the relationship between wholes and their parts—the same questions which led to the denial of the reality of mereological wholes in the first place—would demand ultimate answers that could not be given. In particular, Siderits is concerned that these kinds of questions gives rise to sorites cases which introduce indeterminacy and truth value gaps. Siderits (2009, 64) explains that conventional language involving reference to ordinary objects is by its nature vague or ‘tolerant,’ allowing for variation and indeterminacy of application conditions. A pot, to use Siderits’ example, is subject to sorites problems in a way that its ultimate constituents are not. The state of affairs involving ultimate existing dharmas will always be, according to Ābhidhārmikas, determinate. However, whether or not some such state of affairs is sufficient to determine the  See Siderits (2009), (2015), and (2016) for arguments to this effect.

4

Buddhist Reductionism, Fictionalism, and Expressibility

351

existence of a pot is not, as sorites examples demonstrate. There will be borderline cases where the existence of a pot will be indeterminate, not because the state of the dharmas is indeterminate but because the application conditions for the concept ‘pot’ are not precise enough. In order to avoid these problems Siderits argues that the best thing to do is to insulate ultimate discourse from conventional discourse such that no ultimately true statement could contain expressions referring to conventional entities. Siderits makes this explicit by defining ultimate truth in the following way: (UT) A statement is ultimately true if and only if it corresponds to mind-independent reality and neither asserts nor presupposes the existence of anything not in the final ontology. (2009, 60)

This formulation of ultimate truth precludes the possibility that a claim including reference to anything but an ultimately existing entity could be ultimately true. It follows from this formulation that Buddhist Reductionism could not be ultimately true. On the Sider-style formulation the supervenience claim at the heart of Buddhist Reductionism rest on a counterfactual conditional, reflected in CTS, that could not be ultimately true because it employs a mixed vocabulary. Therefore, Buddhist Reductionism cannot be expressed as an ultimate truth. And for similar reasons concerning the preservation of classical logic at the conventional level, Siderits argues that conventional truth should also be defined so as to preclude mixed vocabulary. No conventionally true statement can contain expressions that refer to ultimately real entities. Consequently, Buddhist Reductionism is not conventionally true either. Since it is articulated using a mixed vocabulary it is ill-­ formed by either standard and thus is meaningless and inexpressible. This inexpressibility is certainly itself an undesirable consequence that undermines the plausibility of the view. Setting aside issues of intelligibility, as Siderits himself is keen to emphasize, reductive accounts that relate different discourses in order of explanatory priority or explain one kind of entity in terms of another are common not only in science but in ordinary discourse. We make such statements all the time, take them to be not only meaningful but often true, and use them in our practical reasoning about the world. If part of the goal of Buddhist Reductionism is to explain ordinary, conventional discourse, then it should also be able to account for the aspect of that discourse that itself involves statements of reduction, and mixed discourse generally. This is particularly important in the Buddhist context given the extent to which the Buddhist Abhidharma tradition itself employs mixed vocabularies and statements of reduction. The Abhidharmakośabhāṣya, for example, begins by emphasizing that the goal of the texts is to generate knowledge on the part of the astute reader about the nature of the ultimate constituents of reality, the dharmas, because it is knowledge of the dharmas that eradicates the defilements that perpetuate the otherwise endless cycles of suffering and rebirth. These texts employ a mixed vocabulary throughout, strongly suggesting that the Abhidharma traditions reflected in texts like the Abhidharmakośabhāṣya, which Buddhist Reductionism is supposed represent, do not insulate ultimate and conventional discourse.

352

L. P. Guerrero

Throughout the Abhidharmakośabhāṣya Vasubandhu often proceeds by drawing analogies with ordinary experienced objects to make the subtle knowledge of the dharmas clearer. For example, when Vasubandhu is characterizing the sound dharmas he classifies them into four types, two of which he describes as “sound cause by the hand or by the voice” and “sound of the wind, of the tree, of water” (Pruden 1990, 65). In the same verse Vasubandhu also describes some of the color dharmas accepted by certain Ābhidhārmikas as “cloud” or “smoke” (ibid., 64). Dharmas, may be, as Sellars would say, first in the order of being, but they are not first in the order of knowing and Vasubandhu, beings seemingly sensitive to this fact, often draws analogies with conventional entities, like those cited above, in order to clarify the nature of the dharmas. It would seem, therefore, that from the way in which the discussion of the dharmas proceeds in these texts that adopting a mixed vocabulary, which bridges conventional and ultimate discourse, is a necessary condition to even be in a position to engage in ultimate discourse. If statements employing mixed discourse are really meaningless or false then it is difficult to see how texts such as the Abhidharmakośabhāṣya could be thought by those traditions to generate knowledge. Reflecting on texts such as the Abhidharmakośabhāṣya it becomes apparent that it is crucial to the epistemic and soteriological projects of the Abhidharma Buddhists that relations of reduction be not only expressible but also possible objects of inquiry and critical reflection. Thus, the argument that inexpressibility follows from Buddhist Reductionism, if correct, would, I think, constitute a reductio of that view.

5 Goodman’s Walton-Inspired Alternative Formulation of Conventional Truth In light of these problems, I propose an alternative fictionalist formulation of Buddhist Reductionism that still captures the sense in which conventional discourse supervenes on ultimate discourse but that avoids the concerns that motivated the semantic insulation that rendered mixed vocabulary discourse inexpressible. A version of the fictionalist account of conventional truth I propose, based on the account of fiction given by Kendall Walton (1990, 1993), has already been defended by Charles Goodman (2005). However, Goodman’s view contains an important misunderstanding of Walton’s view that distorts the fictionalist account made on its basis. After correcting that misunderstanding I will argue that the fictionalist formulation of Buddhist Reductionism based on Walton’s view provides an effective general account of conventional truth for Abhidharma traditions of Buddhism that is both expressible and consistent with the general approach to cognition and knowledge adopted in the classical Indian context in which Abhidharma thinkers were working. Siderits (2009) criticizes Goodman’s account, which he considers to be a kind of paraphrase account of fiction, on the grounds that it implausibly requires that every conventionally true claim, while strictly speaking false, is thought to be true because it is confused for a “nearby semantic neighbor” that is strictly speaking true (61).

Buddhist Reductionism, Fictionalism, and Expressibility

353

Siderits, like Sider, is skeptical that every conventional truth will have a nearby semantic neighbor for which it could be easily confused. The claim that there is an aggregate of dharmas in such-and-such arrangement over there does not, for example, seem to be semantically very close to the claim that there is a pot over there. A benefit of the Sider-style fictionalism, Siderits argues, is that it avoids having to make such implausible claims. This criticism of Goodman’s view is drawn, however, on the basis of a misunderstanding of Walton’s account that Siderits seems to inherit from Goodman. Goodman says that “[a]ccording to Walton, concise sentences that appear to presuppose the existence of fictional characters have the same meaning as claims that have no such presuppositions but are complex and difficult to state” (385 emphasis mine). On the basis of Goodman’s characterization, it seems, Siderits distinguishes his Sider-­ inspired account of Buddhist Reductionism from Goodman’s fictionalism on the grounds that Buddhist Reductionism does not employ the kind of paraphrase account of fiction that Goodman’s statement suggests (Siderits 2009, 61). However, contrary to what Goodman says, on Walton’s account claims that are fictional do not mean the same thing as the claims articulating the real world conditions that, on Walton’s view, serve as props for the fiction. In order to see that Walton’s view is not a type of paraphrase view, it is helpful to make explicit the structure of Walton’s view of fiction. According to Walton, normatively constrained fictions rely on real world conditions to structure and coordinate collective participation in them. Walton explains this grounding in terms of “props” and “generation principles.” According to him, props—real, intersubjectively available entities, either natural or designed—in conjunction with implicitly understood and culturally determined generation principles determine what is “fictional” with respect to a particular activity of make-believe and thus prescribes what is to be imagined by the participants in the fiction. For a particular proposition to be fictional it is not necessary that any individual actually imagine it, only that such an imagining is prescribed by some game of make-believe in light of some real world conditions and the generation principles of the game. This prescription provides an effective level of objectivity that normatively constrains such fictionalist discourse. Walton uses a child’s game of pretending that stumps are bears to illustrate his view. In this game, the rule ‘stumps are bears’ makes it the case that wherever there is a stump, a bear is to be imagined as being where the stump is. In other words, the rule makes it fictional that there is a bear amongst the trees if there is a stump there, even if no one sees the stump and no one imagines a bear there. Walton explains that, props insulate fictional worlds from what people do or think, conferring on them a kind of objective integrity worthy of the real world and making their exploration an adventure of discovery and surprise. (1990, 67)

Here Walton highlights the way in which real world props in conjunction with generation principles constrain what we ought or ought not pretend when engaging in a fiction. The fictional world can be revealed to us and our playmates as we participate in our collective imaginings in way that is intersubjectively available and in some sense objective. The props and the rules allow us to play the same game, occupy the

354

L. P. Guerrero

same imaginative world, by enjoining us to imagine the same things on the basis of the real world props present in our environment. According to the fictional account outlined by Walton, the norms governing participant attitudes for a fictionalist domain are determined by generation principles that express rules such that with respect to a specified fictionalist domain, when some real world condition, or prop, obtains, some fictionalist state of affairs is to be imagined. We can represent generation principles generally as having the following form: ‘according to fiction f, if p, then imagine x,’ where p represents the props and x represents what is to be imagined.5 With this view as a model, conventional entities can be defined as entities that are to be imagined by those participating in convention in virtue of the rules that govern conventionality, including its generation principles. On this formulation, participating in conventional ways of thinking and talking involves being governed by norms of participation that determine how the conventional world is to be imagined. This need not be explicit in a participant’s mind. Just as ordinary native speakers of a language skillfully use their language in a way that is responsive to the rules of the language without necessarily being able to explicitly articulate them, ordinary people engaged in fictional thought and speech need only be implicitly aware of, and responsive to, the rules that govern their participation. With Walton’s view in mind, Goodman characterizes the make-believe game of Buddhist conventional discourse in the following way: The game I have in mind involves pretending that various very general propositions about the existence and nature of composite things are true… If we pretend that they are true, then the regularities in the appearance and disappearance of momentary simples often make it appropriate to pretend that there are composite material substances. When the game makes it appropriate to pretend that a statement is true, then the statement is conventionally true. But if the game makes it appropriate to pretend that a statement is false, then that statement is conventionally false. (2005, 386–387)

Like Siderits, Goodman takes conventional discourse to primarily involve talk about mereological wholes, which he agrees Abhidharma Buddhists take to be unreal. Since we should not, in light of their unreality, believe claims about mereological wholes, these Buddhists advocate only imagining or pretending claims about them. What is appropriate to imagine or pretend is determined by the real world props, which in the Buddhist case are “the regularities in the appearance and disappearance of momentary simples,” and the implicitly understood generation principles that employ those props to generate the conventional fiction. Once Goodman’s Walton-style fictionalism is cleansed of misunderstanding, we can see that a Walton-style fictionalism shares with Siderits’ Sider-style account the notion that claims are conventionally true in virtue of some underlying truths that support them, without transforming through paraphrase conventional claims into ultimate ones. Claims are fictional in virtue of what is ultimately true without

 This is the formulation of Walton’s account that I give in Guerrero (2018), where I discuss socalled Buddhist global fictionalism in the context of Mahāyāna traditions of Buddhism. 5

Buddhist Reductionism, Fictionalism, and Expressibility

355

reducing those claims to the ultimately true claims on which they supervene. On the Walton-style formulation conventional truth would be defined as follows: (CTW) A statement C is conventionally true if there is some ultimately true statement U such that U in conjunction with some operative generation principle P entails that C is to be imagined.

Like Sider’s view, conventional truths supervene on ultimate truths without being equivalent to them. The connection of fictionality to props through generation principles does allow one to elliptically point to the real world props by using fictional discourse in what Walton calls ‘prop-oriented make-believe’ (1993).6 But this is not the same as claiming the fictional claim means the same thing as, or is a paraphrased form of, the claim articulating the real world props on which it depends. Returning to Walton’s example, pretending that there is a bear in the yard when one is engaged in the game is not a way of thinking about a stump, even if pretense could be used that way in a different kind of game. Rather, the stump merely serves to support our imagining a bear. It is the bear that is the subject of our imaginings, not the stump.

6 Expressibility The Walton-style formulation of Buddhist Reductionism outlined above can be expressed as an ultimate truth. This is a desirable outcome as Siderits (2009) himself says, “if the reductionist thesis is true, surely it must be more than merely conventionally true” (65). A Walton-style formulation is expressible because generation principles serve as rules that prescribe what is to be imagined by its participants and thereby what is conventionally true for that discourse. They are not, therefore, the kind of things that can themselves be assessed for truth or falsity. Generation principles do not describe what people do nor are they claims about composition or identity, rather they prescribe what ought to be conventionally imagined on the basis of what ultimately exists. The kinds of issues concerning identity and composition that Siderits raises with respect to CTS that motivate semantic insulation do not, therefore, arise in the context of the Walton-style account. To use Walton’s example again, pretending that stumps are bears does not require or entail that bears are identical to stumps or that bears be composed of stumps. Bears are to be imagined on the

 Sometimes, Walton points out, when we engage in fiction we do so to direct attention to the props that we employs as part of the fiction. For example, we might draw someone’s attention to a particular person by characterizing him as Santa Claus.This is possible in light of the generation principles which make it the case that by imagining some fictional entity, such as Santa Claus, one can assert something about the real world conditions, in this case a person, on which that fiction depends. Walton goes on to explain that sometimes it is helpful to use the medium of fiction to assert something of real objects that might otherwise be difficult to assert directly. But it is important to his view that while one can indirectly assert something about the real world using fiction, what one asserts fictionally does not mean the same thing as, and is not reducible to, the underlying truth on which it supervenes and to which it can indirectly indicate. 6

356

L. P. Guerrero

basis of stumps, and so in some sense supervene on them, but they do so in a way that does not involve issues of identity or composition. In the same way, a generation principle such as ‘dharmas arranged thus and so are pots’ does not give rise to any questions about whether or not, or when, pots are identical to or composed of dharmas. For this reason, there is no need to insulate conventional and ultimate discourse on the Walton inspired view. Statements of Buddhist Reductionism understood as generation principles that involve a mixed vocabulary are, therefore, expressible. Siderits here may worry that while CTW may not face the issues of identity and composition that CTS faces, it still faces indeterminacy in cases where it is unclear when a generation principle applies. In these cases all the details about how the dharmas are arranged are determinate but it is unclear, due to the imprecision of the rule, for example, whether or not one is to imagine a pot on their basis. In such a case, the kinds of questions that Siderits is concerned about that generate sorites issues seem to arise: Is it ultimately true or false that a pot is to be pretended? And when does it become determinate that the rule stops or starts applying? This indeterminacy could still threaten bivalence at the ultimate level. In response, it is worth first noting, as Siderits himself seems to acknowledge, that it is very likely that vagueness and imprecision are an inevitable feature of conventional, fictionalist discourse given that it is generated by finite creatures like us in light of our interests and cognitive limitations (2009, 64). In light of this it seems reasonable to allow that while ultimate level discourse is generally bivalent, ultimate discourse specifically about what is conventionally true may not be. This does not seem like a general threat to bivalence at the ultimate level since the possibility of bivalence failure will be limited only to claims about what is conventionally true—an outcome we would expect given the nature of conventional discourse. Containing the indeterminacy to the conventional level, one could even reasonably say in such cases that it is ultimately true that it is indeterminate whether or not something is conventional true or conventionally false. Similarly, just as we find it reasonable to say that it is true that it is neither fictionally true nor fictionally false that Sherlock Holmes wears a size ten shoe (since Sir Conan Doyle did not indicate one way or another), it would be equally reasonable to say that it is ultimately true that it is neither conventionally true nor conventionally false that there is a pot, for example, in cases where the imprecision of the generation principle concerning how pots are to be imagined leaves the matter undecided. Such statements at the ultimate level acknowledge and confine bivalence failure to the conventional level. Such an approach might even help explain some of the matters that the Buddha famously would not provide answers for: perhaps there were no conventional answers to be had. However, Siderits (2009) is concerned not just to preserve bivalence at the ultimate level but at the conventional level as well. If according to Abhidharma Buddhist views bivalence is in fact an important feature of conventional discourse, the Walton-style fictionalist formulation of Buddhist Reductionism can explain how Abhidharma Buddhist can address the problem of conventional vagueness in particular cases. First of all, the principle of bivalence can itself serve as a generation

Buddhist Reductionism, Fictionalism, and Expressibility

357

principle that prescribes that our conventional discourse not allow contradiction, indeterminacy, or incompleteness. This would preserve bivalence at the conventional level as a regulative ideal that would prescribe that those engaging in conventional discourse make more precise the relevant generative principles when vagueness interferes with our imaginative activities. The rules governing conventional discourse are meaningful, possible objects of meta-conventional reflection. As rules, not truths, they are also open to revision in light of pragmatic considerations. If conventional discourse is thought of as a tool that is used by creatures like us for navigating our environment in light of our interests and cognitive limitations, as Siderits often emphasizes, then we are at liberty to collectively adjust the tool when circumstances demand it. This can take the form of making our generation principles more precise, or adding or subtracting rules. Most of the time our conventional rules operate well enough that imprecision is not an obstacle. When it becomes an obstacle, then it can, on the Walton-style formulation of Buddhist Reductionism, be a matter of collective decision how, or if, to adjust the rules of our conventional fiction.

7 The Ultimate Truth About Conventional Truth While generative principles are not themselves truth-apt, they do make it the case that true and false things can be thought or said at the ultimate level about what is conventionally true. Hicks (2010, 2015) explains how on a Walton inspired view we can still say serious, ultimately true things about the things we pretend.7 We can engage in serious discourse about conventional, fictionalist discourse by appealing to what the generation principles determine is to be pretended in that discourse. The kind of objectivity that props and generation principles afford the participants in a fiction, or in the Buddhist case conventional discourse, also allow for this kind of serious, ultimate discourse about what is conventional. This serious discourse does not require one to actually engage in the pretense but only to be sensitive to the way the content of fiction-directed thoughts are fixed by the fictions of which they are part. Furthermore, Hicks argues that while our capacity to think about fictions is tied to our ability to engage in them, we need not necessarily be aware that that is what we are doing. For example, Hicks argues that one cannot think of Sherlock Holmes, the particular character created by Sir Arthur Conan Doyle, except by way of the stories Conan Doyle wrote (2015, 1558). He says, “[f]or an utterance of, e.g., ‘Holmes was a detective’ to express a thought (at least to express the ‘right’ thought), it must be knowingly dependent on a game of make-believe… There is no such thing as slipping into fiction without knowing it” (1561). What Hicks means here is that thinking about the fictional character Holmes can only stem from the fiction of

 The view Hicks defends is not defended by Walton but is developed from his account of fiction.

7

358

L. P. Guerrero

which Holmes is part. Without familiarity with that fiction—that way of pretending—one cannot think about Sherlock Holmes. In this respect, “serious fiction-­ directed thought is a different kind of thought than world-directed thought” (1560); the content of fiction-directed thought is tied to the fictional games of which it is part and so unlike world-directed thought, is mediated by those games. However, thinking about fictional entities does not actually require that one know that what one is doing is engaging with the fiction. It only requires that one be sensitive to the fiction such that one is responsive to what is fictional and what is anti-fictional, i.e., the rules of the fiction. “Knowing” here can be of the implicit kind. In order to grasp the content of a fiction-directed thought, one must be sensitive to the fact that it is fictional, but this primarily involves familiarity with the relevant game and the capacity to pretend in the appropriate way. This kind of analysis also helps differentiate fictional thought from erroneous thought involving, for example, hallucination (1561). While this kind of thought may involve the imagination of something unreal, the content of the imagination is not connected to a fiction, and its normative constraints, the way fictional thought is. So while such thought is imaginary in some sense, since it is not mediated by fiction in the right kind of way, it has a different kind of content than fiction-directed or world-directed thought. Hicks points out that a young children typically can engage in pretense-­dependent thought without being able to distinguish pretense from reality or have the concepts of either (1569). They can effectively engage in pretense and in reality before they can meta-cognitively think about what they are doing. Hicks says that this suggests that “pretending does not require the meta-conceptual ability to judge that one is pretending.” He further explains that, the capacity to engage in self-conscious mature thought about reality—thought involving a concept of reality—presupposes the ability to successfully distinguish those thoughts with a fiction-directed profile from thoughts whose character is about reality: one comes to realize that there is a difference in the character of the two sorts of thought, both of which one already employed. The difference in profile between the two types of thought is not to be identified by adverting to the speaker’s concept of fiction (or reality)—rather, one’s concept of fiction (reality) is to be understood as abstracted from one’s ability to think about fiction (reality), abilities one can possess without possessing concepts of fiction or reality (1570).

Here Hicks argues that meta-cognitive reflection about what and how we are thinking is a separate more sophisticated activity than simply engaging in the thinking. We can think about the world and about fiction without knowing that is what we are doing. In an externalist vein that is in keeping with a general classical Indian understanding of intentionality, we can add that on this view we can be mistaken about what the content of our thoughts is. What makes a cognition a world-directed one or a fiction directed one depends on whether or not the cognition is mediated by fiction, and since this is something the cognitive agent herself may be ignorant about, it is possible that she is engaging in pretense, following cognitively the appropriate imaginative rules, and not know that that is what she is doing. In this way she may come to be mistaken about the content of her thoughts, taking that content to be real when really it is just imagination.

Buddhist Reductionism, Fictionalism, and Expressibility

359

Returning to Abhidharma Buddhism, Hicks account helps to explain how for these traditions it could be that the vast majority of us unenlightened engage in conventional pretense all the time without realizing that is what we are doing. One might object to Buddhist fictionalism on the grounds that it implausibly ascribes a pretending attitude to people engaging in conventional discourse that such people simply do not have. Buddhist texts, such as the Abhidharmakośabhāṣya, describe conventional truth as a kind of truth and ordinary people likewise typically describe the things they believe about the ordinary world as being true. So Buddhist fictionalist accounts may seem to be off base. However, while it is true that texts such as the Abhidharmakośabhāṣya do not talk in terms of pretense and that ordinary people do not typically describe their attitudes as those of pretending, Buddhists do generally maintain that the ordinary person is deeply mistaken about the nature of their own experience. In the Abhidharma context, the unenlightened who are not well versed in Abhidharma are thought to mistakenly think that the conventional objects of their experience are real in a robust realist sense of real. This simple and profound error is taken to be a significant source of suffering that engaging in Abhidharma analysis can help remove. The Walton-style fictionalist account, as understood by Hicks, provides a way to explain the nature of this error and how Abhidharma can be effective at removing it. On the Walton-style view, Abhidharma can be understood as offering the meta-­ cognitive reflection that allows us, the unenlightened, to come to understand that what we are doing when we engage in conventional thought is imagining and then gives us an account of the structure of that imagining. Abhidharma provides an analysis of the generative principles that have been regulating our participation in conventional discourse so that we can better understand what is real and what is merely imagined and thereby learn to adopt the appropriate attitudes towards the different kinds of cognitions we have. On this account, the purpose of Abhidharma analysis is precisely to make apparent to those unenlightened who mistake fictions for reality that the content of their conventional imaginings are merely fiction, and thereby make clear that they ought not to be believed but only pretended. We have always been able to imagine what convention prescribes we ought to imagine; the mistake we make is believing these things we only imagine. The enlightened can still imagine alongside the rest of us when they engage in conventional thought and discourse, but such enlightened people will see the imagination for what it is and adopt the appropriate pretense attitude, reserving belief for those cognitions that actually reflect the underlying reality of the dharmas.

8 Conclusion Putting Buddhist Reductionism forward as an ultimate truth is in keeping with the way in which texts such as the Abhidharmakośabhāṣya describe their Abhidharma projects as reflective of the deep insight of the Buddha. A Walton-style fictionalist formulation of Buddhist Reductionism still captures the basic asymmetrical

360

L. P. Guerrero

dependence relation of conventional truth on ultimate truth that Siderits was rightly concerned to capture. Like Siderits’ version of Buddhist Reductionism, the formulation defended here articulates the project of Abhidharma to make clear that our conventional understanding of the world is a fictional projection grounded in a more fundamental reality. However, unlike the Sider-style formulation, the Waltoninspired version of Buddhist Reductionism can be expressed as the ultimate truth that Abhidharma traditions take it to be.

 References Benovsky, Jiri. 2017a. Buddhist Philosophy and the No-Self View. Philosophy East and West 67 (2): 545–553. ———. 2017b. The Self, Agency, and Responsibility: A Reply to Mark Siderits. Philosophy East and West 67 (2): 558–564. Bronkhorst, Johannes. 2009. Buddhist Teaching In India. Boston: Wisdom Publications. Collins, Steven. 1982. Selfless Persons. Cambridge: Cambridge University Press. Cox, Collett. 2004. From Category To Ontology: The Changing Role of Dharma In Sarvāstivāda Abhidharma. Journal of Indian Philosophy 32: 543–597. Crittenden, Charles. 1981. Everyday Reality as Fiction – A Mādhyamika Interpretation. Journal of Indian Philosophy 9: 323–333. D’Amato, Mario. 2013. Buddhist Fictionalism. Sophia 52: 409–424. Davids, Rhys T.W., trans. 1890. The Questions of King Milinda. Vol. 35. The Sacred Books of the East. Oxford: Clarendon Press. Duerlinger, James. 2003. Indian Buddhist Theories of Persons: Vasubandhu’s “Refutation of the Theory of a Self”. London: RoutledgeCurzon. Ganeri, Jonardon. 2012. The Concealed Art of The Soul. Oxford: Oxford University Press. Garfield, Jay L. 2006. Reductionism and Fictionalism: Comments on Siderits’s Personal Identity and Buddhist Philosophy. APA Newsletter on Asian and Asian-American Philosophers and Philosophy 6 (1): 1–7. Goodman, Charles. 2005. Vaibhāṣika Metaphoricalism. Philosophy East and West 55 (3): 377–393. Guerrero, Laura. 2018. Buddhist Global Fictionalism? Ratio 31 (4): 424–436. Hicks, Michael. 2010. A Note On Pretense and Co-Reference. Philosophical Studies 149: 395–400. McDaniel, Kris. 2019. Abhidharma Metaphysics and the Two Truths. Philosophy East and West, 69 (2): 439–463. ———. 2015. Pretense and Fiction-Directed Thought. Philosophical Studies 172: 1549–1573. Ronkin, Noa. 2005. Early Buddhist Metaphysics: The Making of a Philosophical Tradition. London: RoutledgeCurzon. Sauchelli, Andrea. 2016. Buddhist Reductionism, Fictionalism about the Self, and Buddhist Fictionalism. Philosophy East and West 66 (4): 1273–1291. Sider, Theodore. 1999. Presentism and Ontological Commitment. Journal of Philosophy 96 (7): 325–347. Siderits, Mark. 2006. Replies to Garfield, Taber and Arnold. APA Newsletter on Asian and Asian-American Philosophers and Philosophy. http://c.ymcdn.com/sites/www.apaonline.org/ resource/collection/2EAF6689-­4B0D-­4CCB-­9DC6-­FB926D8FF530/v06n1Asian.pdf. ———. 2009. Is Reductionism Expressible? In Pointing At The Moon: Buddhism, Logic, Analytic Philosophy, ed. Mario D’Amato, Jay L.  Garfield, and Tom J.F.  Tillemans, 57–69. Oxford: Oxford University Press. ———. 1991. Indian Philosophy of Language: Studies in Selected Issues. Dordrecht: Kluwer Academic Publishers.

Buddhist Reductionism, Fictionalism, and Expressibility

361

———. 2015. Personal Identity and Buddhist Philosophy. 2nd ed. Ashgage: Hampshire. ———. 2016. Thinking on Empty: Madhyamaka Anti-Realism and Canons of Rationality. In Studies in Buddhist Philosophy, ed. Jan Westerhoff, 24–38. Oxford: Oxford University Press. Vasubandhu. 1990. Abhidharmakośabhāṣya. Translated by Leo Pruden and Louis de La Vallée Poussin. Vol. 3, 4 vols. Berkeley: Asian Humanities Press. Walser, Joseph. 2005. Nāgārjuna in Context: Mahayana Buddhism and Early Indian Culture. New York: Columbia University Press. Walton, Kendall. 1990. Mimesis As Make Believe. Cambridge: Harvard University Press. ———. 1993. Metaphor and Prop Oriented Make-Believe. In Fictionalism in Metaphysics, ed. Mark Eli Kalderon, 65–87. Oxford: Oxford University Press. Williams, Paul. 2000. Buddhist Thought: A Complete Introduction to the Indian Tradition. London: Routledge.

The Presupposition Strategy and the Two Truths John Spackman

1 Introduction In his insightful and influential writings on Madhyamaka thought, Mark Siderits has sought to develop what he calls an anti-realist interpretation of the ultimate truth of emptiness, according to which there is no ultimate truth.1 One of the central tools on which he relies in this endeavor is what might be called the presupposition strategy. The presupposition strategy is an approach to understanding the catuṣkoṭi or tetralemmas, and other related statements, that appear in Madhyamaka works, most importantly Nāgārjuna’s Mūlamadhyamakakārikā (MMK). This strategy has, indeed, played an important role in the work of a wide range of other interpreters, both ancient and modern, from Candrakirti (1913), to Ruegg (1977), Garfield and Priest (2009), and Westerhoff (2009). In their negative instances, the catuṣkoṭi have the form of a rejection of four statements, p, not p, both p and not p, and neither p nor not p. The presupposition strategy proposes that, appearances notwithstanding, we can view the rejections of these four statements as logically consistent if we regard each of the statements as in some way semantically deficient, due to the fact that each makes a false presupposition. The presupposition in question, of course, is that there are entities that  See for instance Siderits (2007, 2016a, b, 2019).

1

This essay is offered as an expression of gratitude to Mark Siderits, for the tremendous amount that I have learned from his writings, and for the high standard he has set for analytical approaches to Buddhist philosophy, which has been so beneficial for the field. J. Spackman (*) Middlebury College, Department of Philosophy, Middlebury, VT, USA e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_19

363

364

J. Spackman

have svabhāva, an independent existence or intrinsic nature – what I will call the svabhāva-presupposition.2 The central claim of the MMK is precisely that this presupposition is false, that all things are empty of svabhāva. For Siderits, in particular, the presupposition strategy plays a crucial role in establishing the anti-realist claim that for Nāgārjuna no statements are ultimately true, since it allows him to say that at the level of ultimate truth, neither p, nor not p, nor both, nor neither, are true. The four statements of the catuṣkoṭi are meaningless, and thus do not conflict. This strategy for understanding the catuṣkoṭi has much to recommend it. But recent discussions of the strategy have done little to engage with the extensive work that has been done on the nature of presupposition within linguistics and philosophy in recent years, and this work raises challenges to the presupposition strategy that have not been widely addressed. Some have argued that presupposition is not a real linguistic phenomenon at all (Lycan 2000; Kempson 1975). Others have argued that rather than viewing presupposition as semantic, as is generally assumed by the presupposition strategy, we should see it as a pragmatic phenomenon (Stalnaker 1974; Wilson and Sperber 1979). Either way, the ability of the presupposition strategy to resolve the apparent inconsistencies among the statements of the catuṣkoṭi, and so too Siderits’ anti-realist account of emptiness, is threatened. My aim in this essay is twofold, first to draw attention to these challenges as issues that need to be more fully addressed by advocates of the presupposition strategy, and second, to consider the viability of the strategy in the face of these challenges. I will argue that there are in fact good reasons, from a Buddhist perspective, to view the svabhāva-presupposition as fundamentally pragmatic rather than semantic in nature. But I will also argue that this position is compatible with holding that the presupposition strategy can indeed resolve apparent inconsistencies in the catuṣkoṭi. To see how this is possible will require a consideration of the nature of emptiness and the two truths, and I will maintain in particular that what I call a nonconceptualist account of emptiness provides a better model for preserving the viability of the presupposition strategy than an alternative interdependence account of emptiness.3 As we shall see, the nonconceptualist account has implications for our understanding of the relation between the two truths. Rather than following Siderits in adopting the anti-realist view that there is no ultimate truth, the nonconceptualist account favors a view which takes seriously the tenor of MMK 24:18, that the ultimate truth is the conventional truth.

 The term svabhāva has received a variety of translations, from “essence” (Garfield 1995) to “intrinsic nature” (Siderits and Katsura 2013) to (in certain uses anyway) “substance” (Westerhoff 2009). Given the diversity of these translations, I will here generally leave the term untranslated. 3  In Spackman (2014) I referred to the account I support as a “conceptualist” interpretation of emptiness, because, as we shall see, it views Nāgārjuna as making a claim about the concept of existence. But it seems more perspicuous to call it, as I do here, a nonconceptualist account. 2

The Presupposition Strategy and the Two Truths

365

2 The Presupposition Strategy The historical forerunners of Nāgārjuna’s use of the catuṣkoṭi lie at least in part, no doubt, in the Buddha’s responses to the indeterminate questions (avyākṛta) in the Nikāyas, as for instance in the Aggivacchagotta Sutta when he answers the question of whether a Tathāgata exists after death by saying that he holds neither that he exists after death, nor that he does not, nor both, nor neither (Bodhi and Ñānamoli 1995, 590–94). The catuṣkoṭi are similar in form, and raise similar interpretive issues.4 One of the most succinct instances of a catuṣkoṭi occurs at MMK 22:11: “It is empty” is not to be said, nor “It is non-empty,” nor that it is both, nor that it is neither; [“empty”] is said only for the sake of instruction.5 Jan Westerhoff (2009, 67-80) argues plausibly that if the tetralemma is taken to have propositional form, the verse can be viewed as a rejection of four alternatives (taking A to be “‘Empty’ should be asserted”): 1. A 2. ~A 3. A ∧ ~A 4. ~ (A ∨ ~A) As Westerhoff points out, if the rejection of each of these alternatives is taken to be standard, truth-functional negation, and the negation operator in the second alternative is also taken to be truth-functional negation, then the rejection of the first two alternatives clearly generates a contradiction. The truth-functional negation of (1) is ~A, and the truth-functional negation of (2) is ~~A, which by Double Negation Elimination is equivalent to A. A second interpretive difficulty is that, again assuming that both the rejection of the alternatives and the negation internal to the alternatives is truth-functional negation, the rejections of alternatives 3 and 4 would seem to logically equivalent. Applying DeMorgan’s Law and Double Negation Elimination, both of these rejections would be equivalent to A ∨ ~A. It might be argued that this result is unlikely to be what Nāgārjuna intended, and also that it saddles the interpreter with the question of why the author would include an alternative that is logically superfluous. The central idea of the presupposition strategy is to seek to resolve such interpretive difficulties by claiming that each of the four alternatives makes a false presupposition, and that statements with false presuppositions are subject to semantic or, sometimes, pragmatic, anomaly. It is helpful to distinguish several different variants

 The literature on the catuskoṭi is extensive. See in particular Ruegg (1977), Westerhoff (2009), Garfield and Priest (2003), Tillemans (2009). 5  śūnyam iti na vaktavyam aśūnyam iti vā bhavet/ubhayaṃ nobhayaṃ ceti prajñaptyarthaṃ tu kathyate. In what follows I will employ the translation of Siderits and Katsura (2013) unless otherwise noted, with emendations of my own where appropriate. 4

366

J. Spackman

of the strategy, however, which identify different explanations of why false presuppositions lead to anomalies. Siderits suggests that any statement with a false presupposition is meaningless, the implication presumably being that contradictions cannot arise between meaningless statements. Thus he says, in relation to the Buddha’s statements about a Tathāgata’s existence after death noted above, that each of the statements presupposes that there is such a thing as a person, and “any statement about the enlightened person lacks meaning at the level of ultimate truth” (Siderits 2007, 72). A distinct claim, also suggested by Siderits, is that any statement with a false presupposition “lacks a truth-value” (Siderits 2016c, 117), that bivalence fails for such statements. Westerhoff, by contrast, rejects the view that statements with a false presupposition lack a truth-value – he argues that Nāgārjuna accepts the Law of the Excluded Middle for the relevant type of negation (Westerhoff 2009, 77) – and considers two different versions of the presupposition strategy. Both versions draw on the well-­ known distinction within the Indian philosophical tradition between two kinds of negation, prasajya negation and paryudāsa negation. Prasajya negation, as Westerhoff presents it, is presupposition-cancelling propositional-negation, while paryudāsa negation is presupposition-preserving term-negation (Westerhoff 2009, 69–71). If, for instance, I speak of someone as a “non-brahmin”, I negate the term “brahmin”, and if this negation is paryudāsa, then I presuppose that the person in question belongs to one of the other three castes. On the other hand, I might say “This is not a brahmin” with the intention of negating the whole proposition (“This is a brahmin”), and without presupposing that the person in question is a member of any of the castes. In this case the negation is prasajya. Similarly, viewed as a paryudasa negation, the statement ‘The number seven is not green” presupposes that numbers are the kinds of things that can have colors, so that any number is either green or not green. Viewed as a prasajya negation, though, the statement makes no such assumption. The distinction between prasajya negation and paryudāsa negation is often presented in terms of an ambiguity with regard to the negation operator. The statement “x is not a brahmin” may be understood to mean either x is a non-brahmin, or It is not the case that x is a brahmin. The first of these is sometimes said to involve an internal negation, and the latter an external negation, where this distinction is drawn in terms of an ambiguity in the scope of the negation operator. Expressed formally, the statement may be understood as having either the form ∃x(~Bx) or the form ~∃x(Bx). One version of the presupposition strategy that Westerhoff considers seeks to avoid the apparent problems for the tetralemma simply by appeal to the contrast between prasajya and paryudāsa negation. If we view the overall rejection of each of the alternatives as prasajya negation, and the negation internal to alternative 2 (~A) as paryudāsa negation, then the apparent contradiction between alternatives

The Presupposition Strategy and the Two Truths

367

(1) and (2) disappears.6 If for instance we take “The number seven is green” as an instance of (1) and “The number seven is not green” as an instance of (2), the conjunction of the prasajya negations of these statements denies that the number seven is either green or not green. There is no contradiction here  – the conjunction is simply true  – because the prasajya negation amounts to a rejection of the false assumption made by both statements that numbers are the kind of things that can have colors. Westerhoff attempts to skirt the apparent difficulty that alternatives 3 and 4 seem to be logically equivalent in a similar vein. He takes the outer negation in (4) ~(A ∨ ~A) to be prasajya negation, and the inner negation to be paryudāsa, so that (3) and (4) say quite different things. The result is that on this approach, all four alternatives of the catuṣkoṭi can simultaneously be true. I take the first three versions of the presupposition strategy that we have considered to be semantic, in the sense that they seek to avoid the problems with the negative tetralemma by appeal to facts about the meaning or truth values of its constituent statements.7 A final variant, advocated by Westerhoff and also discussed by Siderits, is a pragmatic one. According to this version, the outer negations of each of the four alternatives should be viewed as illocutionary negations. Illocutionary negation is a type of illocutionary force, specifically the non-assertion of a proposition.8 I can, on the one hand, assert the negation of the proposition “The chair is red” by asserting “The chair is not red.” If I engage in illocutionary negation toward this proposition, by contrast, I simply do not assert it. There are a variety of reasons why one might engage in illocutionary negation toward a given proposition, but Westerhoff suggests that among these is the fact that the proposition makes a certain presupposition that one wishes to avoid. On Westerhoff’s view, Nāgārjuna advocates illocutionary negation of the four alternatives of the tetralemma because it is appropriate to apply prasajya negation to them (Westerhoff 2009, 78). Notice, as Siderits points out, that if I assert a proposition which should have been subject to illocutionary negation, the result is an inappropriate or infelicitous utterance, not a false one (Siderits 2019). That a sentence is subject to illocutionary negation is a pragmatic rather than a semantic fact about it; this is thus a pragmatic version of the presupposition strategy.

3 The Pragmatic Turn in Presupposition Theory What are we to make of these appeals to the notion of a presupposition in order to dispel the apparent inconsistencies of the catuṣkoṭi? Contemporary theorizing about presupposition has its origin, of course, in the familiar debate between Strawson  Ruegg notes that both Candrakīrti and Bhāvaviveka view the outer negations involved in the catuṣkoṭi concerning causation in MMK 1:1 as prasajya negation (1977, 4). 7  The third version can be considered semantic in this sense in that the contrast between prasajya and paryudāsa negation is used to resolve the apparent contradiction resulting from the truth values of the first two alternatives. 8  Searle (1969), 31–33. 6

368

J. Spackman

(1950, 1964) and Russell (1905, 1957). In response to Russell’s view that the proposition “The King of France is bald,” due to its false presupposition that there is a present King of France, is significant but false, Strawson argues that this statement lacks a sense, and that for this reason, the question of its truth or falsity does not arise. In claiming that statements with false presuppositions are meaningless or lack a truth-value, Siderits would thus appear to be siding, on this issue at least, with Strawson. In the wake of the Strawson-Russell debate, a great variety of linguistic phenomena, beyond the referential expressions that concerned them, were taken by many linguists and philosophers to fall under the umbrella of presupposition.9 To take just a few of these: Referential expressions 5a. The King of France is 29 years old. 5b. The King of France is not 29 years old. 6. (Presupposition) There is a King of France. Factives 7a. Ariel knows that she is the tallest in the class. 7b. Ariel does not know that she is the tallest in the class. 8. (Presupposition) Ariel is the tallest in the class. Cleft Sentences 9a. It was James who asked the first question. 9b. It wasn’t James who asked the first question. 10. (Presupposition) Someone asked the first question. In each of these cases, both the affirmative sentence and the corresponding negative sentence have been held to have the indicated presupposition. Presuppositions are said, in this sense, to project from embeddings under negation. Projection is commonly taken to be the hallmark of presupposition, and presuppositions have generally been held to project from operators other than negation, as well (e.g. modal and conditional operators). Early discussions of presupposition tended to assume that it is a semantic phenomenon, but since at least the 1970s there has been a marked turn toward the analysis of presupposition in terms of various pragmatic features. Before we consider some of the arguments presented in favor of more pragmatic accounts, let’s clarify more fully what’s involved in the semantic conception of presupposition. According to semantic accounts, presupposition is to be analyzed in terms of the content or truth-conditions of the sentences in question. As we have seen, on Strawson’s account, for instance, if a statement S presupposes a statement P, its negation ~S presupposes P as well, and in cases in which P is false, both S and ~S will lack a truth-value. Presupposition is thus, on this view, a semantic relation that  For an overview of some of these cases, see Simons (2006), Beaver and Geurts (2011).

9

The Presupposition Strategy and the Two Truths

369

is distinguished sharply from entailment, and indeed is incompatible with it. Consider for instance: 11. The King of France is less than 30 years old. Sentence (5a) (“The King of France is 29  years old.”) entails sentence (11), but unlike in cases of presupposition, its negation (5b) clearly does not. Entailment is a relation in which the truth of a second sentence S2 follows necessarily from the truth of a first sentence S1, but nothing at all follows from the falsity of S1. By contrast, in presupposition, viewed as a semantic relation, both the truth and the falsity of the presupposing sentence S necessitate the truth of the presupposed sentence P. The pragmatic critique of the semantic account has come from a number of angles. Some proponents of a more pragmatic account are what we might call eliminativists, arguing that in fact presupposition does not constitute a distinct linguistic phenomenon, that it is analyzable in terms of a variegated set of more familiar linguistic features such as entailment and conversational implicature.10 Others, who we might call reductionists, allow that presupposition is a real phenomenon, but hold that it is a purely pragmatic one, a feature of speakers, and only derivatively of sentences, propositions, or speech acts.11 Let’s consider some of the arguments for these two perspectives. One argument that has played a central role in many eliminativist critiques of the semantic account is that, contrary to what we assumed above, presuppositions are in fact cancelable in many contexts – that is, a sentence’s presupposition can be denied without creating semantic anomaly. The line of thought here is perhaps most clearly expressed by Ruth Kempson (1975). As Kempson points out, even Strawson, in his (1964), acknowledged that in some contexts presuppositions are cancelable, but she takes this to hold for all presuppositions. On her view, a sentence such as 12. The King of France didn’t visit the exhibition – France hasn’t got a King. might be viewed as simply true, contrary to what one would expect if “The King of France didn’t visit the exhibition” presupposed the existence of the king. She would thus argue that while “The King of France is 29 years old” entails “There is a King of France,” its negation, “The King of France is not 29 years old,” does not, since it could be true while “There is a King of France” is false. Since in order for a sentence S to semantically presuppose P, both S and ~S must necessitate P, the relation between a referential expression and the corresponding existence claim cannot be one of semantic presupposition (Kempson 1975, 85–89). Now Kempson acknowledges that a sentence such as (12) will generally be heard as infelicitous. “The King of France didn’t visit the exhibition,” she admits, is more naturally interpreted as an internal negation, which would entail the existence of the French king, rather than as an external negation, which would not. But she argues  See e.g. Kempson (1975), Boër and Lycan (1976), Lycan (2000).  See e.g. Stalnaker (1974, 2002), Wilson and Sperber (1979). Simons (2006) refers to accounts of both of these kinds of accounts together as “reductionist,” but it seems useful to distinguish the reductionist from the eliminativist. 10 11

370

J. Spackman

that the business of semantics is to predict all possible interpretations of each wellformed sentence in the language, not to determine which interpretations are more natural or likely (Kempson 1975, 80–81). The latter job falls to pragmatics. She therefore maintains that the relation between the statement “The King of France is not 29 years old” and the statement “There is a King of France” is not one of entailment or presupposition, but rather generalized conversational implicature.12 The reductionist account of presupposition differs from the eliminativist account in that it acknowledges the reality of presupposition, but seeks to reduce it to the status of a fundamental pragmatic phenomenon. Perhaps the most influential version of such a view is Robert Stalnaker’s. On Stalnaker’s formulation, while a semantic account of presupposition views it as a relation between sentences, propositions, or speech acts, a pragmatic account sees presuppositions as properties of speakers, as something akin to their background beliefs (Stalnaker 1974, 50). Presuppositions are viewed as features of sentences, propositions, or speech acts only derivatively from their role as psychological states of speakers. A speaker’s presuppositions are, more specifically, those beliefs which she takes to constitute the “common ground” for the conversation. This common ground  – speaking roughly at first – is constituted by the beliefs shared among all parties to the conversation (Stalnaker 2002, 704). Thus for instance, on this view the sentence “The King of France is 29 years old” – and also “The King of France is not 29 years old” – can be uttered appropriately only if the speaker believes that it is part of the common ground that there is a King of France at the time of the utterance. Stalnaker complicates this picture, however, in order to account for a variety of examples in which the common ground of a conversation may diverge from the common beliefs of its participants. He thus defines the common ground by reference to a notion of acceptance which is more general than the concept of belief, encompassing for instance belief, assumption, and acceptance for the purposes of an argument. The core idea is that “to accept a proposition is to treat it as true for some reason” (Stalnaker 2002, 716). For instance, a speaker might accept as true for purposes of conversation, without believing, his interlocutor’s presupposition that the baby girl his arms is a boy. Stalnaker has offered a number of general arguments that support a pragmatic account of presupposition over a semantic account.13 Most importantly for our purposes, the pragmatic account is more sensitive to contextual variation in presuppositions. For instance, the pragmatic view allows one to hold that if someone said, in one context, “My cousin isn’t a boy anymore,” they might be taken to assert that their cousin is grown up, and to presuppose that he is male; while in another context, the speaker might be asserting that their cousin is undergoing gender transition, and presupposing that the cousin is young. On the semantic account, by contrast, one

 See Böer and Lycan (1976) for a compelling account of the reasoning underlying this kind of conversational implicature. 13  See e.g. Stalnaker (1974), 53–55. 12

The Presupposition Strategy and the Two Truths

371

would have to view the sentence as ambiguous, thus yielding a more cumbersome semantic theory. I want finally to consider a third challenge, one that might be taken to support a pragmatic account of the svabhāva-presuppositions that concerned Nāgārjuna in particular. This challenge focuses not on the nature of presupposition, but on the consequences of presupposition failure. Several researchers have recently maintained that while there are cases in which presupposition failure is “catastrophic,” in the sense that it undermines the entire assertive enterprise, as Strawson held, there are others in which it is “non-catastrophic,” in which assertion is actually preserved.14 The challenge for the presupposition strategy is that, on the face of it, Nāgārjunian examples in which there is a failure of svabhāva-presuppositions look more like the non-catastrophic cases than the catastrophic ones. But if statements with false svabhāva-presuppositions actually succeed in making an assertion, it seems that they should be straightforwardly assessable as either true or false, rather than being meaningless or non-bivalent. Why hold that there are some cases in which presupposition failure is non-­ catastrophic? We can distinguish between what Yablo, following Strawson, calls “radical failure of the existence presupposition,” in which “there is just no such particular item at all” to which the speaker aims to refer (Strawson 1964, 103), and “non-radical presupposition failure,” where although the description used is not uniquely satisfied by the referent, nonetheless the speaker does have a unique item in mind. “The King of France is bald” is an example of radical reference failure. One example of non-radical presupposition failure is Keith Donnellan’s (1966) well-known counter-example to Strawson. Donnellan claims that if someone says “The man drinking a martini is a famous philosopher,” and it turns out that the indicated man in fact has water in his glass, we should nonetheless regard the statement as true (assuming that he is in fact a famous philosopher), rather than neither true nor false. This is one kind of case that Yablo views as involving non-catastrophic presupposition failure, for the assertive enterprise is not entirely disrupted. Now why might we take the failure of svabhāva-presuppositions to be non-­ catastrophic? The key consideration involves Nāgārjuna’s distinction between ultimate and conventional truth. The ultimate truth I take to be, roughly, that all things are empty of svabhāva; false statements at the ultimate level of discourse, by contrast, are claims that entail that things exist with svabhāva. The conventional truth is aptly expressed by Garfield as being “the truth about things as they appear to accurate ordinary investigation;” correspondingly, falsehoods at the conventional level are statements that lack this accuracy (1995, 212). It is clear that for Nāgārjuna, at the level of ultimate truth, nothing exists and no facts obtain, for nothing exists with svabhāva. But it is also accepted by all but proponents of the nihilist reading of Nāgārjuna that at the level of conventional truth, we can say truly that things exist

14

 See for instance Yablo (2006), von Fintel (2001), Bezuidenhout (2016).

372

J. Spackman

and facts obtain.15 Thus, the statements of ordinary, deluded human beings, which presuppose that things exist with svabhāva – for example, “The pear tree is blooming” – although false at the ultimate level, nonetheless express conventional truths. If this is the case, however, then it could be argued that the presupposition failure in question is non-catastrophic rather than catastrophic. What causes the failure in the pear tree example is not that the tree in question does not exist at all, for conventionally, it does exist; it is rather that the tree does not exist under a certain description, that is, it does not exist with svabhāva. The assertive enterprise is thus not entirely demolished, for the statement does make an assertion, namely that the pear tree is in bloom. So contrary at least to the semantic versions of the presupposition strategy, on this argument it looks like statements subject to presupposition failure are not semantically deficient. Indeed such an argument might be taken to support a pragmatic view of svabhāva-presuppositions more generally. For on this view, svabhāva-presuppositions are not simply part of the meaning of all existence claims, since existence claims at the conventional level do not carry this presupposition. Svabhāva-presuppositions are context dependent, in force in ultimate but not conventional contexts, and context dependence, for Stalnaker anyway, is a mark of a pragmatic view of presupposition.

4 Assessing the Pragmatic Challenges How are we to assess the ability of these three arguments to undermine the presupposition strategy? Let me point out first that since each of these arguments in one way or another analyzes presupposition in terms of pragmatic phenomena, one natural response would be to hold that while they undermine the semantic versions of the presupposition strategy, they are compatible with the pragmatic version, that is, the illocutionary negation account. The illocutionary negation account, after all, accepts that the only effects of presupposition failure on the statements of the tetralemma are pragmatic rather than semantic, at the level of felicity and infelicity rather than truth and falsehood. But while the illocutionary negation view may provide an insightful account of some aspects of speakers’ attitudes toward the statements of the tetralemma, I do not think that it can be the whole story. Recall that on this view, the reason it is appropriate to adopt illocutionary negation toward these statements is because they make a presupposition we wish to avoid, namely that there are things that have svabhāva. But what is the status of the statement that things have svabhāva – is it true or false? This is, of course, the subject of MMK 22:11: “‘It is empty’ is not to be said, nor ‘It is non-empty,’ nor that it is both, nor that it is neither….” That is, the statement that things are empty of svabhāva is itself subject to the negative

 The nihilist reading holds that nothing exists either at the ultimate or at the conventional level. See Wood (1994). For an argument against this reading, see Spackman (2014). 15

The Presupposition Strategy and the Two Truths

373

tetralemma. But according to the illocutionary negation account, each of the statements of the negative tetralemma is subject to illocutionary negation. We can ask, then, for the reason why we should adopt this illocutionary attitude toward the statements concerning emptiness, and presumably the answer will be that they make the false presupposition that emptiness itself is svabhāva. But then this presupposition itself will be subject to the negative tetralemma – and so on. We are off on an infinite regress, in a futile search for the reason why we should adopt an attitude of illocutionary negation toward the statements of the negative tetralemma. The upshot is that by itself, the illocutionary negation account cannot provide us with such a reason. And if it cannot give us any reason why speakers should adopt such an attitude toward these statements, it hardly amounts to a plausible means of avoiding the apparent inconsistencies between them.16 I will thus assume that if the presupposition strategy is salvageable, the semantic versions of it are more plausible than this pragmatic one. Do the arguments we have considered undermine these semantic versions? It seems to me that Buddhists have good reason to accept that svabhāva-presuppositions are (contra the eliminativist) real linguistic phenomena, and that they are in fact fundamentally pragmatic rather than semantic phenomena, in the specific sense delineated by Stalnaker, that they are properties not of sentences or speech acts, but of speakers. But I will also argue that their pragmatic nature does not preclude their producing anomalies at the semantic level, the level of truth values. It has not been widely discussed that the svabhāva-presuppositions are, in certain respects, quite different from most of the presuppositions discussed by linguists. Unlike the claim that “John knows p” presupposes that p, the claim that certain statements have svabhāva-presuppositions is a theoretical claim that Nāgārjuna seeks to argue for, and which his opponents resist.17 More importantly for our purposes, though, svabhāva-presuppositions are pervasive, in the sense that they are widespread across types of utterance and speech-act. Just how widespread they are is a matter for debate. Clearly they are not limited to any specific types of semantic construction, as are the presuppositions of, for instance, factives and cleft sentences. It might be argued that they are carried only by a specific type of speech act, namely assertion. Recall that one of the central examples of the negative tetralemma, MMK 22:11, enjoins us not to engage in a certain speech act (vaktvavyam) with regard to  It will not do to respond to this argument by suggesting that far from undermining the illocutionary negation approach, the infinite regress described here is precisely what one would expect, because this regress is just another way of describing the emptiness of emptiness, that emptiness itself does not have an intrinsic nature. The regress described here does not concern the nature of emptiness at all, but rather whether the illocutionary negation approach can supply any reason for why a speaker should refrain from asserting the four alternatives of the catuṣkoṭi. The objection is that if it is not possible for this approach to specify such a reason, then it cannot explain why a speaker has any more reason to adopt illocutionary negation in regard to the statements of the catuṣkoṭi than in regard to conventionally true statements like “the earth orbits the sun,” which are not subject to catuṣkoṭi. I am grateful to an anonymous reader for raising this challenge. 17  They are thus somewhat akin to what Böer and Lycan call cases of “theoretical implication.” Böer and Lycan (1976), 11. 16

374

J. Spackman

“It is empty,” “It is not empty,” and so on, and vaktavyam has sometimes been translated – not implausibly – as “assertion.”18 So it might be argued more generally that all and only assertions carry svabhāva-presuppositions. I think there is good reason, however, to suppose that svabhāva-presuppositions are even more pervasive than this. If a deluded person says “I promise you I’ll be home by six,” or commands “You’d better be home by six,” presumably, if he is thinking of the persons and actions in question as having svabhāva, these speech acts carry svabhāva-presuppositions. Similarly, if a deluded person imagines or thinks hypothetically about Santa Claus, without actually believing or asserting that he exists, it seems he will be thinking of Santa as something with svabhāva. If this is so, a very wide range of utterance types and speech acts should be regarded as at least capable of carrying svabhāva-presuppositions. It thus seems better to think of this kind of presupposition not as a property of any particular utterance-type or speech-act, but as an attitude or state of mind that subjects bring to bear on various speech-acts and utterance types – that is, as properties of speakers rather than sentences or speech acts. This way of thinking of svabhāva-presuppositions is supported by the notion, developed especially in Yogācāra thought, that the seeds of ignorance lie in deep unconscious impulses in the mind. If svabhāva-presuppositions are indeed fundamentally pragmatic rather than semantic in nature, this establishes a framework for thinking about them that is different from the semantic approach. However, this framework need not preclude viewing svabhāva-presuppositions as semantic in Stalnaker’s derivative sense, or, in principle, as having effects on the truth-conditions of sentences with those presuppositions. In this derivative sense, to speak of a sentence S as having a svabhāva-­ presupposition p is a kind of shorthand for talking about speakers presupposing that p, in cases where p is not merely a conversational assumption but forms a part of the meaning of S. Indeed Stalnaker explicitly allows that viewing presuppositions as pragmatic in his sense is perfectly consistent with these presuppositions sometimes, if not always, being explained in terms of semantic presuppositions (in the derivative sense) (Stalnaker 2002, 702).19 This conception of svabhāva-presuppositions as fundamentally pragmatic also provides us with the resources to respond to the argument for eliminativism concerning presuppositions. Kempson’s argument, recall, was that contrary to what is required for semantic presupposition, a negative sentence like “The King of France is not 29 years old” does not entail “There is a King of France,” and that although this sentence is more naturally heard as an internal negation, which entails the existence of the King, rather than as an external negation, which does not, this fact falls within the domain of pragmatics rather than semantics. But Kempson’s main target,  See e.g. Garfield (1995), 280; Westerhoff (2009), 73.  Note that the direction of explanation suggested here by Stalnaker might not be appropriate for cases involving svabhāva-presuppositions, since it seems implausible that subjects hold svabhāva-­ presuppositions because of the linguistic rules governing any particular locutions. But this direction of explanation does not seem required in order to view svabhāva-presuppositions as semantic in the derivative sense. 18 19

The Presupposition Strategy and the Two Truths

375

throughout her argument, is the notion that presupposition is a purely semantic phenomenon, one insulated from pragmatic processes. Once we acknowledge that presupposition is, from the ground up, the product of pragmatic processes, it becomes possible to recognize that it may have consequences at the semantic level, the level of meaning and truth-conditions. In particular, even if we accept that the ambiguity between the internal negation reading and the external negation reading has a pragmatic source, this need not prevent us from viewing it as having consequences at the level of the truth-conditions of the negated sentence. Regardless of the source of the ambiguity, the fact remains that on the internal negation reading – which is the most natural one – “The King of France is not 29 years old” does entail that there is a King of France. Thus in this case, both S and the most natural reading of ~S, entail P. If presuppositions can be viewed as semantic in this derivative sense, presupposition failure must inevitably, in order to maintain logical consistency, carry with it some sort of semantic deficiency. I cannot here argue fully for which of the types of semantic anomaly we considered earlier is most plausible – that is, whether sentences with failed presuppositions are best viewed as meaningless, non-bivalent, or subject to presupposition-canceling negation. But my inclination is to think the last of these is the most straightforward to support, given that the statements of the catuṣkoṭi certainly seem to be meaningful, and given the apparent evidence of Nāgārjuna’s support for the law of the excluded middle. Viewing the four negative statements of the catuṣkoṭi as simultaneously true, as the presupposition-canceling negation approach does, also makes sense of the notion that what the negative tetralemma expresses is the ultimate truth. There are, however, other potential barriers to viewing svabhāva-presuppositions as semantic in even this derivative sense. If the third challenge considered above is correct, and statements with failed svabhāva-presuppositions can be seen as making assertions, perhaps svabhāva-presuppositions are best seen as wholly pragmatic after all. Let’s consider this argument in more detail, then.

5 Nonconceptualist and Interdependence Accounts of Emptiness There is one kind of reading of Nāgārjuna according to which, as I understand it, the non-catastrophic account of the failure of svabhāva-presuppositions gets things just right. On this reading, we should view Nāgārjuna as holding precisely that the failure of svabhāva-presuppositions rules out assertion at the ultimate level, but allows it at the conventional level. One type of interpretation of Nāgārjuna that takes this line is based on what we might call an interdependence account of emptiness.20 I would argue, however, that the issue raised here represents a more significant

 See e.g. Garfield (1995), Garfield and Samten (2006). For other criticism of interdependence accounts of emptiness, see Siderits 2007, 200–202, Siderits (2016b), 40n, and Siderits (2019). 20

376

J. Spackman

challenge to the viability of the presupposition strategy than the interdependence account would allow. Although the interdependence view does have the resources to defend the presupposition strategy, it can only do so at the cost of a misconceived view of the relation between ultimate and conventional truth. We can achieve a better view of this relation, and a better model for preserving the presupposition strategy, if we adopt an interpretation that views emptinesss not in terms of interdependence but in terms of nonconceptuality. Let’s clarify more fully what’s involved in these two accounts of emptiness. For present purposes I will focus on the interdependence account of emptiness defended by Garfield, which takes many of its cues from the reading of Nāgārjuna associated especially with Tsong kha pa. Garfield takes the negative tetralemma of MMK 22:11 to express Nāgārjuna’s view of the functioning of language at the ultimate level, and thus holds that at this level nothing at all can be asserted about reality.21 At the conventional level, however, he holds that true assertions can be made. This is because, for him, assertion is the ascription of a property to an entity, and at the conventional level, entities and properties may be said to exist: “corresponding to these conventional assertions are real propositions that make them true or false – entities with or without the ascribed properties” (Garfield 1995, 213). Since the constituents of the thoughts expressed by assertions are concepts, it seems clear that underlying this notion that assertion is possible at the conventional but not at the ultimate level is a parallel distinction between the concept of existence employed at the ultimate level and the concept of existence used at the conventional level. The concept of existence involved at the ultimate level is a concept of things having svabhāva, independent existence. To say that things exist conventionally, by contrast, is to say that they exist dependently on their parts, their causes, their materials, and on human conceptualization and language. The conventional concept of existence is thus a concept of interdependent existence. Since for Nāgārjuna the ultimate truth and the conventional truth are said to be the same (MMK 24:18), the ultimate truth that everything is empty of svabhāva becomes just another way of expressing the conventional truth that everything exists interdependently. Emptiness is interdependence. According to the interdependence view, to speak at the level of ultimate truth is to make a statement with a svabhāva-presupposition, a presupposition that is in fact false. Notice, though, that on this view, since conventional assertion is possible, it must be the case that conventional level statements do not have svabhāva-­ presuppositions. The nonconceptualist account takes a different view, which may at first seem counterintuitive. It maintains that statements at both the ultimate and the conventional levels have svabhāva-presuppositions, and for this reason, that strictly speaking presupposition-failure is catastrophic rather than non-catastrophic. Strictly speaking, assertion is possible neither at the ultimate nor at the conventional level.  Garfield (1995), 280–82, 330–31. We should note that more properly, given his adherence to a dialethist understanding of Nāgārjuna, which allows for the possibility of true contradictions, Garfield holds both that at the ultimate level no true assertions are possible, and that true assertions can be made. See Garfield and Priest (2003). 21

The Presupposition Strategy and the Two Truths

377

We can clarify more fully the commitments of the nonconceptualist position by considering some arguments that favor it over the interdependence view. The central motivations for the nonconceptualist account are twofold, first, textual evidence from the MMK, and second, the alleged incoherence of the notion of universal ontological interdependence adopted by the interdependence reading. I have developed both of these arguments more fully elsewhere (Spackman 2014), and in the present context, I will focus mainly on the first. I cannot here discuss the second in any detail, but let me note in passing its central idea. The notion of universal ontological interdependence has not infrequently been viewed as logically incoherent, especially by advocates of the nihilist view of Nāgārjuna, both ancient and modern.22 The main objection is, in short, that it makes no sense for all existent things to be dependent on other existents, because if that were the case there would be no independently existent things for the dependent things to depend on. Paul Williams puts it succinctly when he maintains that in ancient India, Nāgārjuna’s view was often regarded as nihilistic because if everything was interdependent, then “all things would be constructs with nothing for them to be constructed out of” (Williams 2000, 150).23 As to the textual evidence in support of the nonconceptualist account, there are a number of passages from the MMK that appear to claim that whatever exists must be either svabhāva or parabhāva – that is, as I understand it, that whatever exists must either exist independently, or depend for its existence on something else that exists independently. Let’s call such statements ESP statements, for “existents are svabhāva or parabhāva.” One clear example appears at MMK 15.4: Further, without intrinsic nature (svabhāva) and extrinsic nature (parabhāva) how can there be an existent (bhāva)? For an existent is established given the existence of either intrinsic nature or extrinsic nature.24 On the face of it, these verses appear to claim that anything that exists must be either svabhāva or parabhāva. But there are different ways of interpreting these lines. The nihilist account of emptiness holds that for Nāgārjuna nothing exists at all, neither at the ultimate nor at the conventional level, and reads these lines as supporting that view. For if whatever exists is either svabhāva or parabhāva, and nothing is svabhāva, then it seems to follow that nothing exists at all (Wood 1994). Garfield, on the other hand, glosses these lines in a way consistent with the interdependence interpretation. He reads the term “existent” (bhāva) in each line as referring not simply to existent things, but to “inherently existing entities,” that is, existents viewed as svabhāva (Garfield 1995, 221). On this account, what the verses reject is not the existence of things per se, as the nihilist would have it, but simply the  See for instance Wood (1994), Burton (1999).  For contemporary philosophers who have raised similar doubts about the coherence of the idea of universal ontological interdependence, see for instance Blackburn (1990), Heil (2003). 24  svabhāvaparabhāvābhyām ṛte bhāvaḥ kutaḥ punaḥ / svabhāve parabhāve ca sati bhāvo hi sidhyati. I have added the first two of the italicized terms in the first verse for clarity. 22 23

378

J. Spackman

existence of things as svabhāva. Conventionally, things may still be said to exist, as interdependent. The nonconceptualist interprets these lines in yet a different way, a way that can be seen as falling in-between the nihilist and interdependence accounts. The nonconceptualist agrees with the interdependence theorist, contra the nihilist, that things may be said to exist at the conventional level.25 But she views the interdependence account as failing to do justice to what we actually find in these lines of the Mūlamadhyamakakārikā. There seems to be little textual basis for interpreting the occurrences of “bhāva” in the verses above as meaning “inherently existing entities” rather than simply “existents.” If that is what Nāgārjuna meant, he could after all have used the term “svabhāva” instead. Garfield’s reading would also make each of these two verses into elaborate tautologies. On this account, what the first line says is “Further, without svabhāva or parabhāva how can there be an existent with svabhāva?” If the sense of “bhāva” in these lines is not restricted to “existents with svabhāva,” the nonconceptualist proposal is that we view it as referring to the concept of existence in general, so that the gist of the verses is that we cannot make sense of the concept of existence except in terms of what is svabhāva or parabhāva. That is, we should view ESP statements such as this one as expressing a claim about the concept of existence itself – namely, what we might call the conceptual claim: The conceptual claim  – The presupposition that whatever exists is either svabhāva or parabhāva is a core part of the meaning of the concept of existence itself.26

If the ESP statements are understood in this way, the implication is that any use of the concept of existence – whether at the ultimate or at the conventional level – is governed by what we might call the svabhāva-parabhāva (sp-) presupposition. In contrast with the interdependence theorist’s view that there are two distinct concepts of existence, on this view we have only one concept of existence, the concept described by the conceptual claim. Given that for Nāgārjuna nothing is svabhāva, strictly speaking this concept cannot be truly applied at either the ultimate or the conventional level. This is why this reading is called nonconceptualist. Strictly speaking, assertion is possible neither at the ultimate nor at the conventional level; presupposition failure occurs at both levels, and so is catastrophic. The obvious question raised by this view is how the nonconceptualist can hold both, contra the nihilist, that for Nāgārjuna things may correctly be said to exist at the conventional level, and also that strictly speaking assertion is not possible at this level. What allows the nonconceptualist to consistently adopt these positions is her

 For a fuller argument to this effect, see Spackman (2014), 155–57.  Saying that this presupposition is a core part of the meaning of the concept of existence might invite the challenge that this view is committed to an essentialist view of concepts that Nāgārjuna would find problematic. But in the sense in which I intend it, to speak of a belief as being a core part of the meaning of a concept is not to imply that it is an essential or necessary part of it in a way that would raise Quinean scruples. It is only to say that if the specified belief is false, we begin to lose a grip on what an application of the concept means in those circumstances. 25 26

The Presupposition Strategy and the Two Truths

379

account of the conventional truth, which differs significantly from that offered by the interdependence theorist. For the nonconceptualist, to speak at the conventional level is not to employ a distinct concept of interdependent existence, but rather to employ the sole concept of existence we have, with its sp-presupposition, in a less than fully strict way, a way that abstracts from the sp-presupposition that forms part of its content.27 What distinguishes discourse at the level of ultimate truth from discourse at the level of conventional truth, on this view, is the attitude of speakers at the two levels towards the sp-presupposition. Speakers at the ultimate level treat this presupposition as if it were, as it in fact is, part of the meaning or truth-conditions of existence claims; that is, they accept any existence claim as true only if they believe the presupposition that the existing things are either svabhāva or parabhāva is satisfied. By contrast, speakers at the conventional level  – or at any rate, liberated speakers  – adopt the contrary attitude: they do not treat the sp-presupposition as part of the meaning or truth-conditions of claims at this level. Recall that on this account, sp-­ presuppositions are part of the meaning of conventional existence claims; liberated speakers simply treat them as if they were not. Given the fundamentally pragmatic notion of presupposition developed earlier, this talk of sp-presuppositions being part of the meaning of existence claims will obviously be shorthand. On this view, it is part of the common ground, part of the shared system of beliefs, for existence claims, that all speakers have a deep-seated belief that whatever exists is either svabhāva or parabhāva. Indeed, given the nonconceptualist claim that universal ontological interdependence is incoherent, it is part of this common ground that it is not possible for speakers to believe that something could exist which is neither svabhāva or parabhāva. To say that the sp-presuppositions are part of the meaning of existence claims, then, is a shorthand way of saying that all speakers take this deep seated belief to be part of the common ground for existence claims. However, on this account, what distinguishes the attitude of liberated speakers is that they are willing to accept these conventional claims as true, despite these deep-seated beliefs.28 The attitude of liberated conventional speakers toward sp-presuppositions, on this account, may be helpfully compared to the attitude of acceptance described by Stalnaker. Recall that for Stalnaker, to accept a proposition is not to believe it is true, but “to treat it as true for some reason” (Stalnaker 2002, 716). To speak at the conventional level, on this view, is to accept claims as true for practical purposes, even though, because the sp-presupposition is not satisfied for them, they are not ultimately true. Whereas for the interdependence theorist, conventional truth consists in correspondence to conventional truth-conditions, for the nonconceptualist, it consists in acceptance for practical purposes of claims that are not strictly speaking true.  I borrow the notion of using concepts in more and less strict ways, ways that abstract more and less from their content, from Mark Johnston. See e.g. Johnston (1992). 28  Deluded speakers, at least of what might be called a common sense realist type, will of course not adopt this attitude, since for them in order for a statement to be conventionally true its sp-­ presupposition must be satisfied. 27

380

J. Spackman

Though I have dwelled on one appearance of an ESP statement in the Mūlamadhyamakakārikā, it should be emphasized that similar statements occur in multiple places in the work. Consider for instance MMK 1:10: “Since things devoid of intrinsic nature are not existent/‘This existing, that comes to be’ can never hold.”29 The first half of this verse appears to say that there are no beings or existents (bhāva) that are not svabhāva. Similarly, the first half of MMK 5:2 reads “Nowhere does there exist any such thing as an existent without defining characteristic.”30 If the notion of a defining characteristic (lakṣaṇa) is interpreted, in keeping with the account of the apparently Abhidharmika opponent against which Chapter 5 is directed, as meaning an essential property, the property that defines the svabhāva of a thing, then this verse would appear to imply that nothing exists without having svabhāva. In each of these cases, as in our first case, the interdependence theorist must resort to ad hoc interpretive hypotheses unmotivated by Nāgārjuna’s text in order to preserve a reading in keeping with the interdependence account of emptiness. And in each of these cases, the nonconceptualist urges that we can develop a more consistent and less ad hoc interpretation if we view these verses as expressing the conceptual thesis.31 I take the evidence we have considered  – both the incoherence of the idea of universal interdependence and the textual examples – to provide strong support for a nonconceptualist account of emptiness. Why should we think that the nonconceptualist view offers a better foundation than the interdependence account for preserving the viability of the presupposition strategy? As we have seen, the nonconceptualist views the presuppositions relevant to the catuṣkoṭi as semantic, in the derivative sense, in that the sp-presupposition is part of the very meaning of existence claims. It is thus in a position to regard the failure of these presuppositions as affecting the truth-values of the statements of the catuṣkoṭi, and resolving their apparent contradictions. Now it is true that there is also a sense in which the interdependence theorist might also view the svabhāva-presupposition as semantic. After all, such a theorist holds that svabhāva-presuppositions are part of the meaning of the concept of existence employed at the ultimate level, the concept of independent existence. And even though they are not, on this view, part of the meaning of the conventional concept of existence, it is only at the level of ultimate truth that statements are subject to semantic deficiency, so the independence theorist could still potentially explain the semantic anomalies of the catuṣkoṭi by reference to presupposition failure. While such an account in many ways offers an attractive explanation of how svabhāva-presuppositions might be viewed as semantic, the point of my earlier arguments for the nonconceptualist account was precisely to undermine such a picture. This picture depends for its cogency on a view of the operation of language at  bhāvānāṃ niḥsvabhāvānāṃ na sattā vidyate yatah/satīdam asmin bhavatīty etan naivopapadyate 30  alakṣaṇo na kaścic ca bhāvaḥ saṃvidyate kva cit 31  See Spackman (2014) for a fuller development of the argument in relation to these additional examples of the ESP statements. 29

The Presupposition Strategy and the Two Truths

381

the ultimate and conventional levels of discourse as in fundamental respects semantically independent. Statements at the ultimate and conventional levels are treated as semantically independent in the sense that they are underwritten by two semantically independent concepts of existence – independent existence and interdependent existence – and thus they have truth-conditions that are independently satisfiable. It is this view of the language at the two levels of truth as functioning independently that I have suggested misconstrues the relationship between the two truths, and the notion of emptiness itself. Such a view depends on the claim that there is a coherent concept of universal ontological interdependence that operates independently of the concept of independent existence. We have seen both that there is reason to doubt the coherence of this concept, and that there is evidence, in the ESP statements, that Nāgārjuna would not countenance the notion of two distinct concepts of existence. It is for these reasons that the nonconceptualist account offers the most plausible model for accounting for the semantic features of svabhāva-­presuppositions and preserving the presupposition strategy.

6 Conclusion: Nonconceptualist and Anti-Realist Accounts of Emptiness The recent movement in philosophy and linguistics toward a reconceptualization of presupposition as pragmatic rather than semantic is, I have suggested, one that should be viewed congenially in the Buddhist context. But I have argued that this reconceptualization need not undermine the viability of the presupposition strategy. I have argued as well that the best model for preserving the cogency of the strategy is one that adopts a nonconceptualist rather than an interdependence account of emptiness, as the latter comes at the cost of an implausible account of the relationship between the two truths. But the view of the two truths offered by the nonconceptualist also differs from that presented by anti-realists such as Siderits, and I want finally to consider these differences. The nonconceptualist account might be thought to fall prey to the difficulties confronting a view of emptiness that Siderits has often warned against, the view that ultimate reality is ineffable, wholly beyond the reach of language and rationality (Siderits 2016b, d, 2019). After all, on this account, to say that things are empty is to say that they cannot be conceptualized, either at the level of ultimate truth, or, strictly speaking, even at the level of conventional truth. But the nonconceptualist is in no way committed to saying that there is an ultimate reality that is ineffable. The key move in Siderits’ defense of the anti-realist view that there is no such thing as ultimate reality is the claim that Nāgārjuna’s critique applies even to his own account of ultimate truth, the doctrine of emptiness. This is what he takes to be the purport of MMK 22:11, which applies the logic of the catuṣkoṭi to emptiness itself, yielding the result that emptiness is itself empty. But this move is available as much to the nonconceptualist as it is to the anti-realist. The nonconceptualist holds not that there

382

J. Spackman

is an ultimate reality that is not conceptualizable, but simply that no concepts can be employed in true assertions, neither ultimately nor, strictly, conventionally. The claim that there is an ultimate reality is itself subject to the logic of the negative catuṣkoṭi, and so cannot simply be asserted to be true. Nonetheless, the nonconceptualist account is importantly different from Siderits’ anti-realist approach in its account of the relation between the two truths. As Siderits presents it, the anti-realist interpretation holds simply that “the ultimate truth is that there is no ultimate truth – there is only conventional truth” (2016, 41). Such a view follows naturally from Siderits’ preferred approach to the presupposition strategy, which regards the four alternatives of the negative tetralemma as meaningless or without truth-value, due to the failure of their common presupposition. If there were an ultimate truth, the negative tetralemma would express it, but all of its statements are meaningless. What the anti-realist account does not recognize, I would suggest, is the way in which claims at the level of ultimate truth, the claims of the negative tetralemma, may have profound effects on the character of the conventional truth. I have emphasized that on the nonconceptualist view, even at the conventional level, strictly speaking the concept of existence does not apply to things, though of course existence claims may still be considered true in the sense that they are accepted for practical purposes. What this means is that everyday things like rocks and pear trees and persons do not, in the full sense of the concept, exist; whatever these things are doing, it is not existing, in the full sense of the term. On the nonconceptualist account, we have only one concept of existence, employed at both the ultimate and conventional levels, and if this is so, then the fact that the ultimate truth is emptiness will have consequences for how we think of everyday reality. If, instead of viewing the four negations of the tetralemma as meaningless, we regard them simultaneously true, as I have proposed, a different account of the relation between the two truths is suggested. This account is the one that is proclaimed prominently in MMK 24:18, according to which the ultimate truth is the conventional truth: “Dependent origination we declare to be emptiness.” It is worth noting that Siderits shies away from accepting the identity of the two truths, suggesting that it might be a mistake to take MMK 24:18 at face value (Siderits 2016b, 40n). But if we view the negations of the tetralemma as true rather than meaningless, there is no difficulty in viewing the two truths as identical. We can maintain, in keeping with Leibniz’ law, that to say that the two truths are identical is to say that whatever is true at the ultimate level holds of conventional reality, and whatever is true conventionally holds of ultimate reality. Thus, while rocks and pear trees and persons may be truly said to exist conventionally, in the sense that we can accept claims about them for practical purposes, the negative catuṣkoṭi applies to them as well: strictly speaking (in the full sense of the concept) they neither exist, nor do not exist, nor both, nor neither. In this sense, on the nonconceptualist view, the ultimate truth is the conventional truth.32  The nonconceptualist account has affinities with the perspective offered by Garfield and Priest (this volume), since both affirm that there is an ultimate truth. The two views differ, however, in that Garfield and Priest view ultimate reality as ineffable, while the nonconceptualist rejects such a notion. For the nonconceptualist, ultimate reality is eminently effable, for ultimate reality is identical to conventional reality, and we utter positive conventional truths all the time. 32

The Presupposition Strategy and the Two Truths

383

References Beaver, D.I., and B. Geurts. 2011. Presupposition. In Stanford Encyclopedia of Philosophy. https:// plato.stanford.edu/entries/presupposition/. Accessed 19 Jan 2019. Bezuidenhout, A. 2016. Presupposition Failure and the Assertive Enterprise. Topoi 35: 23–35. Blackburn, S. 1990. Filling in Space. Analysis 50 (2): 62–65. Böer, S.E., and W.G. Lycan. 1976. “The Myth of Semantic Presupposition.” Ohio State Working Papers in Linguistics, No. 21. Department of Linguistics, Ohio State University, Columbus. Bodhi, B., and B. Ñānamoli, trans. 1995. The Middle Length Discourses of the Buddha: A Translation of the Majjhima Nikāya. Boston: Wisdom Publications. Burton, D. 1999. Emptiness Appraised: A Critical Study of Nāgārjuna’s Philosophy. London/New York: Routledge. Candrakirti. 1913. Prasannapadā Mūlamadhyamakavṛttiḥ. ed. Louis de la Vallée Poussin. St Petersburg: Bibliotheca Buddhica IV. Donnellan, K. 1966. Reference and Definite Descriptions. Philosophical Review 75: 281–304. von Fintel, K. 2001. Would You Believe It? The King of France is Back! (Presuppositions and Truth-Value Intuitions). In Descriptions and Beyond, ed. A.  Bezuidenhout and M.  Reimer. Oxford: Oxford University Press. Garfield, J. 1995. The Fundamental Wisdom of the Middle Way: Nāgārjuna’s Mūlamadhyamakakārikā. Oxford: Oxford University Press. Garfield, J., and G. Priest. 2003. Nāgārjuna and the Limits of Thought. Philosophy East and West 53: 1–21. ———. 2009. Mountains are Just Mountains. In Pointing at the Moon: Buddhism, Logic, Analytic Philosophy, ed. J.L.  Garfield, T.J.F.  Tillemans, and M.  D’Amato, 71–82. Oxford: Oxford University Press. Garfield, J., and N. Samten, trans. 2006. Ocean of Reasoning: A Great Commentary on Nāgārjuna’s Mūlamadhyamakakārikā. Oxford: Oxford University Press. Heil, J. 2003. From an Ontological Point of View. Oxford: Clarendon Press. Johnston, M. 1992. How to Speak of the Colors. Philosophical Studies 68 (3): 221–263. Kempson, R. 1975. Presupposition and the Delimitation of Semantics. Cambridge: Cambridge University Press. Lycan, W.G. 2000. Philosophy of Language: A Contemporary Introduction. New York/London: Routledge. Ruegg, D.S. 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: 1–71. Russell, B. 1905. On Denoting. Mind 14: 479–493. ———. 1957. Mr Strawson on Referring. Mind 66: 385–389. Siderits, M. 2007. Buddhism as Philosophy. Indianapolis/Cambridge: Hackett Publishing. Searle, J. 1969. Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press. ———. 2016a. Nāgārjuna as Anti-Realist. In Studies in Buddhist Philosophy, ed. M.  Siderits. Oxford: Oxford University Press. ———. 2016b. On the Soteriological Significance of Emptiness. In Studies in Buddhist Philosophy, ed. M. Siderits. Oxford: Oxford University Press. ———. 2016c. Contradiction in Buddhist Argumentation. In Studies in Buddhist Philosophy, ed. M. Siderits. Oxford: Oxford University Press. ———. 2016d. Distinguishing the Mādhyamika from the Advaitin: A Field Guide. In Studies in Buddhist Philosophy, ed. M. Siderits. Oxford: Oxford University Press. ———. 2019. The Prapañca Paradox. Journal of Indian Philosophy 47: 645–659. Siderits, M., and S.  Katsura, trans. 2013. Nāgārjuna’s Middle Way: Mūlamadhyamakakārikā. Boston: Wisdom Publications. Simons, M. 2006. Foundational Issues in Presupposition. Philosophy Compass 1 (4): 357–372.

384

J. Spackman

Spackman, J. 2014. Between Nihilism and Anti-Essentialism: A Conceptualist Interpretation of Nāgārjuna. Philosophy East and West 64: 151–173. Stalnaker, R. 1974. Pragmatic Presuppositions. In Context and Content, ed. R. Stalnaker. Oxford: Oxford University Press. ———. 2002. Common Ground. Linguistics and Philosophy 25 (5–6): 701–721. Strawson, P.F. 1950. On Referring. Mind 59: 320–344. ———. 1964. Identifying Reference and Truth-Values. Theoria 30: 96–118. Tillemans, T. 2009. How Do Mādhyamikas Think? Notes on Jay Garfield, Graham Priest, and Paraconsistency. In Pointing at the Moon: Buddhism, Logic, Analytic Philosophy, ed. Mario D’Amato, Jay L. Garfield, and Tom J.F. Tillemans, 83–100. Oxford: Oxford University Press. Westerhoff, J. 2009. Nāgārjuna’s Madhyamaka: A Philosophical Introduction. Oxford: Oxford University Press. Williams, P., and A.  Tribe. 2000. Buddhist Thought: A Complete Introduction to the Indian Tradition. London/New York: Routledge. Wilson, D., and D.  Sperber. 1979. Ordered Entailments: An Alternative to Presuppositional Theories. In Syntax and Semantics Vol 11: Presupposition, ed. C.-K. Oh and D.A. Dinneen. New York: Academic. Wood, T.E. 1994. Nāgārjunian Disputations: A Philosophical Journey Through an Indian Looking Glass. Honolulu: University of Hawai’i Press. Yablo, S. 2006. Non-catastrophic Presupposition Failure. In Content and Modality: Themes from the Philosophy of Robert Stalnaker, ed. J.J.  Thompson and A.  Byrne. Oxford: Oxford University Press.

Part IV

Morality

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist Amber Carpenter

I confess: I am a Buddhist-Platonist. There is, I think, an important and meaningful affinity between Buddhist and Platonic ethics – one that would enrich and enlarge our understanding of possible spaces within the moral terrain, and enable us in particular to appreciate a fuller range of manifestations of Buddhist ethical thought. I will call this area of salient overlap in moral outlook ‘impersonal idealism’. It is not, of course, a view that any Buddhist thinker will have set out in theoretical terms, and even regarding Plato one must analyse together a whole set of diverse claims across several texts in order for the structure I wish to articulate to come into view. In this sense, then, one might consider this project to be an exercise in rational reconstruction, in the spirit of Mark Siderits. I will, however, take it that it is an advantage of this particular reconstruction that it leads with the priorities and questions that we find in the ancient thinkers themselves.

We all owe a tremendous debt of gratitude to Mark Siderits for his career of trail-blazing work, bringing Buddhist philosophy into the field of critical Anglophone philosophical inquiry. I remember in particular a lively exchange with Mark in Oxford, when I was only at the very beginning of my own work in this area, about whether Buddhist ethics is consequentialist. This paper is a continuation of that provocative conversation, offered with heartfelt appreciation for the lively philosophical friendship Mark showed in his critical engagement then, and always. This piece also owes a great deal to audiences at Melbourne and Oxford, whose engagement with earlier drafts substantially contributed to the improvement of this final version. And it has its proximate roots in work with the Moral Beacon Project (Templeton Religious Trust), which supported my first inquiries into Buddhist-Platonism. A. Carpenter (*) Yale NUS, Department of Philosophy, Singapore, Singapore e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_20

387

388

A. Carpenter

Buddhist-Platonism may sound absurd on the surface of it – and perhaps not only on the surface.1 For ethics is closely bound up with metaphysics, for both Platonist and Buddhist2; yet their respective metaphysical views could hardly be more opposed. Plato thought there were unchanging realities, intelligible, well-ordered and good, and responsible for the good, intelligible order of the sensible cosmos. Buddhists reject all that, from the unchangingness, to the intelligibility and good order. If there is anything on which the classical Indian Buddhists may be taken to speak with one voice it is on impermanence, no-self and suffering as the marks of mutually dependently arising existence. In spite of this, I argue, the impersonal nature of reality for both parties, as well as the centrality assigned to knowing this impersonal reality, constitute a significant and distinctive approach to ethics.3 This might be spelled out in terms of a phenomenon and an ethical structure. The phenomenon is the formation and re-formation of character by the ethical ideals one holds, by the depictions of these ideals, and by the practices around these; Buddhist and Platonic moral thinking are similar, I suggest, in foregrounding the formation and re-formation of character around ambitious ideals. The Buddhist aims to become an Arhat or a Bodhisattva; the Platonist aims to become a Philosopher or ‘properly dialectical’, as that is described in the Republic, for instance, or the Philebus. This ideal structures accounts of moral development, its mechanisms and its character, and subordinates rules of action and moments of choice as intelligible and relevant only with reference to these. Beyond this basic idealist-­phenomenological structure, the structuring ideals in Buddhist and Platonic ethics share key features: Both put an overriding priority on knowing an impersonal reality, as the central and indispensable ideal which will transform one’s character in the necessary way.4 Both recommend as an ideal state a perspective which appreciates the fundamentally impersonal character of reality as a whole, and a consequent divestment of personal categories. And for both, this freedom from personality and from person-­ oriented concerns grounds, in different ways, an ability, moment-to-moment, to recognise responsively each particular situation in all of its complexity and singularity. I will approach the discussion by way of an examination of virtue ethics. This is not because I hold out great hopes for a finally illuminating classification of that  I am, however, gratified to be in the company of Iris Murdoch (especially, but not only, Murdoch (1956, 1970), who picks up on the same aspects of Plato’s moral thinking that I will bring out here, as well as their affinity with Buddhism (especially prominent in her novels, including for instance The Sea, The Sea, The Nice and the Good, The Green Man) – in particular, the dedicated attention to an impersonal reality, and the moral effect of drawing us out of ourselves and transforming what she calls ‘vision’, which is prior to all action and choice, and indeed conditions what options reality appears to give us. 2  As I argue in Carpenter (2014), 4 and passim. 3  Throughout, ethics must be construed broadly, as part of what is under discussion is precisely how this terrain should be demarcated; a feature that Buddhist and Platonic thought in this area share is that the defining concerns are not always, and never entirely, those of modern Euro-Anglophone moral philosophy. 4  Carpenter (2018) argues for this reading Buddhaghosa’s ‘Samādhi’ chapter of the Visuddhimagga. 1

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

389

mongrel non-entity ‘Buddhist ethics’ into some one or another of the pre-given options of Anglophone analytic moral philosophy of the late 20th and early 21st Century. In fact, I argue that Plato is ill-served by pressing him into the virtue ethics mould5; and that classical Indian Buddhist ethical thought is likewise ill-served for the same reason.6 This should identify a convergence in the two views that goes to the heart of how we conceive of our moral lives and our ethical project. But this convergence gets obscured when we, on the Buddhist side, dismiss out of hand a range of texts and practices as irrelevant to philosophy and moral theory; and when we, on the Platonist side, assimilate all the ancient Greeks and Romans to the Aristotelian virtue ethics mould. On the Buddhist side, short narratives abound in the tradition; they travel widely throughout the Buddhist world, and persist over time. But by and large Anglophone philosophers, at least, have been strangely incurious about this phenomenon – setting them aside as morality tales central to Buddhist culture, perhaps, but not to moral theory.7 This has largely persisted in spite of Charles Hallisey and Anne Hansen’s excellent ‘Narrative, Sub-Ethics, and the Moral Life’ (1996), which calls on scholars and philosophers to do otherwise.8 To understand Buddhist ethics properly, we must take on board the fact that the distinctively Buddhist ethical outlook is formed through the telling and retelling of Buddhist tales.9 Consider for comparison how ethically vital Plato took myth or story-telling to be: those interested in his thought know that understanding that he thought it so, and why, and what sort of tales he thought ought to be told, is essential to understanding his ethical view.10 What is obvious in the case of Plato should be no less obvious in the case of Buddhist moral thought.11 What sort of ethos is it that is necessarily  – not just  Notice how three concepts central to the virtue ethics tradition (arete, eudaimonia, phronesis) are decidedly off-centre in Plato, who for instance converts all talk of virtue into talk of knowledge. Carpenter (2017c) argues the case using the debate with Polus in the Gorgias to focus the discussion. See also White (2015) for debate over whether Plato should be read as a virtue ethicist. 6  So although I agree with Harvey (2000) that, taken extensively, “the rich field of Buddhist ethics would be narrowed by wholly collapsing it into any single one of Kantian, Aristotelian, or Utilitarian models” (50), I do hope to bring out something salient that runs through much of Indian Buddhist moral thinking which each of those categories would obscure. 7  Siderits (2015), 121, for instance, distinguishes moral theory from moral life in a way that would exclude significant philosophical consideration of narrative tales as relevant to moral theory. 8  Adam (2018) is a move towards correcting this; and Heim (2014), Ch. 4, is excellent on this. Appleton (2016) and Ohnuma (2017) engage seriously with the story literature of Buddhism; and Stepien, ed. (2019) tackles various aspects of Buddhist literature and philosophy. Obeyesekere and Obeyesekeye (1990) offers early and rich comment on the ethos shaped by Budddhist tales, ‘The Demoness Kali’ story in particular. 9  “Looking back at my childhood,” Ranjani Obeyesekere writes (Obeyesekere (1991), x), “I realize we were never given religious instruction as such... We participated in Buddhist rituals and ceremonies...and listened to many, many Buddhist stories. That is how we learned to be Buddhists.” 10  For instance Gill (1985), Smith (1999), Janaway (1995), Kamtekar (1998),  Scott (1999), Jenkins (2015). 11  Buddhist jātaka tales and avadānas are as restrictive in their content and in their style in practice as Plato recommends for ideal ethical formation in Republic II-III (though not perhaps as restrictive as the recommendations in Republic X). I discuss details of this in Carpenter (2019, 2020). 5

390

A. Carpenter

accidentally – formed and reformed through stories? What is morally salient and essential when it matters which stories are told? And what does it mean that it is just these narratives, tales of this very sort  – in the Buddhist case, the jātakas, the avadānas, the stories collected in the Dhammapada Commentary and in the vinaya? On the Greek virtue ethics side, the long shadow of Aristotle occludes our ability to appreciate Plato’s distinctive ethical approach and outlook. The Anglophone Aristotelian Revival partially constitutes and thoroughly influences what is today called ‘virtue ethics’ – a moral theory or a way of doing ethics that positions itself in distinction to (on the one hand) Kantian theories that ground moral goodness in the rational will and (on the other) to consequentialist theories that vest all goodness in results. While Kant taught us to lump all ‘the Ancients’ together as eudaimonists12 – theorists of happiness and not of morality – it is to Aristotle in particular that Anscombe (1958), Foot (1958, 1958-59, 1969) and later McDowell (1979),13 MacIntyre (1981), and Hursthouse (1999) looked in their efforts to recover a different way of approaching moral philosophy.14 And so Aristotle has set the template for what we understand today as virtue ethics – set the questions, topics and categories constitutive of virtue ethics  – so that it becomes nearly synonymous with neo-­ Aristotelian naturalism, even when it departs widely from the particular virtues Aristotle himself recognised – as it does, for instance, in discourse about ‘Confucian virtue ethics’.15 The resurgence of interest in Aristotelian-style ethics has enriched Anglophone moral philosophy, broadening the range of questions to be asked, and reasserting the philosophical value of a realistic moral psychology. As ‘neo-Aristotelianism’,16 it is a rich field for philosophical inquiry. Under the name of ‘virtue ethics’, however, and its relative ‘eudaimonism’, all ancient Greek ethical thought is swept into the

 KpV 64; ‘One must regret that the acuteness of these men...was applied infelicitously in excogitating identity between extremely heterogeneous concepts’ (viz., happiness and virtue, KpV 111 in Pluhar, tr. (2002)). This lesson, that the gulf between the ancients and the moderns lies in whether they rejected or accepting the dualism of practical reasoning, was taken up by Sidgwick, who calls it ‘the most fundamental difference’, and is discussed by Frankena (1992). 13  Published the year after Foot’s essays collected in Virtues and Vices. 14  See also Sherman (1991), Lovibond (2002), Kraut (2007), Thompson (2008). The dissident revivalist is Iris Murdoch, who was much more interested in recuperating Plato in her moral philosophy (witness the title of her lectures, Sovereignty of Good)  – and also, not accidentally, a Buddhist outlook in her novels; also Simone Weil by whom Murdoch was influenced. This dissident strain in the Anglophone tradition is picked up by Holland (1980), Gaita (1991) (second edition (2004)), Blum (1994), and more recently Chappell (2014), and to some extent by Brewer (2009) and Vogt (2017); and, with a Christian twist Adams (1999). 15  An indicative collection of the extensive work in this area is Angle and Slote (2013). Of course one need not go so far afield as China: MacIntyre (1981) advocates the recovery of a different set of virtues from those Aristotle identified, drawing on Jane Austen’s understanding of human flourishing and good, and the Christian tradition in which it is embedded. 16  or even neo-Aristotelian naturalism, for not all interpreters of Aristotle agree with the neo-Aristotelians cited above that Aristotle himself seeks to ground ethics in human nature (see, for instance, Berryman (2019)). 12

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

391

same category,17 and what is distinctive, insightful and even challenging in Plato’s ethics, at least, vanishes from sight. What vanishes, astonishingly enough, is the Form of the Good – its role in Plato’s moral thought, and its implications for the rest of life and thought.

1 Buddhist Ethics as Virtue Ethics? It is a commonplace that classical Indian philosophy has no moral theory.18 This verdict, of course, rests rather heavily on what one counts as moral theory, which texts one allows might be doing such a thing, and how one reads them.19 But as Siderits observes (2015, 121–24), the fact that classical Indian thinkers neither go in for debate over the real sources of normativity, nor battle over why we ought to be moral, suggests at least that they typically have rather more affinity with classical Greek ethical styles than with modern European moral theory and its contemporary Anglophone heirs; and this in turn suggests taking classical Indian moral thinking as a type of virtue ethics.20 However, Indian Buddhist moralists cannot be eudaimonists, Siderits claims, and must rather implicitly hold a consequentialist moral theory.’21 Indeed it may seem obvious that any view on which the ultimate aim is to  According to Julia Annas, all ‘ancient writers belong to the same ethical tradition as Plato. They are eudaimonists’ (Annas 1999: 2); and, starting with Plato’s Socrates, ‘Ancient theories unanimously locate virtue within our overall aim of happiness’ (Annas 2003: 3–4). Striker (1996), Vlastos (1999) (105–36, 108) and Irwin (1995) (53) concur, regarding Plato. Not all contemporary virtue ethics positions are eudaimonist but most are, for typically a virtue is that which either reliably delivers eudaimonia, or is a proper part of what eudaimonia consists in. (Virtue ethics without eudaimonism, such as Slote (1997), are still typically Aristotelian in their focus on ‘agents’ and thereby actions; Slote 2001’s Humean virtue ethics is an outlier.) Siderits (2015) focuses his discussion on eudaimonism. 18  Matilal (2002): 19; see also Tillemans (2010-11) and Barnhart (2012). 19  “There is a tendency today to think that the job of ethical theories is to give an answer to the question of moral motivation (Why should I be moral?),” writes Siderits (2015, 120), “and to provide decision criteria that help us solve moral dilemmas. This makes of ethics something whose chief focus is moral reasons—reasons that concern other-affecting actions—and not prudential reasons”. Winch (1972), 171–73 offers sharp critique of this ‘problem solving’ and action-guiding conception of ethics or moral philosophy, which prevails nonetheless. It is true, then, that if this is what moral theory is, one looks in vain for moral theory in ancient India and, as Siderits goes on to argue, this is one reason to find ‘virtue ethics’ rubric apt. 20  Regarding Buddhist ethics, Damien Keown (1992) was agenda-setting in conceiving it as virtue ethics, and he pursues the line in great specificity, including tying virtue ethics specifically to Aristotelianism. 21  He is by no means the only one to suggest this; Goodman (2009) argues with great subtlety for a distinctively Buddhist form of character consequentialism. Clayton’s precise and lucid study of Śāntideva’s Śiṣkāsamuccaya patiently considers the virtue ethical and consequentialist aspects of the text, before concluding that it offers a provisional, agent-based ‘supererogatory character ethic’ for those on the path, but ultimately a consequentialist ethic for those of fixed good motivation (Clayton 2006, 100, 114–15). 17

392

A. Carpenter

eliminate suffering can only be consequentialist – Bodhisattvas’ license to suspend ordinary constraints on conduct, when their perfected insight enables them to discern that doing so would reduce suffering, only makes the point more sharply. If the Buddhist texts, especially the Mahāyāna texts, also talk about paramitās – perfections of character, or virtues, including generosity and patience  – this makes no odds, for after all even Kant and Mill had their theories of virtue. Granted: virtue ethics is not just virtue theory, plus an obstinate refusal to answer the normative question, ‘What makes an action right?’ It is the positive assertion of an alternative understanding of ethical inquiry, for which Aristotle offers the template. Still, so understood, I think we capture more of the richness and subtlety of classical Indian Buddhist moral thought, and appreciate better its own priorities, if we understand it as virtue ethics.22 If there is something that nevertheless fits ill in such a characterisation, this stems from taking Aristotelianism in particular as our template for what virtue ethics is. For Buddhist ethics is much closer to Platonist ethics in respects that both of these labels obscure – namely, in (i) the central importance of knowing reality; (ii) the impersonality of the reality to be known; (iii) the resultant disinterest in the human good, and (iv) the concomitant transparency of moral exemplars.23 First, however, consider the dimensions of useful good fit in regarding Buddhist ethics as a form of virtue ethics. Virtue ethics prioritises questions about character and its excellences (virtues), and how it comes to have or lack these; what a life with or without these might look like; what the human good is and how and why the human excellences realise that good. Correctness of decisions, principles of choice and particular actions are set into the larger social and psychological context from which they gain their meaning, and subserve primary concerns.24 Virtue ethics has no ambition to offer us a decision procedure for awkward cases, and no c­ ommitment to denying the possibility of tragic conflict.25 Such an approach may imply theories of human nature, of virtue and the good,26 of the domain of the moral and its nature; but it resists assimilation into consequentialist or deontological paradigms by insisting on the logical priority of virtue and goodness, of character and lives, over  While it would be absurd to claim that no Indian Buddhist text engages in consequentialist reasoning, it is telling that when typical problems of consequentialism (about distribution, say, or about the connection between agents and acts) are brought to bear on Buddhist ethical discourse, they seem simply to miss the mark – and not because some Buddhist philosophers thought up an especially clever form of consequentialism that answers all such objections (but see Goodman (2009), who would no doubt disagree). 23  Appreciating these dimensions of affinity responds, I think, to Siderits’ worries about the relation between self and the final end on a virtue ethics view; and it avoids the objection (raised by both Goodman and Clayton) that (Mahāyāna) Buddhist ethics cannot be virtue ethics, because the former demands sacrificing one’s happiness (or being willing to) while the latter cannot do so, since happiness is after all that in terms of which virtue is defined. On the view to be sketched here, neither Platonic ethics nor Buddhist are eudaimonist in the first place. 24  MacIntyre (1981), Ch. 15, argues this point effectively. 25  Nussbaum (1986) is eloquent on this point of Aristotelianism. 26  Kraut (2007) and Hursthouse (1999) are significant instances. 22

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

393

decision points and principles of action. In this way, virtue ethics may be naturally associated with particularism of a certain sort,27 when it is not put forward as a distinct theory in its own right. So conceived, virtue ethics aptly brings out important features of Indian Buddhist ethical thought, in the following ways: 1. Buddhist moral thought is dominated by concern with character  – or, more strictly, with patterns in the arising of person-constituting dispositions, inclinations and reactions  – and with the excellences (Greek: aretai) or perfections (Skt: paramitā)28 of this character-bundle, and how it comes to have or lack these.29 It is preoccupied with what a life with or without these might look like; what real happiness is and how the virtues relate to it.30 A consequentialist needs virtue to be ‘whatever gets the job done’; but this is not how Buddhist texts describe the paramitās. Even if such a retrospective reconstruction may be possible, the texts consistently emphasise that generous, for instance, is a good way to be – it is part of what real happiness is, and it tends towards the reduction of suffering and the attainment of the final goal. Nāgārjuna’s Ratnāvalī is quite explicit about the excellent qualities of character being both constitutive of real happiness, and tending towards the final good – and his position in this respect is typical, rather than innovative. Paramitās are qualities of the practitioner the perfecting of which constitute the path towards the ultimate goal. Or, as Aristotle might say, they are the states and dispositions on account of which a human being counts as ‘good’ and attains the good. Not only do generosity and self-­ restraint lead to happiness,31 they constitute it. 2. Like most virtue ethics, Buddhist ethics has no ambition to offer a decision procedure for awkward cases, and no principled commitment to denying the possibility of tragic conflict. Buddhist ethics emphasises quality of mind and intention32  – this is part of what speaks strongly, though not insurmountably, against forcing Buddhist ethics into the consequentialist mould. And yet it does

 Not so much Dancy (1993) and (2004), perhaps, as the perception-based particularisms of McDowell (1981), Nussbaum (1986), McNaughton (1988), Nussbaum (1990) (Chs. 2 and 5), or Blum (1994). 28  The brahmavihāras of metta, karuṇā, muditā and uppekha might also be thought of as virtues in this sense. 29  Thus Garfield (2021) and (2015) characterise Buddhist ethics as ‘moral phenomenology’; Garfield (2010-11) offers detailed treatment with respect to Śāntideva in particular. 30  Consider for instance how Nāgārjuna, in the Ratnāvali, first corrects our notion of happiness, arguing that only a moral life, constrained by śīla, is actually a happy one at all. (This is before he shifts the goalposts to enlightenment as a goal  – the distinctively Platonist move, discussed in Carpenter (2015a)). 31  the good life, abhyudaya, identified with pleasure-happiness, sukha, RĀ I.4 32  Gombrich (1991) goes so far as to call Buddhist ethics an ‘ethic of intention’ (289). For as the Aṅguttra Nikaya famously declares, intention is karma (AN iii.415); and the Dhammapadā declares in its opening verse, mind is the forerunner of all things. ‘The moral value of an act’, Heim (2003) reminds us, ‘rests on the intention or volition underlying it’ (531). 27

394

A. Carpenter

not recommend assimilation to the deontological, because principles and imperatives of action – the ‘thou shalt not’ of deon – are not the measure of intention. In Buddhist moral psychology, an intention might be more readily characterised as a generous one, or a greedy one; and it is ways of perceiving reality – made habitual by experience, including imaginative and historical experience – that generate one quality or another of intention. In Buddhist ethical thought as in virtue ethics and ancient Greek moral psychology generally, questions of decision, principles of choice and particular actions are set into the larger social and psychological context from which they gain their meaning, and subserve primary concerns. Taking the whole social-psychological context as relevant, there is then no great gulf between ‘moral’ and ‘prudential’. Such an outlook resists assimilation into consequentialist or deontological paradigms by insisting on the logical priority of virtue and goodness, of character and lives, over decision points and principles. This resistance to imperatives and principles of action is one reason Buddhist ethics has also insightfully been called particularist.33 The vinaya’s preservation of the circumstances surrounding the first articulation of a precept, for instance, is not a freak interest in the historical record, but a mechanism for developing moral perception in each new generation, and thus an implicit assertion of the priority of perception and context over principle. 3. The consequentialist label, as the deontological, is unhelpful regarding the ‘path’ literature comprising the greater bulk of Buddhist moral thought. These texts are simply ‘untheoretical’, and therefore irrelevant to the ‘theory’ point of view. But one can and ought to ask why descriptions of the path from ordinary to extraordinary is a dominant mode of moral discourse. Categorising Buddhist ethics as virtue ethics can be helpful here, for virtue ethics is centrally concerned not only with qualities of character, but with the development of character. Eudaimonist virtue ethics, like Buddhist ethics, takes the whole of the person’s life and psychology as relevant to its domain, and takes a central area of concern to be the trajectory over time from less good to better. Our task is not to do right in each case, but to become perfect. Because of this virtue ethics, unlike consequetialism, takes story-telling as central to its concerns, not an adjunct. From within virtue ethics, it is obvious why there is such a prevalence of story literature in Buddhist ethical thought, and it is part of the domain of inquiry to consider why just these types of stories, what were their different uses and what the expected effects of so using them. A notion of ethical theory that strips away these virtue ethics concerns tends to make of classic texts like the Bodhicaryāvatāra an awkward amalgam of arguments (mostly tendentious) and the ‘roadshow ­ Powerpoint presentation of a motivational speaker’ (Siderits 2015, 129), instead of grasping the whole as engaged in protreptic argument34; while other texts,  See Hallisey (1996) and Barnhart (2012).  or an attempt to ‘turn the soul’ in the right sort of way, by the right sort of means. Carpenter (2019a) argues that this protreptic reading does not make the text any less philosophical, or any less in the business of offering reasons. 33 34

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

395

such as Vasubandhu’s Abhidharmakośabhāṣya or Twenty Verses are read stripped of any ethical purpose whatsoever.35 A final sort of motivation for taking Buddhist ethics to be a form of virtue ethics brings us to the edge of the usefulness of the categorisation. This is the will to see Buddhist ethics as naturalistic, in spite of karma.36 Just this is the good that a return to Aristotle – and not Plato – was supposed to give us: A legitimacy in our talk of good and bad which did not rest on ‘spooky’ entities (such as ‘rational intuition’), but which also did not unrealistically externalise all good into the biological incentives of pleasure and pain. Those interested in classifying Buddhist ethics as virtue ethics may be motivated by the thought that doing so grants Buddhist ethics this attractive feature of virtue ethics – namely, maintaining robust talk of good and bad without relying on anything that is not found in nature.37 From this angle, Buddhist and Aristotelian ethics can look like happy bedfellows indeed. But on closer examination, this easy fellowship starts to unravel. Buddhists, after all, insist in their various ways that there is no substantial self (no enduring agent-­ subject unity38) while Aristotelian virtue theory presumes a self as the bearer of traits, underlying moral development, and ensuring it is me who was bad and is now good39; surely this alone debars Buddhists from being Aristotelian-style virtue ethicists. Yet this does not quite get to the heart of the matter. The Aristotelian commitment to a substance-self40 may be incidental to virtue theory,41 and a less metaphysically loaded sense of self suffice to sustain talk of virtue and character development. A Buddhist’s bundle of inter-connected dispositions, tendencies, feelings and desires may do the job of securing talk of moral character, without positing some core substratum to which the virtues belong. Instead of a virtue being a  A view which their respective auto-commentaries belie. For instance, the opening of AKBh. I, and the framing remarks of AKBh. IX (the Sanskrit can be found in Pradhan (1967), the English translation in Pruden (1991)) both frame what is to come in terms of the final goal (enlightenment), and the means to attain it (which I do not take to contrast with some other ‘moral’ domain; see note 3). The Twenty Verses (English and Sanskrit both found in Anacker (2005)) places its arguments for mind-only (prajñaptipātra) in the context of the path at Viṃś-A 10 and 21–22. 36  which perhaps can itself be naturalised; see Wright (2005) and Keown (1996). 37  Although he rejects the virtue ethics label for Buddhist ethics, Siderits himself shares this motivation towards naturalism, and typically defending distinctively Buddhist positions that are naturalistic, or consistent with naturalism. Westerhoff describes Siderits’ naturalism as ‘a view that begins from an epistemological position, seeing science as our best route to knowledge’ (2016, 2). Of course, the very question of what is and is not found in nature is itself a contested question; see Meyers (2018) for thoughtful discussion of the question in this context. 38  According to Vasubandhu, “The three kinds, of grasping after self are grasping for one central entity, grasping for an ‘enjoyer’, and grasping for a ‘doer’,” Discussion of the Five Aggregates (Anacker 2005, 74). 39  As observed by Siderits (2016), 272. 40  Individual animals, and indeed human beings, are the paradigm examples of ‘primary substances’ in Categories 5 and at Metaphysics Z.8. 41  Indeed the Buddhist even argues that such a substance self positively precludes the possibility of moral development. 35

396

A. Carpenter

disposition of the soul, it might be a pattern of occurrent mental events and conditions for other mental events and actions located within a psycho-physical bundle – but this change is no significant impediment to conceiving of a view along virtue ethics lines.42 If Buddhist non-categorial metaphysics can at all provide a viable alternative to Aristotelian categorial metaphysics, then it should be adequate to providing a somewhat revisionary but sustainable account of the virtues (and if it cannot, there are bigger problems here). The more insurmountable difficulty in this area is not selves as substrata, but selves as agents – which Aristotelian adult humans are. Indeed it is this agency that makes the doings of Aristotelian selves liable to praise and blame, and makes humans capable of that distinctively human good, eudaimonia. Actions, in order to fall within the domain of the moral at all, must be voluntary, or ‘up to us’ (eph’hēmin, EN III.1–5). This is not quite setting ‘free will’ as a requirement on moral responsibility, but it is the seed of that tradition.43 It sets up the capacity for decision as the crux of moral value, even if particular decisions and moments of choice remain evaluable as expressions of virtue.44 Contemporary virtue theory might underplay this theme, but the Buddhist outlook is fundamentally different in kind regarding agency. Intentional action may be related to distinctive effects and warrant distinctive responses from us as most efficacious in reducing suffering; but Buddhist interest in the quality of one’s intentions in no way leads to an interest in whether our choices are ‘free’, or our actions ‘up to us’,45 and there is no special kind of moral responsibility arising exclusively from moments of choice.46 For the Buddhist, deliberation and choice hold no privileged position in characterising ‘who we really are’; we are not essentially or truly deliberators, and so deliberating collectively about the human good in particular plays no special role in constituting human

 “A Buddhist Reductionist,” Siderits observes, “might welcome a narrative view of self” – such as is supported by the neo-Aristotelian Alasdair MacIntyre (Siderits 2011, 299). 43  See Sauvé-Meyer (2014) for discussion of how it is not a ‘free will’ claim in Aristotle (cf. Frede 2014); and Frede (2011) for insightful discussion of how that claim arose from this tradition, via Stoicism, and flowered within Christian thought. 44  How choice as Aristotle discusses it at EN III.3 relates to responsibility when we take into account responsibility for our character and its development is an enormously complicated question (on which Bobzien (2014) is careful and illuminating; for Aristotle on moral responsibility generally, see Sauvé-Meyer 1993/2011). The use of decision or will to mark out the morally responsible from the rest took on a life of its own over the course of the development of European moral thought, a life closely related to the emergence of ‘free will’. 45  This may partly explain why the question never arose in those terms among classical Indian Buddhists. See Siderits (2008) for a rational reconstruction of what an Abhidharma position on ‘free will’ might be; see also Riccardo Repetti ‘s (2010, 2012a, b, 2014) for a comprehensive state of the discourse and useful bibliography of the Buddhist free will discourse. Repetti, ed. (2017) contains excellent and nuanced contributions to the discussion from contemporary scholars. 46  Unlike choice, intention (cetanā) – in a very wide sense of that term that covers awareness of what one is doing (or lack thereof, Milindapañha II.6.viii) – does have a special connection to action in most Buddhist texts. Carpenter (2017a) considers how the significance of quality of motivation can be maintained without intention marking out a special domain as ‘moral’. 42

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

397

goodness.47 The Aristotelian agent must still be a special sort of uncaused cause, autonomous, voluntarily initiating action. But this autonomous agency, or the special sort of choice that is truly us and our own (the excellent functioning of what most makes us what we are), is for the Buddhist just another version of the mythical agent-self that causes us so much misery. So this is one way in which Buddhist ethics sits ill with virtue ethics: Buddhists do not just happen to have a different vocabulary or a different way of thinking about agency; they have a diagnosis of the Aristotelian way of thinking about agency as positively pernicious. And this Aristotelian concept of agency cannot be substituted for something a Buddhist could endorse without destabilising the whole organisation of virtue ethics. This is related to a second point of ill fit. When Aristotelian naturalism does avoid appeal to spooky entities like the ‘rational will’, it does so by appealing to natures.48 The nature of a thing is what makes it the thing it is, and determines what, for that thing, counts as an excellence or virtue. Buddhists of course deny there are essential natures, whether they do this Abhidharma-wise or Madhyamaka-wise. But again, this doctrinaire point alone is not yet decisive in cleaving Buddhist from Aristotelian ethics, for some neo-Aristotelians argue that we need not be essentialist in our claims about human nature.49 Instead of much-criticised Aristotelian forms or natures, virtue ethicists can work ground up, empirically; we just look and see what all human beings experience, are capable of, and confront due to their particularly human situation, and from this generate an account of the distinctively human and thus what is the human good. What could be more Buddhist in spirit than to work from experience as our baseline? According to the much-cited ‘Kālāmā Sutta’ (AN iii.65) even reasoning is not to be trusted over our own experience (PTS A.I.190). The difficulty, however, is not the virtue ethicist’s account of human nature, but the use she makes of it. A virtue is a characteristic that makes that which has it good; and which features make it good depends upon what it is. In the case of human beings, it is the fact that humans are such-and-such-like that determines which states of any particular human beings are virtues – namely, those states which enable one to enjoy distinctively human forms of flourishing, or to avoid characteristically human vulnerabilities, and in general to be a good thing of the human kind. This becomes the naturalistic ground on which ethical value is built. On a virtue theoretic account, good human beings are good at being human beings. In the perfectly virtuous, human nature is perfectly realised, and virtues are those qualities of the individual which both enable and constitute this perfect realisation of human nature. Thus our primary practical aim and ultimate goal is to become good-things-of-ourkind, exemplars of humanity.

 Contrast neo-Aristotelians such as MacIntyre (1981) (‘The good life for man is the life spent seeking the good life for man’, 219); Hursthouse (1999), Ch. 10, On Virtue Ethics, Chapter 10; McDowell (1995), esp. §10; and even, in a qualified way, Kraut (2007), §53. 48  As I read him, something like this objection and the previous are what lead Siderits (2015) to conclude that Indian Buddhist ethics, alone among classical Indian ethics, cannot be eudaimonist. 49  Richard Kraut (2007) is a prominent example of such an approach. 47

398

A. Carpenter

But now just consider the crashing dissonance of supposing a Buddhist hortatory text enjoining you to become a good thing of your kind, or to be the best human being you can be.50 There is for the Buddhist no ‘good for a human being’ distinct from ‘good for an elephant’ – the elephant Pārileyyaka is good and virtuous for just the same sorts of reasons as any human being (compassion and devotion to the Dharma or to the Buddha) – and likewise the monkey, partridge and elephant of Vin. ii.161–162.51 Although this is often overlooked, the same is true for Plato – unlike Aristotle, Plato is relatively uninterested in what makes a human being human. It is, of course, useful to know the material we are working with when we are trying to improve it; but what makes a human being human is not a source of natural normativity, on which a theory of virtue depends. On the contrary, it is only the Good itself that makes the difference in whether justice or temperance is good or bad52 – and even these are not qualities special to human beings because of what human beings are. Both Buddhists and Platonists, then, have in effect no particular interest in the human good.53 On both views, our ultimate aim is simply to become good, full stop – and if we cannot become perfectly good (because both views set this bar very high), then our life’s task is to strain every nerve to come as near to it as possible.54

 One might be tempted to think this was because of karma-rebirth theory; yet without appeal to it, Plato does not think in such terms, while with karma-rebirth to hand, the ethos of the Rāmayāna is perfectly able to put forward Rāma as the perfect man. 51  quoted by McDermott (1989), 270. Contrast Aristotle’s claim that only adult human beings can be eudaimon (EN I.9); only they can be properly praised or blamed, and only they can be virtuous. Even human children are only so-called as a courtesy, in light of what we hope they will become (EN I.9, 1100a2–5). 52  Republic 504a-509b; cf. Charmides 161a1–2. In the Philebus, only forms of goodness (proportion, beauty and truth) determine which things make a life good, practically bye-passing virtue altogether. 53  This is obscured in the Plato discourse by the way that scholars such as Vlastos, for instance, elide Socrates’ talk of ‘good’ and ‘happiness’ into a single subject, ‘the human good’ (Vlastos 1999, 109n20). Plato may seem closest to posing the question of a specifically human good in the Philebus, which sets the agenda with a bald claim that ‘pleasure is (the) good’ (Philebus 11b4), unqualified by whose good it may be; against which Socrates gives as his own view that cognition is ‘better and more desirable than pleasure for all things insofar as they are able to share in [it]’ (Philebus 11b8-c1). Contrary to Vogt (2017), I take one of the primary lessons of the Philebus to be that even the ‘second contest’ into the cause of goodness in a life (Philebus 22d) cannot be addressed without reference to the good, grasped here in three forms (beauty, proportion, and truth, 65a1–2), which is the necessary orientation point and criterion for understanding goodness in a life. 54  This focal point for development given by a fixed ideal (the Good, Awakening) is what may raise doubt that particularism is the most apt way to characterise either Buddhist or Platonic moral thought. 50

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

399

2 Reality Is Impersonal; the Task Is to Know It The disinterest in a specifically human good rests in both cases on the impersonal nature of reality, and the way we should relate ourselves to it. For Plato, the only source of goodness is the Good itself, the wholly impersonal cause of an intelligible reality undifferentiated according to human and non-human, devoid of anything essentially personal. It is this impersonality of Plato’s Good that Aristotle objects to when he complains in Nicomachean Ethics I.6 that to investigate such supra-­ categorial good (were it even incoherent) would be useless for ethics, which directs our attention to the human good, determined by human capacities, and their deployment in actions.55 Plato’s ‘Good’, Aristotle rightly saw, is not beholden to individual, or even human concerns. The reality it stands for is wholly impersonal. It does not contain persons or the central categories of personhood (agent, action, decision, choice). On the contrary, if we are to deploy such concepts as agent, action and choice aptly and fruitfully, then we must understand them and their relevance in terms of the impersonal, intelligible reality that is prior to any merely human perspective or concern. Our task with respect to this reality, and the Good itself, is to know it.56 But doing so will not, as Aristotle rightly saw, give us maxims for action or even information about what the virtuous person would do in a particular situation. Rather, the right relation to reality – namely, knowing it – is itself the morally improving activity. It will so shape our perceptions of everyday, particular and messy realities as to engender motivations which cause us to cause goodness in and around us, as the opportunity arises. Similarly, the Buddhist focus is on knowing reality as it is – and how it is, is personless. The first Noble Truth is to be known; and from Buddhaghosa in the Theravada tradition to Śāntideva in the Mahāyāna, the perfection of wisdom or insight is the culmination of the whole trajectory of ethical development, so that from beginning to end the Buddhist path is permeated by the orientation towards

 “For even if there is some one good predicated in common, or some separable good, itself in its own right, clearly that is not the sort of good a human being can achieve in action or possess; but that is the sort we are looking for now” (EN I.6, 1096b30–35, tr. Irwin and Fine). “Moreover, it is a puzzle to know what the weaver or carpenter will gain from his own craft from knowing this Good Itself, or how anyone will be better at medicine or generalship from having gazed on the Idea Itself. For what the doctor appears to consider is not even health [universally let alone good universally], but human health, and presumably the health of this human being even more, since he treats one particular patient at a time” (EN I.6, 1097a10–15). 56  Thus Plato makes no distinction between practical and theoretical wisdom (see Lane 2005, 343). Plato’s concern with knowledge and what is fully knowable is pervasive in the dialogues; the Symosium’s discussion with Diotima, the Phaedrus’s palinode, the central books of the Republic and the culmination of the Philebus are conspicuous expressions of the unparalleled value attached to knowing reality. 55

400

A. Carpenter

knowing an impersonal reality.57 Practices of seeing reality as it is are improving and clarifying all along the way, even if different Buddhist thinkers have different views of what the reality is into which perfected wisdom has insight.58 Such insight is not just liberating (the insight into reality itself sets one free from suffering); it is awakening. The reality one is awaked to is devoid of persons as elements, and not properly characterised by the categories proper to personhood (agent, subject, individual, unifier, controller). If the aim is the transformation of our phenomenology, awakened to reality is that transformed phenomenology, not merely the external route to it. And, as with Plato, the discipline of knowing impersonal reality as it is should also have knock-on effects on action via the transformation of our perceptual and motivational engagement with the particular, transient world around us. The Buddhist no-self claim, in its many competing varieties, is the claim that individual persons and the categories pertaining to them are not fundamental or prior, and in this there is substantial agreement with the Platonic perspective according to which real reality is an impersonal Good which grounds the being of fully intelligible and impersonal reality of eternal Forms.59 Whether knowing reality is knowing the fixed explanatory relations between unchanging non-sensible intelligibles, or whether it is knowing reality as inter-dependent, transient and suffering, the reality to be known will not contain or be governed by the central categories and features of personhood. In both cases, our central task is to know this reality, and to understand our experiences in light of this and in relation to this. And in both cases, such insight transforms the arising mental events constitutive of character and leading to action.

 Consider as indicative the Abhidharmakośabhāṣya’s opening praise of the Buddha: the first quality to praise in the Buddha is that he ‘has destroyed all blindness’, where ‘blindness is ignorance, for ignorance hinders the seeing of things as they truly are. By this, the Buddha, the Blessed One is sufficiently designated’; he is secondarily praiseworthy for having thereby ‘drawn out the world from the mire of transmigration’. 58  and even dispute over whether it is, in a certain sense, real, as when Siderits disputes that Nāgārjuna is making metaphysical claims at all, and interprets the ultimate truth to be that there is no ultimate reality (a ‘semantic’ interpretation, on which understanding emptiness means ‘one abandons the idea of an ultimate nature of reality’ (Siderits 2007, 182–190; and see Siderits 2003a). 59  Mann (2000), Part II, even argues plausibly that Plato’s metaphysics is not one of substantial individuals, but rather one of mixtures of property-occurrences, making Plato’s understanding of sensible reality rather akin to some Buddhist understandings. Theaetetus 184b-186b might suggest Platonic souls are at least bare unifiers of apperception, which of course is at odds with most Buddhist views; but Plato shares the Buddha’s caution about the pernicious way that ‘I’ leads to a damaging ‘mine’ (Republic V.462c). The most significant difference between Buddhist views and Plato’s, I will argue at the end, is not whether there is a unified soul, but whether we ought to aim to make ourselves unified. 57

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

401

3 Knowing Impersonal Reality and Ethical Formation So on both views, reality is impersonal, and our task is to know it. Moreover, accomplishing this task should have a transformative effect on our dispositions, inclinations, thoughts, beliefs, feelings. Afflictive emotions will be permanently quelled, and there will be consequently no impetus to actions that give rise to suffering for self or others. Socrates in the Republic (500b8-c2) notes this connection between attending to impersonal reality and emotional engagement with persons, “No one whose thoughts are truly directed towards the things that are has the leisure to look down at human affairs or to be filled with envy and hatred by competing with people”. It may well be that, in contrast to Plato, there is not, according to many Indian Buddhist lines of thought, some other reality we should know, apart from the one around us. But the practical implications of seeing reality selflessly are still evident, even in early Buddhism, where it is disputed whether the Buddha actually thought that there is no self, or indeed cared much about metaphysics at all.60 What the Buddha certainly did recommend was to stop thinking in terms of self: stop seeking the self, stop looking for things to identify with, or supposing the answer is to be found by ‘looking inside’ and discovering the ‘real you’. It is theories of the self,61 the associated assertions that ‘this am I, this is mine’,62 obsessive concern with getting this right – as if by this I would have laid hold of something truly valuable63– that causes the misery. Seeing reality impersonally, such damaging mental practices of identification, self-assertion and appropriation become incoherent, and ultimately find no foothold. Turning our attention towards this impersonal reality does not so much inform us as it forms us. Directing attention away from ourselves, away from the pre-­ constituted human-shaped familiar ways of categorising experience, and instead orienting ourselves towards impersonal reality, gives us categories, priorities and perspectives which provide a truer understanding of everyday, messy, sensible reality. The locus classicus for this in Plato is the philosophers’ return to the cave in Republic VII, where after grasping the wholly impersonal structure of  For an excellent discussion of the original indeterminacies in the early sūtras, see Priestly (1999), Chapter 1. Schmidthausen (1973) and Frauwallner (1953) deny that outright rejection of self was an original part of the Buddha’s teachings. 61  “I too do not see any doctrine of the self that would not arouse sorrow, lamentation, pain, grief, and despair in one who clings to it” (MN 22.22). See also the claim that moral improvement is made impossible by any view identifying the self (SN 12.35). 62  ΜΝ 22; cf. Ratnāvalī I.27–28. Harvey (1995), discussing a text from the Khuddaka Nikaya, (Niddesa II.278–82, in particular), notes that “Self is practically equivalent to ‘what pertains to Self’, I, mine, ‘I am’,” (Selfless Mind, 50). Vasubandhu confirms this. Rejecting “the view of self has two parts: to say ‘I’ and to say ‘mine’,” he observes “If the idea of ‘mine’ were different from the idea of ‘I’, then the ideas expressed in the other grammatical cases, such as mayā (by me) or mahyam (to me) would thus constitute so many new views” (AKBh. V.9a-b, tr. Pruden). 63  MN 109 argues against holding a view of self on the grounds that there is nothing worthy of being considered a self (cf. MN 22.26–27); by implication in looking for the self, we are looking for something worthy to be the self, not just something that solves a metaphysical puzzle. 60

402

A. Carpenter

reality – located outside the engulfing cavern of the human social world – the philosophers recognise all the particular things they had known before their transformative appreciation of impersonal reality, but now they can correctly appreciate these things for what they are (Rep. 520c). This is familiar in Buddhist discourse as the phenomenon of recognising our conventional designations as conventional. To improve ourselves, then, we do not introspect and tinker with our insides; we look rather outside ourselves, to reality as distinct from wishful appearance – for Plato, to the impersonal good itself which informs the structure of reality; for the Buddhist, to the transient, essenceless mutual dependency of reality which cannot in truth sustain the construction our desires and fears put on it. Doing this not only changes our minds – we literally have different thoughts; it changes our mode of engagement with, indeed our very perception of ordinary, everyday matters.64 By recognising of our old person-centred categories and concerns that they are constructed  – by circumstances, society, by desires and ignorance – we thereby recognise that they are not forced on us by reality. Our relationship to these categories and concerns is transformed tout court – we are no longer in the grip of them, driven on by their logic. We are liberated from compulsive behaviours and emotions by seeing reality. So similarly for Plato true knowledge of what impersonally exists liberates us from the tyranny of sensual desires, pleasures and pains, which ‘make the soul believe the truth is what the body says it is’ (Phaedo 83d6–7). It does so by focusing the attention on an impersonal reality, which guides our understanding of and relationship to everyday experience. Understanding reality as it is liberates us from the compulsions of wayward appetites and misguided pride.65 The liberating effects of knowing impersonal reality arise from the comprehensiveness of the process or activity of reaching understanding. The sort of knowing we strive for commandeers, organises and orients every part of the soul. “[T]he power to learn is present in everyone’s soul,’ as Plato has Socrates say (Republic 518c-519a), but it “cannot be turned around... without turning the whole soul”. Thus every discussion of virtue in Plato turns into a discussion of knowledge, and discussion of knowledge is an opportunity to ‘turn the whole soul’ from its engrossment in the everyday familiar affairs towards the unchanging intelligible reality that is necessary, on his view, to make proper sense of sensible reality. Attend with your whole soul to intelligible reality, says Plato, and transform your whole soul – your values, ambitions, expectations, hopes and desires66; what is ordinarily recognised as good action will follow (cf. Rep. IV.441e-44e). This reverses the direction of neo-Aristotelian moral development, according to which we become just by doing just deeds. Through doing right actions, says the neo-Aristotelian, we become better able to appreciate the right reasons for them, and from this come to acquire a stable motivation to do just deeds because they are  I argue the case specifically for the Philebus in Carpenter (2006) and Carpenter (2021).  This theme is a woven throughout the Republic especially, but also the Gorgias, and Symposium and Phaedrus. 66  I explore the mechanisms of this with respect to the Philebus in Carpenter (2017b), and with respect to the Republic in Carpenter (2023). 64 65

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

403

just. Eventually this motivation becomes part of how we look at the world and new situations. Buddhist ethics, by contrast, like Plato’s, typically leads with understanding, in this sense: the priority is revising our basic outlook, what we see when we look at the world, whether this be effected through engagement with narratives, with arguments, or with explicit practises in recognising impersonal reality as it is.67 This prioritisation is the fall-out of recognising that all our choices and ‘rational deliberations’ are embedded, not autonomous; they arise as they do due to mental causes and conditions.68 If we expect the choices to change, we need to change their causes and conditions – the mental factors of perception and conception that lead to them. Our fundamental understanding of reality operates at the level of the background conditions of any more specific experience, so that changing this basic understanding transforms the way we experience particular, everyday situations – in particular, we experience them in ways that do not invite envy, resentment and fear, but rather as occasions for compassion and care.69 Feeling care and joy, lacking resentment and fear, motivations for harmful actions are simply absent. Seeing reality sub specie doloris opens up new possibilities for responsiveness. Such wholesale transformation of fundamental outlook or orientation does not come by getting to the QED at the end of a proof about the nature of reality, but works rather at the level of what makes it possible to draw the conclusion and appreciate its meaning. Such reorientation and reworking of the categories through which experiences take shape at all requires sustained, repeated engagement.70 This is why both narratives and, for Buddhists, meditation practices, have the ethical significance they do.71

4 Modes of Transformation: Meditation and Narratives Although Plato emphasises the repeated practice and effort required for wholesale transformation, the Platonic tradition did not hand down a set of tools and array of practices by which we might all advance further in seeing reality as it is,  “I can only choose within the world I can see”, writes Iris Murdoch (1970, 36).  Republic VIII and IX are studies in this phenomenon; and consider also the Timaeus’ advice that ‘it is not right to reproach people’ for their wicked deeds, for all bad deeds are ‘a result of one or another corrupt condition of the body and an uneducated upbringing’, neither of which anyone would will to have (Tim. 86e). See Carpenter (2017a). 69  Plato does not use the language of care; but he insists that goodness wants only goodness and has only good effects on others (Republic II), and that the gods, being good, have no envy, but rather want everything to share in the goodness they enjoy (Timeaus). 70  The training for philosophers in the Republic includes ten years of higher mathematics before even beginning a five-year training in dialectic, which only concludes fifteen years later with comprehension of the Good (Republic 538b-d,539e-540a; cf. Meno 85c8-d2; Parmenides 135d-36a). 71  On the way that narratives might specifically target our way of reading the world and reality, see Hansen (2002) and Carpenter (2019, 2020a, b). For meditation as an ethical activity, see Dreyfus (1995). 67 68

404

A. Carpenter

transforming ourselves fundamentally along the way.72 The so-called ‘Socratic method’ is hardly methodical; and there is no clarity or consensus about how the much-vaunted ‘dialectic’ should actually be practiced. The Buddhist tradition, by contrast, developed and transmitted a handsome suite of meditative practices. Buddhaghosa’s 5th C. c.e. Visuddhimagga (Path of Purification) is one comprehensive and authoritative source for these, and dedicated meditation manuals (such as Yogāvacara’s Manual73) and Abhidharma handbooks which incorporated detailed meditation instructions (such as Anuruddha’s Abhidhammattha Sangaha) are not uncommon. A great part of these meditational exercises aim at appreciation of reality as non-­ self, transient and suffering, whether locally or globally. Common meditational practices, which predate Buddhaghosa, such as meditation on the breath or the body, tend to take something we are ordinarily personally involved in and change our relation to and experience of it into something impersonal. ‘In this way, one abides contemplating the body as body...its nature of both arising and vanishing. Or else mindfulness that ‘there is a body’ is simply established in one to the extent necessary for bare knowledge and mindfulness. And one abides independent, not clinging to anything in the world’ – and so on for the feelings, thoughts, and objects of thought or perception (MN 10).74 Through meditation, one knows person-­ constituting elements as transient, and regards each as being simply itself, not attached meaningfully to any narrative. This is detachment. Of Buddhaghosa’s forty suitable meditation objects for the practitioner looking to advance along the path, seventeen are clearly impersonal,75 while a further twelve de-personalise something ordinarily strongly regarded as personal.76 Meditations on the ten foulnesses, for instance, like exercises in the Satipaṭṭhāna Sutta, focus on the (dead) human body. Practices of careful definition of the corpse in a specific state of decay (by its colour, its shape, its direction and location; by its joints, its openings, its concavities and convexities), converts it into something we regard impersonally, and therewith both sensual desires and the concomitant revulsion towards corpses is dissolved (Vism. VI). Recollections of the Buddha, Sangha and deities might be expected to be invitations to meditate explicitly on persons  – just the opposite of cultivating an  Pierre Hadot’s Philosophy as a Way of Life has become a standard point of reference for understanding classical Mediterranean philosophies as practices. None of these, however, show us how to practice what Plato recommends. Certain elements of Stoic practice come close, but their emphasis on prohairesis (rational choice) writes person-involving categories ineradicably into the fabric of reality. Plotinus and those influenced by his reading of Plato may have had such practices, but communities of practitioners never became widespread. 73  translated as Manual of a Mystic by F.L. Woodward (1961). 74  Ñāṇamoli and Bodhi, trs. Middle-Length Discourses of the Buddha (Majjhima Nikāya); the sutta is repeated in an expanded form at DN 22 (Long Discourses of the Buddha, tr. Walshe). 75  the ten kasinas (earth, water, fire, air, blue, yellow, red, white, light, limited-space); the four immaterial states (boundless space, boundless perception, nothingness, neither perception nor non-perception); the recollections of Dharma and cessation; the defining of the four elements. 76  The ten foulnesses; the Recollections of the body, breath, and death. 72

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

405

impersonal view of reality. But as Buddhaghosa describes this exercise, meditation on the Buddha, for instance, has nothing to do with recounting the legend of Prince Siddhartha, or the noble and selfless deeds done on the path to awakening, and everything to do with rehearsing reality as the Buddha taught us to see it, such that he brought suffering to an end. In fact, on Buddhaghosa’s account the Buddha becomes oddly transparent. To contemplate the Buddha is to contemplate his teaching of reality as selfless, dependently arising, and transient; so that in looking to the Buddha, we look through the Buddha to his Dharma, and specifically to his teachings of the personless nature of reality. While Buddhaghosa in other places rehearses various tales of moral success and failure,77 here in the Recollection of the Buddha, the person of the Buddha becomes wholly transparent, dissolved into the impersonal reality which liberates when seen. Transformation is effected not by doing as the good man does, but by seeing as the good man sees. But the brahmaviharas  – especially the cultivation of goodwill, care and joy (metta, karuṇā, and muditā) – look necessarily person-involving. One cannot ubiquitously exude the wish, ‘May all beings be free from enmity, affliction and anxiety, and live happily’ without directing it towards beings (satta), understood ‘in accordance with ordinary speech’ (Vism. IX.55).78 And yet even here there are surprising elements tending towards impersonalisation, and knowing reality as devoid of personal characteristics. Several of the techniques Buddhaghosa recommends encourage us to dissolve persons into non-personal elements, or to regard reality impersonally. More significantly, the first three brahmaviharas are followed by a fourth, deeply impersonal Divine Abiding, upekkhā – specifically introduced to correct the personalisation inherent in the previous three. One must ‘see the danger in the former [three divine abidings] because they are linked with attention given to beings’ enjoyment..., because resentment and approval are near, and because their association with joy is gross’ (Vism. IX.88). Thinking in terms of beings may be necessary temporarily for proper self-cultivation.79 Yet there is a danger inherent in it just insofar as it requires thinking in terms of beings, and a deliberate corrective to this must be built into the Path. Ultimately, even the Divine Abidings point towards prajñā, the insight into reality as suffering, transient, and impersonal. According to Asaṅga’s alternative account of the brahamviharas, they are not even essentially person-involving in the first place. A form of loving-kindness directed at beings is to be supplanted by a form directed at impersonal elements; and this in turn supplanted by utterly objectless loving-kindness.80  Heim (2014), Chapter 4 discusses Buddhaghoṣa’s use of Buddhist tales in his work.  Visuddhimagga IX.50, tr. Ñāṇamoli (1976). This defines metta cultivation in particular; karuṇā cultivation consist in enlarging the feeling that ‘This being has indeed been reduced to misery’ – or may become so – ‘if only we could be freed from this suffering!’ (Vism. IX.78–81); and muditā, ‘This being is indeed glad. How good, how excellent!’ (Vism. 85). 79  Compare Śāntideva’s advice that if I-grasping increases due to compassionate engagement with (what are delusionally thought of as) beings, practice selflessness (BCA IX.76–77). 80  Asaṅga, Bodhisattvabhūmi, translated by Atermus B. Engle, The Bodhisattva Path to Unsurpassed Enlightenment. Snow Engle (2016), 402–402. 77 78

406

A. Carpenter

In addition to meditation exercises, narratives are another key tool the tradition deploys for inculcating the ethos, the habits of mind – of orientation and appreciation and responsiveness – constitutive of a wholesome way of being in the world. Republic II-III infamously insists that children be exposed only to the right kind of music and poems. Plato recognises that the stories we are told, and how they are told, present us with a pre-theoretical sense of the space of reasons and actions. They shape our sense of which concepts and categories are available to be deployed in making sense of experience. While Platonists, however, left no body of actual stories expected and proven by experience to transform the soul in the appropriate way, Buddhist communities by contrast adapted and generated a rich body of narrative literature of the sort that Socrates in the Republic only talks about. The telling and retelling of Buddhist narratives constitutes an induction into and inculcation of an outlook on reality as dependently arising and without self, everything at once affected and affecting, without any locus of overall control.81 The narratives show us a world that is familiar, yet repeatedly shown in the unfamiliar light of insight into the fundamental mutability of the world, its lack of purpose, of providence, and of self-determining agents acting freely upon rational deliberation of the options. One of the modes by which this inculcation is effected is the repeated presentation of the Buddha, the Bodhisattva, or accomplished arhats, engaging with and describing situations and persons in ways that reflect their own appreciation of the deep personlessness of reality. The repetition and the minuteness here of perception is important for the effect. Consistently, devotion to the Buddha and Dharma is admired (particularly in the avadānas); repeatedly arhats view themselves and their actions as opportunities for others to become good and exercise virtue. Over and over again beings are traced over lifetimes, and repeatedly the Buddha’s perspective highlights the hazards and futility of attempting any ultimate judgements of origins or blame.82 One of the mechanisms for displacing our ordinary view and replacing it with a better one is the drastic under-characterisation of the stories’ moral examplars. The jātakas may trace the Bodhisattava’s trajectory over countless rebirths; but they do not give us stories of his personal emotional development. There is nothing personable about the Arhats and accomplished laity we encounter in the tales, and we no more identify with them than with the perfected Buddha. There is no character development that would help us to see how we, too, could become like that. The  Hansen (2002) brings out the centrality of dependent arising conveyed through narratives, and the Gatilok in particular. Heim (2014), Ch. 4 draws out the inter-personal entanglement complicating intention in the narratives. Hallisey and Hanson (1996) emphasise how the karmic configuration of the stories creates possibilities for Bandhula to be both innocent and yet getting what he deserved. ‘The Demoness Kali’ beautifully illustrates, as Obeyesekere and Obeyesekere (1990) show, Buddhist narratives’ tendency to deliberately confuse and obscure any purported clear lines of agency and individuation. Historical narratives may similarly serve to convey basic understandings of reality and the character traits and responses appropriate to that (Berkwitz (2003), which because it is heavily Aristotelian conducts the discussion in terms of emotions, although the Pāli for gratitude, as Berkwitz tells us, may be literally translated as ‘knowing what was done’). 82  Carpenter (2019b) discusses in detail the connection between of narratives and blame. 81

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

407

radical sparsity of characterisation of persons the Buddhist tales present as normative means we cannot use personal engagement or identification with these characters in the way Martha Nussbaum, for instance, presents the moral edification of engaging with Henry James’ novels.83 The mode of presentation positively resists this, offering nothing of the personalities of accomplished persons for us to become engrossed in and distracted by. This is not an accidental feature of the genre, but directly related to the ethos to be inculcated.84 Our liability to become distracted by personality continually confounded Socrates’ attempts to turn his interlocutor’s attention away from him and towards a shared examination of reality; and Plato thus continually highlights the moral danger in attaching oneself to any particular person, including Socrates, instead of looking directly to impersonal reality, or the Good itself. The exemplary persons of Buddhist narratives resist such attachment in the same way that Plato’s Socrates rejected attempts to make him into something special  – but much more effectively. Both Buddhist and Platonic exemplars are thus distinctly self-effacing. If the view they are articulating and recommending is correct, then their personalities are the least interesting – the least real, certainly the least important – thing about them. As in Buddhaghosa’s account of the specific meditative practice of recollection of the Buddha, here too in the Buddhist tales we see a transparency of moral exemplars – to see an arhat or a bodhisattva is to see through them to the way they see the world. With little else by way of engaging personality or psychological drama to be distracted by or to hold onto emotionally, this outlook recommends itself to the auditor or reader directly, and then becomes available as a perspective they might adopt for themselves. In place of ‘doing as the good man does’, the ethos which focuses on the primacy of knowing impersonal reality advocates ‘seeing as the good man sees’. Orienting oneself towards impersonal reality pulls one out of oneself, not just in the ordinary sense of making one less self-absorbed (and taking in other persons’ equally self-absorbed perspectives instead); this is rather a practice that

 Further discussion of the importance of this difference can be found in Carpenter (2019, 2020). Compare here Plato’s aversion, in Republic II, to story-telling which requires impersonation, because of the negative effects of identification. 84  There is much to be said about the genre of Buddhist narratives, which are neither fable nor folktale (though some folk-tales were appropriated and repurposed) nor quite parable; and within which there are significant and interesting species — jātaka, avadana, Dhammapadāṭṭakathā collection, and later renderings of these (Ārya Śūra’s Jatakamāla, for instance) each being quite different in kind, construction and theme from each other, yet preserving across these differences this intentional effacing of personality. Such effacement is not ubiquitous in ancient narrative, nor in short ancient narrative in particular, and it is handled in a deliberate way, with specific effect, within Buddhist story-telling. 83

408

A. Carpenter

pulls us out of the personal altogether. The very categories needed to make sense of pride, anger, greed, and so on are displaced – or in fact vanish altogether.85

5 Conclusion Both Platonic and Buddhist ethics offer an ambitious ethical idealism quite different from what is usually called ‘virtue ethics’. Neither supposes that the human is the proper measure of our ethical task. Each rather takes there to be something profoundly out-of-joint with the everyday world and our ordinary experience of it; each would, therefore, categorically reject the Aristotelian impulse to suppose that, generally and for the most part, most of us have got it mostly right, morally speaking, so that we could look to the good deeds of the good folk around us in order to learn what is good. On the contrary, on both of these views we aim to attain ideal comprehension of impersonal reality. Pursuit of such knowledge organises all other facets of moral development, transforming the whole of our phenomenology, and changing the most basic ways in which the everyday appears to us.86 We are not just directed away from our individual selves, but from the whole way of looking at the world as involving persons.87 Seeing reality as impersonal frees us from the clamouring demands of the personal and social, enabling us to see and to feel particular situations as articulations of ultimate reality, be that suffering and dependent-arising, or the intelligible Good Itself. Does it, in the end, make much difference, ethically, whether this impersonal reality is the intelligible Good or transient dependent arising? Perhaps orientation towards reality is good for us regardless, just because it is a disciplining of our experiences according to the preference for reality over appearance.88 Yet even if this is so, there are still significant differences in how a practice of knowing impersonal reality affects us, according to whether that reality is Platonic or Buddhist. The differences arise, however, not simply from the metaphysics, but from the accompanying radically different conceptions of what it is to know such a reality. The ethical differences between the Buddhist and the Platonic forms of impersonal idealism arise most sharply from their epistemological differences.  If person-involving concepts are later built back into our thinking and motivations, they are radically revised, determined by and subordinated to impersonal concerns. Think here of the philosophers returning to the cave, who see ordinary things as the partial approximations of real beauty, unity, etc., that they are Rep VII.520; think also of the bodhisattva who remains active in saṃsāra by recognizing that what are known to be impersonal processes are conceived and felt by others as objects and projects. This transformed reintroduction of essentially person-involving categories is in the area of what Siderits (2003b), 99–111, calls ‘ironic engagement’. 86  This point is highlighted by Murdoch (1956, 1970). 87  It is perhaps no surprise then that both parties have faced criticism of not allowing to personal relationships the sort of overriding importance they are commonly expected to have. Regarding Plato, see Vlastos (1981). 88  I emphasise the good Plato saw in this minimal truth-orientation in Carpenter (2017b). 85

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

409

Plato’s reality is, above all, intelligible – as is captured by the older convention of translating ἰδέα and εἶδος (both of which have connotations of cognition89) as Idea. That the intelligibility of reality requires stability, and independence of changing sensibles, tells us much already about Plato’s famously robust conception of knowledge. To know – in the Meno, the Republic, the Phaedo, the Philebus – is deeply connected to explanation, to giving a logos (an explanatory account) of the nature of unchanging things. The Meno (98a) and the Philebus (18c-d) call grasping the ‘why’ the bond that ties our cognitions together into something sufficiently robust to meet the Republic’s high demand that we, to count as knowers, be able to defend an account of the being of a thing which we can defend against all interrogation (Rep. 534b-c). Explanation is normative, and explanations can only be complete when they are anchored in goodness – ‘that it is good so’ is the only bedrock in explanation (Phaedo 98b-99c). This means that the activity of turning towards impersonal reality is for Plato an activity of drawing increasingly comprehensive explanatory relations between things; discerning the real structure underlying the changing sensible appearances; understanding the ‘why’ of each known thing as well as of its relations, in such a way that one appreciates by what standard these may be judged good, and fitting within a well-ordered, intelligible cosmos. One of the reasons we lack a distinct suite of Platonic exercises for soul-­ transformation is because engaging in theoretical astronomy or theoretical physics or higher mathematics already counts as a reality-orienting exercise of the relevant kind.90 But the deeper reason for the lack is that, because his conception of knowledge is explanation-based, Plato takes questioning to be the exclusive route to knowledge of impersonal reality. Knowing implies an ability to explain that can only come from having posed questions one genuinely has and accepted only those answers that make sense to oneself, for reasons one grasps for oneself and can articulate.91 The effect of such attempts to understand is, above all, unifying: Striving to comprehend coherent explanations of a unified reality organises and unites our disparate psychic resources. This unification around a common and unified purpose eliminates conflicts of interest, countervailing desires, inconsistency between word and deed as well as erratic behaviour. It should also inoculate us against marketing, advertising, propaganda and other forms of manipulation which depend upon making us feel good, or which serve partial interests insulated from their relation to the real good. Orientation towards an integrated, explicable reality makes each thing appear as something to be fitted into this picture and project, taking its significance – and its attractiveness – from how it fits into this larger aim. Plato can agree  εἶδος literally means ‘what is seen’, and therefore the form, shape, look, or even beauty of a thing or person; ἰδέα can mean ‘idea, notion’, in addition to form (LSJ). Both are related to the verb εἴδω, a verb for seeing, more commonly found in its perfect form, οἶδα, to know. 90  Burnyeat (2000) is eloquent and erudite in his articulation of the ‘why mathematics is good for the soul’; Carone (1997) does the same for astronomy in the Timaeus. 91  Carpenter (2023) explores the importance of grasping and giving reasons ourselves for the ethical transformation engendered, including our accountability to one another. 89

410

A. Carpenter

that the psyche is as disunified as the Buddhists say it is; but making it into a real unity by organising our apprehensions of value around the unifying project of grasping a unified reality is a fundamental good, with the host of knock-on benefits identified. When Plato takes knowledge to be the power to explain the nature or being of what is unchanging, he explicitly contrasts these intelligibilia with whatever is known via the senses, of which there can be no knowledge (e.g., Philebus 59b). The Buddhists, by contrast, consistently prefer perception-models of knowing ultimate reality. Taking transience to be a mark of existence – indeed, as one of the most fundamental facts that we must appreciate in order to transform our experiences – it is fitting that perception (whether internal or external) is the favoured epistemic mode for capturing that reality. The Buddhists prefer in it precisely what Plato disdains. Where Plato claims that to be intelligible is to be liable to explanation, graspable in well-defined, well-related, stable and correct concepts, the Buddhist holds that this very thing is just so much conceptual proliferation; the only experience of reality as it is will be non-conceptual and quasi-perceptual. So instead of various examples of inquiry, as we get in Plato (elenchus, dialectic, the method of hypothesis, collection and division), the Buddhist offers various techniques for manipulating, controlling and improving our attentive-perceptual faculties (primarily in meditational exercises, and in narratives), so that we might clear away whatever is obstructing our direct, immediate perception of reality. For the Buddhist, practiced attention to transient reality should likewise transform our experiences, and make us immune to the distortions of the many passing phenomena  – but now these transient phenomena also include the distortions of reasoning and explanation. There is accordingly no promised unifying effect on the soul, nor would this be considered a good thing. Our practice in seeing the disunity, or the accidentalness of transient unity, should rather reduce anxiety about whether we are ourselves such a unity, or whether we have had this momentary unity disrupted. Perceiving ubiquitous dependent arising, we come to expect no overall control alongside pervasive partial responsibility. It engenders then a practical responsiveness to concrete situations which at the same time is divested of the deluded practice of isolating and blaming some ‘real culprit’ distinct from the whole set of inter-connected conditions and causes. Plato’s orientation offers continual practice in distinguishing real explanatory relations from merely accidental juxtaposition, cultivating a love of truth and reality over appearance that so informs and relates our emotional life that egoistic gratification no longer appeals; until we, like the gods, love good and want to be the cause only of good things in and for others and the world generally. The different Buddhist metaphysics favours the cultivation of an appreciation of relationality tout court, undifferentiatedly. Where Platonic inquiry promises to unify the soul, for the Buddhist there is no escape into hard-won independence and reassuringly solid individuation. There is only that escape from suffering, and from creating more suffering, that comes from recognising that this transient, mutually dependent reality is the only reality there is. Such recognition cultivates above all a habit of appreciating the inter-connectedness of things; an ethos of giving patient attention to the

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

411

multitude of causes and conditions giving rise to their varied effects, and a caution about locking down on any one factor as the cause – conversely a way of experiencing others and oneself as always co-conditioned, porous, embedded within multiple conditions without any core independent of these. This should contribute to generating a kind of concern and compassion for the suffering lack of autonomy experienced by all beings by the very nature of things. But this is not the conclusion implied by a consistent belief set. It is rather the fallout of deeply recognising reality as dependently arising, transient and no-self, and of appreciating the causes of suffering in all their fine-grained particularity. The two views have genuine differences, but these differences stem from their shared ethical commitment to knowing an impersonal reality. This shared ethical commitment is what makes their different epistemologies so salient. Seeking explanation-­based knowledge of the impersonal order of the cosmos involves materially different psychological activity and work from seeking a quasi-perceptual knowing of a non-teleological reality. As the labour in search of the end differs, one might expect the effects of that labour on the seeker to differ, too. But whatever the differences in the details here, both the Buddhist and the Platonist consider that striving for knowledge of reality will purify the emotions, draw us out of ourselves precisely because the object to know is devoid of personal connotations and connections; and this purification of the emotions, on both accounts, returns us to the particular, which we are able to see for the first time truly, and with care.

References Adam, Martin T. 2018. Moral Development in the Jātakas, Avadānas, and Pāli Nikāyas. In The Oxford Handbook of Buddhist Ethics, ed. Daniel Cozort and James Mark Shields. Oxford: Oxford University Press. Adams, Robert Merrihew. 1999. Finite and Infinite Goods. New York: Oxford University Press. Angle, Stephen C., and Michael Slote, eds. 2013. Virtue Ethics and Confucianism. New  York: Routledge. Anacker, Stefan, and trans. 2005. Seven Works of Vasubandhu: The Buddhist Psychological Doctor. Rev ed. Delhi: Motilal Banarsidass. Annas, Julia. 1999. Platonic Ethics, Old and New. Ithaca: Cornell University Press. ———. 2003. Should Virtue Make You Happy? In Eudaimonia and Well-Being: Ancient and Modern Conceptions, ed. L.J.  Jost and R.A.  Shiner, 1–19. Kelowna: Academic Printing & Publishing. Anscombe, G.E.M. 1958. Modern Moral Philosophy. Philosophy 33 (124): 1–19. Appleton, Naomi. 2016. Shared Characters in Jain, Buddhist and Hindu Narrative: Gods, Kings and Other Heroes. Routledge. Barnhart, Michael G. 2012. Theory and Comparison in the Discussion of Buddhist Ethics. Philosophy East and West 62 (1): 16–43. Berkwitz, Stephen C. 2003. History and Gratitude in Theravada Buddhism. Journal of the American Academy of Religion 71 (3): 579–604. Berryman, Sylvia. 2019. Aristotle on the Sources of the Ethical Life. Oxford: Oxford University Press.

412

A. Carpenter

Blum, Lawrence. 1994. Moral Perception and Particularity. Cambridge: Cambridge University Press. Bobzien, Susanne. 2014. Choice and Moral Responsibility (NE iii 1-5). In The Cambridge Companion to Aristotle’s Nicomachean Ethics, ed. Ronald Polansky, 81–109. New  York: Cambridge University Press. Brewer, Talbot. 2009. Retrieval of Ethics. Oxford: Oxford University Press. Burnyeat, Myles. 2000. “Plato on Why Mathematics is Good for the Soul”. In T. Smiley (ed.), Mathematics and Necessity: Essays in the History of Philosophy. pp. 1–81. Carone, Gabriela Roxana. 1997. “The Ethical Function of Astronomy in Plato’s Timaeus.” In T. Calvo & L. Brisson (eds.), Interpreting the Timaeus – Critias. Proceedings of the IV Symposium Platonicum. Selected papers. Sankt Augustin, Germany: Academia Verlag. Carpenter, Amber D. 2006. Hedonistic Persons: The Good Man Argument in Plato's Philebus. British Journal of the History of Philosophy 14 (1): 5–26. ———. 2014. Indian Buddhist Philosophy. London: Routledge (Acumen). ———. 2015a. Aiming at Happiness, Aiming at Ultimate Truth  – In Practice. In Moonpaths: Ethics and Emptiness., The Cowherds 21-42. New York: Oxford University Press. ———. 2015b. Ranking Knowledge in the Philebus. Phronesis 60: 180–205. ———. 2017a. Ethics without Justice: Eliminating the Roots of Resentment. In A Mirror Is for Reflection: Understanding Buddhist Ethics, ed. Jake H. Davis, 315–335. New York: Oxford University Press. ———. 2017b. Perfect Knowledge and its Affects. In Plato’s Philebus: Proceedings of the Ninth Symposium Platonicum Pragense, ed. Jakub Jirsa, Filip Karfik, and Štěpán Špinka, 152–166. Prague: OIKOYMENH. ———. 2017c. The Unhappiness of the Great King (Gorgias 461-81). In Rereading Ancient Philosophy: Old Chestnuts and Sacred Cows, ed. Verity Harte and Raphael Woolf, 60–79. Cambridge: Cambridge University Press. ———. 2018. Attention as a Means of Self-Dissolution and Reformation. Ratio 31 (4): 376–388. ———. 2019a. Reason and Knowledge on the Path: A Protreptic Reading of the Guide. In Readings of Śāntideva’s Guide to Bodhisattva Practice. In , ed. Jonathan C. Gold and Douglas S. Duckworth, 45–59. Columbia University Press. ———. 2019b. ‘and none of us deserving the cruelty or the grace’ – Buddhism and the Problem of Evil. In Engaging Buddhist Philosophy: A Cross-Cultural Approach to the Perennial Questions, ed. Steven Emmanuel. Columbia University Press. ———. 2020. ‘Pleasure as Temptress, Pleasure as Complex’, for Plato’s Pleasures. In, ed. J. Aufderheide and M. Erginel. ———. 2019. ‘Transformative Vision: Coming to See the Buddha’s Reality’ for Buddhist Literature as Philosophy, Buddhist Philosophy as Literature. In, ed. Rafal Stepien. SUNY Press., forthcoming. Chappell, Sophie Grace. 2014. Knowing What To Do: Imagination, Virtue, and Platonism in Ethics. Oxford: Oxford University Press. Clayton, Barbra R. 2006. Moral Theory in Śāntideva’s Moral Theory in Śāntideva’s Śikṣasamuccaya: Cultivating the fruits of virtue. New York: Routledge. ———. 2009. Śāntideva, Virtue, and Consequentialism. In Destroying Māra Forever: Buddhist Ethics Essays in Honor of Damien Keown, ed. John Powers and Charles S. Prebish, 15–29. Ithaca: Snow Lion Publications. Dancy, Jonathan. 1993. Moral Reasons. Oxford: Blackwell. ———. 2004. Ethics Without Principles. Oxford: Oxford University Press. Dreyfus, Georges. 1995. Meditation as Ethical Activity. Journal of Buddhist Ethics 2: 28–54. Foot, Philippa. 1958. Moral Arguments. Mind 67 (268): 502–513. ———. 1958-59. Moral Beliefs. Proceedings of the Aristotelian Society 59: 83–104. ———. 1969. Goodness and Choice. In The Is-Ought Question: A Collection of Papers on the Central Problem in Moral Philosophy, ed. W.D.  Hudson, 214–227. London: Palgrave Macmillan.

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

413

Frankena, William K. 1992. Sidgwick and the history of ethical dualism. In Essays on Henry Sidgwick, ed. Bart Schultz, 175–198. Cambridge: Cambridge University Press. Frauwallner, Erich. 1953. Geschichte der indischen Philosophie. Vol. 1, 217–225. Salzburg. Frede, Michael. 2011. In A Free Will: Origins of the Notion in Ancient Thought, ed. A.A. Long . Berkeley: University of California Press.Sather Classical Lectures, vol. 68 ———. 2014. The eph’ hêmin in ancient philosophy. In Studies on Causality and Responsibility in Ancient Philosophy, ed. Pierre Destrée, Ricardo Salles, and Marco Zingano, 351–364. Sankt Augustin: Academia Verlag. Gaita, Raimond. 2004. Good and Evil: An Absolute Conception. 2nd ed. New York: Routledge. Garfield, Jay L. 2010-11. What is it like to be a Bodhisattva? Moral Phenomenology in Śāntideva’s Bodhicaryāvatāra. Journal of the International Association of Buddhist Studies 33 (1–2): 333–357. Garfield, Jay. 2021. Buddhist Ethics: A Philosophical Exploration. New York: Oxford University Press. ———. 2015. Engaging Buddhism: Why It Matters to Philosophy. New  York: Oxford University Press. Gill, Christopher. 1985. Plato and the Education of Character. Archiv für Geschichte der Philosophie 67: 1–26. Gombrich, Richard. 1991. Buddhist Precept and Practice: Traditional Buddhism in the Rural Highlands of Ceylon. 2nd ed. Delhi: Motilal Banarsidass. Goodman, Charles. 2009. The Consequences of Compassion: An Interpretation and Defense of Buddhist Ethics. New York: Oxford University Press. Hallisey, Charles. 1996. Ethical Particularism in Theravāda Buddhism. Journal of Buddhist Ethics 3: 32–46. Hallisey, Charles, and Anne Hansen. 1996. Narrative, Sub-Ethics, and the Moral Life: Some Evidence from Theravāda Buddhism. The Journal of Religious Ethics 24 (2): 305–327. Hansen, Anne. 2002. Story & World: The Ethics of Moral Vision in the Gatilok of Uk Ñā Suttantaprījā Ind. Udaya: Journal of Khmer Studies 3: 45–64. Harvey, Peter. 1995. Selfless Mind: personality, consciousness and nirvana in early Buddhism. Surrey: Curzon Press. Harvey, Peter. 2000. An Introduction to Buddhist Ethics: Foundations, Values and Issues. Cambridge: Cambridge University Press. Heim, Maria. 2003. The Aesthetics of Excess. Journal of the American Academy of Religion 71 (3): 531–554. ———. 2014. The Forerunner of All Things: Buddhaghosa on Mind, Intention, and Agency. New York: Oxford University Press. Holland, R.F. 1980. Against Empiricism: On Education, Epistemology, and Value. Totowa: Barnes & Noble Books. Hursthouse, Rosalind. 1999. On Virtue Ethics. Oxford: Oxford University Press. Irwin, Terence. 1995. Plato's Ethics. New York: Oxford University Press. Janaway, Christopher. 1995. Images of Excellence. Oxford: Oxford University Press. Jenkins, Michelle. 2015. Early Education in Plato's Republic. British Journal for the History of Philosophy 23 (5): 843–863. Kamtekar, Rachana. 1998. Imperfect Virtue. Ancient Philosophy 18: 315–339. Kant, Immanuel. 1788/2011. Kritik der praktischen Vernunft. Otfried Höffe, ed. Berlin: Akademie Verlag. Keown, Damien. 1992. The Nature of Buddhist Ethics. London: Palgrave Macmillan. ———. 1996. Karma, Character, and Consequentialism. Journal of Religious Ethics 24: 329–350. Reprinted in 2005. Journal of Buddhist Ethics 12. Kraut, Richard. 2007. What Is Good and Why: The Ethics of Well-Being. Cambridge, MA: Harvard University Press. Lane, Melissa. 2005. Emploi pour philosophes: l’art politique et l’étranger dans le Politique à la lumière de Socrate et du philosophe dans le Théétète. In Études Philosophique 74/3: 325–45.

414

A. Carpenter

Lovibond, Sabina. 2002. Ethical Formation. Cambridge, MA: Harvard University Press. MacIntyre, Alasdair. 1981. After Virtue: A Study in Moral Theory. London: Gerald Duckworth & Co. Ltd. Mann, Wolfgang-Rainer. 2000. The Discovery of Things: Aristotle’s Categories and their Context. Princeton: Princeton University Press. Matilal, B.K. 2002. In Ethics and Epics: The Collected Essays of Bimal Krishna Matilal, ed. Jonardon Ganeri. New Delhi: Oxford University Press. McDermott, J.P. 1989. Animals and Humans in Early Buddhism. Indo-Iranian Journal 32: 269–281. McDowell, John. 1979. Virtue and Reason. The Monist 62 (3): 331–350. ———. 1995. Two Sorts of Naturalism. In Virtues and Reasons, ed. R. Hursthouse, G. Lawrence, and W. Quinn, 149–179. Oxford: Oxford University Press. ———. 1981. Non-Cognitivism and Rule-Following. In Wittgenstein: To Follow a Rule, ed. Holtzman and Leich, 141–162. London: Routledge. McNaughton, David. 1988. Moral Vision. Oxford: Blackwell. Meyer, Karin. 2018. False Friends: Dependent Origination and the Perils of Analogy in Cross-­ Cultural Philosophy. Journal of Buddhist Ethics. 25: 786–818 Murdoch, Iris. 1956. Vision and Choice in Morality. Proceedings of the Aristotelian Society Supplement 30: 32–58. ———. 1970. The Sovereignty of Good. London: Routledge & Kegan Paul. Ñāṇamoli, Bhikkhu, and tr. 1976. The Path of Purification: Visuddhimagga. Berkeley: Shambhala Publications xx. Ñāṇamoli and Bodhi, trs. Middle-Length Discourses of the Buddha (Majjhima Nikāya. Nussbaum, Martha. 1986. Fragility of Goodness. Cambridge: Cambridge University Press. Nussbaum, Martha. 1990. Love’s Knowledge: Essays on Philosophy and Literature. New York: Oxford UniversityPress Obeyesekere, Ranjini and Gananath Obeyesekere. 1990. ‘The Tale of the Demoness Kali: A Discourse on Evil’. History of Religions 29: 318–34. Obeyesekere, Ranjini. 1991. Jewels of the Doctrine. Albany, NY: State University of New York Press. Ohnuma, Reiko. 2017. Head, Eyes, Flesh and Blood: Giving Away the Body in Buddhist Literature. New York: Columbia. Pluhar, Werner S., and tr. 2002. Kant’s Critique of Practical Reason. Indianapolis, IN: Hackett Publishing. Pradhan, P., ed. 1967. Abhidharm-koshabhāṣya of Vasubandhu. Vol. VIII. Patna: K. P. Jayaswal Research Institute. Priestley, Leonard C.D.C. 1999. Pudgalavāda Buddhism: The Reality of the Indeterminate Self. Toronto: Centre for South Asian Studies, University of Toronto. Pruden, Leo M., and tr. 1991. Abhidharmakośabhāṣyam of Vasubandhu. Translated into French by Louis de La Vallée Poussin. 4 Vols. Berkeley, CA: Asian Humanities Press. Repetti, Riccardo. 2010. Earlier Buddhist Theories of Free Will: Compatibilism. Journal of Buddhist Ethics 17: 279–310. ———. 2012a. Buddhist Reductionism and Free Will: Paleo-compatibilism. Journal of Buddhist Ethics 19: 33–95. ———. 2012b. Buddhist Hard Determinism: No Self, No Free Will, No Responsibility. Journal of Buddhist Ethics 19: 130–197. ———. 2014. Recent Buddhist Theories of Free Will: Compatibilism, Incompatibilism, and Beyond. Journal of Buddhist Ethics 21: 279–352. ———, ed. 2017. Buddhist Perspectives on Free Will: Agentless Agency? Abingdon-on-Thames: Routledge. Sauvé-Meyer, Susan. 2011. Aristotle on Moral Responsibility: Character and Cause. 2nd ed. Oxford: Oxford University Press. (First published 1993).

Ideas and Ethical Formation: Confessions of a Buddhist-Platonist

415

———. 2014. Aristotle on what is up to us and what is contingent. In What is up to us? In Studies on Causality and Responsibility in Ancient Philosophy, ed. Pierre Destrée, Ricardo Salles, and Marco Zingano, 75–90. Sankt Augustin: Academia Verlag. Schmithausen, Lambert. 1973. Spirituelle Praxis und philosophische Theorie im Buddhismus. Zeitschrift für Missionswissenschaft und Religionswissenschaft 57: 161–186. Scott, Dominic. 1999. Platonic pessimism and moral education. Oxford Studies in Ancient Philosophy 17: 15–36. Sherman, Nancy. 1991. The Fabric of Character. Oxford: Clarendon Press. Siderits, Mark. 2003a. On the Soteriological significance of emptiness. Contemporary Buddhism 4, issue 1: 9–24. Reprinted in Siderits, Mark. 2016. Studies in Buddhist Philosophy, ed. Jan Westerhoff, 38–52. Oxford: Oxford University Press. ———. 2003b. Personal Identity and Buddhist Philosophy: Empty Persons. Aldershot: Ashgate Publishing. ———. 2007. Buddhism as Philosophy: An Introduction. Aldershot: Ashgate Publishing. ———. 2008. Paleo-Compatibilism and Buddhist Reductionism. Sophia 47 (1): 29–42. ———. 2011. ‘Buddhist Non-Self: The No-Owner’s Manual’. In The Oxford Handbook of the Self, edited by Shaun Gallagher. Oxford: Oxford University Press. ———. 2015. Does “Buddhist Ethics” Exist? In Moonpaths: Ethics and Emptiness. auth. The Cowherds, 119–140. New York, NY: Oxford University Press. ———. 2016. In Studies in Buddhist Philosophy, ed. Jan Westerhoff. Oxford: Oxford University Press. Smith, Nicholas D. 1999. Images, Education, and Paradox in Plato's Republic. Apeiron 32 (4): 125–141. Slote, Michael. 1997. Virtue Ethics. In Three Methods of Ethics, ed. Marcia Baron, Philip Pettit, and Michael Slote, 175–238. Oxford: Blackwell. Slote, Michael. 2001. Morals From Motives. New York: Oxford University Press. Stepien, Rafal. 2019. Buddhist Literature as Philosophy, Buddhist Philosophy as Literature. SUNY. Striker, Gisela. 1996. ‘Greek Ethics and Moral Theory’. Essays in Hellenistic Epistemology and Ethics. Cambridge University Press: 169–82. Thompson, Michael. 2008. Life and Action: Elementary Structures of Practice and Practical Thought. Cambridge, MA: Harvard University Press. Tillemans, Tom J.F. 2010-11. Madhyamaka Buddhist ethics. Journal of the International Association of Buddhist Studies 33 (1–2): 353–372. Vlastos, Gregory. 1981. The Individual as Object of Love in Plato. In Platonic Studies, 2nd ed., 3–42. Princeton, NJ: Princeton University Press. ———. 1999. Happiness and Virtue in Socrates’ Moral Theory. In Plato 2: Ethics, Politics, Religion and the Soul, ed. Gail Fine, 105–136. Oxford: Oxford University Press. Vogt, Katja Maria. 2017. Desiring the Good: Ancient Proposals and Contemporary Theory. New York, NY: Oxford University Press. Walsh, tr. Long Discourses of the Buddha. Westerhoff. 2016. Introduction. In Studies in Buddhist Philosophy, ed. Mark Siderits and Jan Westerhoff, 1–12. Oxford: Oxford University Press. White, Nicholas. 2015. Plato and the Ethics of Virtue. In The Routledge Companion to Virtue Ethics, 3–15. New York: Besser-Jones and Slote, Routledge. Winch, Peter. 1972. Moral Integrity. In Ethics and Action, 171–192. London: Routledge & Kegan Paul. Woodward, F.L., and tr. 1961. In Manual of a Mystic, Being A Translation from the Pali and Sinhalese Work Entitled The Yogāvachara’s Manual, ed. Rhys Davids. London: Humphrey Milford. Wright, Dale S. 2005. Critical Questions Towards a Naturalized Concept of Karma in Buddhism. Journal of Buddhist Ethics 12: 78–93.

Self-Cultivation Philosophy as Fusion Philosophy: An Interpretation of Buddhist Moral Thought Christopher W. Gowans

It is often observed that there is little or no moral philosophy in classical Indian Buddhist thought.1 This is sometimes believed to be surprising since obviously there is an ethical teaching in Buddhism and clearly there are other forms of Buddhist philosophy. In my view, there is something that can plausibly be called moral philosophy in Indian Buddhism, but it is not quite what many people have expected because they have approached the issue from a specific understanding of philosophy that is too limited. I propose to defend this interpretation by reshaping Mark Siderits’ concept of fusion philosophy. The fusion I will propose reveals a form of moral philosophy that may be called self-cultivation philosophy. I will argue that this was a common activity in ancient Chinese, Greek and Indian cultures—including especially early Buddhist thought and practice. Siderits introduced fusion philosophy as a successor to comparative philosophy and as something similar to fusion music.2 Just as in fusion music features of one tradition were employed to solve problems that arose in another tradition, he said, so in fusion philosophy elements of one tradition may be used to solve problems in another tradition. His primary example was using traditional Buddhist philosophy to defend Parfit’s reductionism in Reasons and Persons.3 In recent years, fusion philosophy so understood has generated impressive results by utilizing resources from classical Indian philosophy to solve problems in contemporary philosophy and

 For example, see Keown (2017). I discussed this issue in Gowans (2015, esp. pp.  54–7 and 161–5), but my perspective here is rather different. 2  See Siderits (2003, pp. xi–xii). 3  See Parfit (1984). 1

C. W. Gowans (*) Fordham University, Department of Philosophy, New York, NY, USA e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_21

417

418

C. W. Gowans

vice versa.4 Still, there are other ways of thinking about philosophy, and I suggest here moving fusion philosophy in a somewhat different direction. The comparison with fusion music is instructive. But I have a different take than Siderits on what fusion music is and this yields a different understanding of what fusion philosophy might be. To revert to my youth, when Santana appeared at the Fillmore West in 1968 they were playing an electric blues-driven rock similar in some ways to many other bands in that venue. But there was a striking difference. There were congas on the stage and much of the music had an unmistakable Latin feel. Carlos Santana was drawing on blues, rock, jazz, African, Latin and other genres and traditions to create something new, unique and exciting (listen to Jingo or Soul Sacrifice), a sound still distinctive to this day, the beginning of a long career in fusion music.5 In my view, what stands out about fusion music—and there are numerous other examples—is that it simultaneously draws on and challenges standard musical genres to generate new perspectives on what music can be. Fusion music on this understanding is the expression of an innovative and subverting attitude, an interrogation of accepted genres and a creation of new sounds that elude those conventional categories— which is why it has often been controversial.6 Could there be a fusion philosophy that does something similar? I believe there can and should be. As an expression of an attitude, fusion philosophy in this sense could be many things. Here I propose just one direction it could take. By bringing several ancient philosophical traditions into our view we can gain a perspective on understanding them, an interpretive framework, that is more evident and fruitful than if we had restricted our attention to only one of these traditions or examined them primarily from a single point of view such as the outlook of contemporary philosophy. This framework is what I call self-cultivation philosophy. What is distinctive in this approach—and here I am drawing on the work of many others—is a clearer realization of what ancient philosophies, including Buddhist philosophies, often were and of what philosophy, in one of its many forms, perhaps might be again.

1 Chinese Philosophy: Xunzi Much of what is considered comparative or fusion philosophy written in English basically begins with Western philosophy as understood in most contemporary graduate philosophy programs. This creates an expectation of what philosophy is supposed to look like and thus shapes what is allowed to be philosophy. As an initial move in challenging conventional genres, I propose to begin elsewhere, with

 For a more recent defense of fusion philosophy, using confluence philosophy as a synonym, and an overview of some work from this approach, see Siderits (2017). 5  For his own account, see Santana (2014). 6  For a similar interpretation of fusion music, see Fellezs (2011). 4

Self-Cultivation Philosophy as Fusion Philosophy: An Interpretation of Buddhist Moral…

419

Warring States Chinese philosophy, and the specific example of the Confucian philosopher Xunzi. This is controversial, of course, because it is sometimes claimed that there was little or no philosophy in ancient China. I will consider this objection in due course. For now, I want to highlight the reason for starting in this way. It is often stated that the ancient Chinese thinkers had a practical rather than a theoretical orientation, that their primary concern was “how to become good.”7 More specifically, it is commonly maintained that many of these philosophers, including both Confucians and Daoists, emphasized self-cultivation practices.8 There is a broader context for these practices that is worth noting. In ancient China, there was considerable interest in a variety of self-cultivation programs. Some of these emphasized physical aspects of life such as health and longevity, and so had more kinship with medicine, while others stressed psychological and moral themes, and hence had more connection with philosophy. But these tendencies were sometimes related to one another in conceptions of embodied virtue. In all these cases, self-cultivation regimes typically focused on the regulation of qi (a vital energy with physiological as well as psychological and moral aspects) and the balance of the complementary movements of yin and yang. This is an indication of the importance, in ancient Chinese culture, of both the body and the significant relationships between medicine and philosophy.9 We will see that these relationships have considerable significance in the analysis to come. Xunzi was a complex third century B.C.E. philosopher who developed and defended the Confucian tradition in response to critics both outside the tradition and within it. His thought is represented in the text Xunzi, a collection of essays (chapters) that he probably composed in part, but which were subsequently edited by others.10 An important feature of the Xunzi is that the first two essays are entitled “An Exhortation to Learning” and “Cultivating Oneself.” These signal that the book is substantially a manual in self-cultivation philosophy.11 According to Xunzi, human beings are born with a set of self-centered desires rooted in our physical nature, desires for such things as life, food, warmth, rest, sensory pleasures, etc. In a world of limited resources, he thought, if people lived solely on the basis of these desires, they would be perpetually dissatisfied and in conflict with one another. Violence and chaos would be the result. But Xunzi believed that the ancient sage kings discovered that people’s desires can be transformed through the development of Confucian virtues such as righteousness (yi) and ritual propriety (li). Life in accord with these virtues is peaceful and orderly, yet  Slingerland (2003, p. 3). Cf. Graham (1989, p. 3).  For example, see Allinson (1989), Ames (1985), Ivanhoe (2000) and Slingerland (2003). Of the traditions I discuss in this paper, classical Chinese philosophy is the place where the phrase “selfcultivation philosophy” has some currency. 9  On these points, see Raphals (2011) and Fall (2017), and Wang (2010) and (2012, ch 5). 10  See Hutton (2014). For a helpful collection of contemporary discussions of Xunzi, see Hutton (2016). 11  For interpretations of Xunzi that emphasize self-cultivation, see Ivanhoe (2000, ch 3), Kline III (2006), Schofer (2000), Slingerland (2003, ch 7) and Stalnaker (2006) and (2016). 7

8

420

C. W. Gowans

still permits moderate satisfaction of our natural desires. Hence people have reason to acquire these virtues, and the ideal figures of the junzi (gentleman or superior person) and the sage have acquired them to an exceptional extent. Xunzi seemed to think all or nearly all human beings could reach these ideals, but he also believed this was very difficult to do. This is because he rejected Mencius’ view that we are born with “moral sprouts” that can be cultivated to mature into the Confucian virtues. In contrast to Mencius’ claim that “human nature is good” because we have these sprouts, Xunzi said that “human nature is bad” because we do not (though both slogans are simplifications of more complex views).12 Nonetheless, Xunzi thought we could acquire the virtues through deliberate effort, though this would require long, hard work. He captured this point in two striking images: Thus, crooked wood must await steaming and straightening on the shaping frame, and only then does it become straight. Blunt metal must await honing and grinding, and only then does it become sharp. Now since people’s nature is bad, they must await teachers and proper models, and only then do they become correct. They must obtain ritual and yi, and only then do they become well-ordered.13

Xunzi’s self-cultivation exercises tell us how to do this. With the guidance of teachers and models, we need to develop the capacities our heart-mind (xin) to be empty, single-minded and still. The formation of these epistemic virtues enables us to understand—impartially, correctly and fully—the way (dao) taught by the ancient wise kings. Though we lack moral sprouts, we do have a natural “ability to know.”14 We do this by studying and reflecting on classic texts such as the Odes and Documents. In addition, since by nature the dispositions for our original desires are deeply rooted, we need to practice living in accord with the Confucian virtues. Through repetition, we can change our habits and develop the know-how needed for a virtuous life. In this way, the junzi “makes his eyes not want to see what is not right” and so “comes to the point where he loves” what he has learned.15 These activities bring about a complete transformation of the ways in which a person perceives, thinks, feels, desires, chooses and acts. In a fascinating chapter, Xunzi says that even music has a role to play in this process because the joy of music is an important part of a virtuous life and proper music has the power “to move the goodness in people’s hearts.”16

 See Ch. 23 of Xunzi for the explanation of Xunzi’s view.  Hutton (2014, p. 248 (23.19–24)). In references to historical texts, where ordinary pagination is supplemented by another system, this is given in parentheses. 14  Hutton (2014, p. 233 (21.363)). For Xunzi’s epistemology, see Ch. 21 of Xunzi as a whole. 15  Hutton (2014, p. 8 (1.221–25)). 16  Hutton (2014, p. 218 (20.13)). In an illustration of the earlier point about the broad concept of self-cultivation, Xunzi goes on to say that proper music can cause “perverse and corrupt qi to have no place to attach itself” to us (p. 218 (20.13–14)). 12 13

Self-Cultivation Philosophy as Fusion Philosophy: An Interpretation of Buddhist Moral…

421

2 The Idea of Self-Cultivation Philosophy There are many features of Xunzi’s outlook, such as his extraordinary emphasis on the importance of tradition and ritual, that may well appear foreign to us and to philosophy.17 Nonetheless, I have briefly sketched his position because it is an outstanding example of a self-cultivation philosophy. We can use this to start developing a general model of self-cultivation philosophy and then see how this model might be helpful elsewhere. In my view, though the context, forms and expressions of self-cultivation philosophies vary greatly, there is an underlying structure that they share and which we can see when we bring different traditions together in the right way. In brief, self-cultivation philosophies propose that we can move from a troubled state of existence to some ideal state of being by undertaking an ensemble of self-cultivation exercises guided by, and partly constituted by, some philosophical analysis. In addition, in making these claims, they typically assert or assume an account of human nature or psychology.18 There could be a self-cultivation program that does not pertain to philosophy, but what is distinctive about self-cultivation philosophies is that they involve philosophy in some significant way: they are guided by some philosophical account of the good life, and a crucial part of the self-cultivation exercises is understanding this account and transforming one’s life in accordance with it. But philosophy in what sense? In fusion music and fusion philosophy both, it is not the case that anything goes (apologies to Cole Porter). Though philosophy is a contested concept to some extent, in my judgment, it is a critical reflective practice that seeks understanding of fundamental assumptions we make in our lives. Since we make many different kinds of assumptions, there are many different kinds of philosophy. The self-­ cultivation philosophies address a variety of different issues, but the focal point of their concern is what Bernard Williams called Socrates’ question: “how should one live?”19 There are many philosophical discussions that do not pertain to this question, and there are many philosophies that are not self-cultivation philosophies. However, for self-cultivation philosophies, this question is central and is commonly an organizing principle for other philosophical concerns such as metaphysics and epistemology (as can be seen in the case of Xunzi). There are two objections that may be raised at this juncture. First, it might be said that philosophy is a theoretical, not a practical, discipline: its goal is knowledge for its own sake irrespective of the practical benefits of this knowledge. Second, it might be claimed that the proper method of philosophy is reason or argument, and  However, it is worth noting the challenge to this assessment in the interpretation of Confucius in Fingarette (1972). 18  In recent years, the expression “philosophy as a way of life” has been employed increasingly to refer to many of the phenomena I call self-cultivation philosophies. However, the phrase “way of life” is not the best for my purpose. A philosophy could be a way of life without having the specific structure of self-cultivation philosophies I outline here. 19  Williams (1985, p. 4). See the whole of Williams’ first chapter for an account of the importance of this question. 17

422

C. W. Gowans

the purported self-cultivation philosophies often do not involve reason or argument. Both contentions may also be deployed in objections that Warring States Chinese thought is not philosophy.20 In response to the first point, though some kinds of philosophy may have little or no practical benefit, we should not accept “theoretical, not practical” as a general characterization of the concept of philosophy itself. It would be very strange to suppose that reflection on the question “how should one live?” would no longer be philosophical as soon as we thought that our answer had some practical benefits in our lives. Moreover, some unquestionably canonical philosophical texts in the Western tradition have a practical aim. For example, Aristotle says in the Nicomachean Ethics that “the present undertaking is not for the sake of theory, as our others are (for we are not inquiring into what excellence (aretē) is for the sake of knowing it, but for the sake of becoming good, since otherwise there would be no benefit in it at all).”21 Self-cultivation philosophies commonly have some theoretical concerns, but they have an overall practical aim and this does not bar them from being philosophies. Similarly, the fact that classical Chinese thinkers such as Xunzi commonly had a practical aim is not a reason to think they were not philosophers. Another way to characterize self-cultivation philosophies in this respect is to say (in ancient terms) that they emphasize practical wisdom at least as much as theoretical wisdom or (in contemporary terms) that they stress “knowing how” as much as “knowing that.” This is implied by the comparison with medicine that, we will see, is significant in all the self-cultivation philosophies we will consider here. We could put this point by saying that these philosophies are concerned with transformative knowledge, that is, knowledge that actually transforms the way we live our lives by changing the problematic habits we start with into the modes of life of the ideal state of being.22 The importance of this is acknowledged by Aristotle himself in his discussion of habituation in the Nicomachean Ethics. The self-cultivation exercises aim at such transformative knowledge. With regard to the second objection, an obvious initial response is that self-­ cultivation philosophers such as Xunzi do offer reasons and arguments for their positions. However, more needs to be said because the central claim of the objection is often taken to be that rational argument is the one and only method of philosophy. We need to examine this familiar claim more carefully. If the aim of philosophy is to understand fundamental assumptions, as I have proposed, then philosophy should employ whatever cognitive capacities human beings have that are needed for this understanding. Reason is certainly an important candidate for these capacities, and it was emphasized by the ancient Greek philosophers. But it is not the only candidate, and in Western philosophy (as well as in other philosophical traditions) there

 For a defense of Classical Chinese thought as philosophy, see Van Norden (2017).  Aristotle (2002, p. 112) (II.2 at the beginning), emphasis added. 22  This is similar to, though not quite the same as, the idea of transformative experience in Paul (2014). 20 21

Self-Cultivation Philosophy as Fusion Philosophy: An Interpretation of Buddhist Moral…

423

has been considerable debate about the nature and value of reason and other cognitive capacities. For example, in recent years attention has been drawn to the different “styles of reasoning” employed in philosophy.23 In addition, as is well known, both Aristotle and Mill (each an author of a work in logic) noted the limitations of the kinds of precision and proof that are appropriate to moral philosophy.24 Moreover, from the ancient Academic and Pyrrhonian skeptics to the present there have been skeptical challenges to the value of reason: even Kant, in some ways very much a rationalist, emphasized the limitations of reason. Finally, it is widely acknowledged—explicitly or implicitly—that philosophy requires more than rational inference. For instance, Descartes says that philosophy needs to take into account what the “light of nature” teaches us.25 More generally, there are frequent appeals in philosophy to various forms of experience or awareness as well as such things as common sense, intuitions, appearances (Aristotle’s phainomena), introspection, instinct, conscience, innate ideas, testimony, meditation and the like. All this suggests that, in philosophy as a whole, a variety of cognitive capacities have been appealed to, and it is a philosophical question (not a defining feature of the discipline) which of these is appropriate. This should not surprise us: in the form of contemporary virtue epistemology called virtue reliabilism (anticipated in India much earlier by Nyāya), it is commonly claimed that there is a plurality of cognitive capacities that are truth-­ conducive such as perception, reason, introspection, intuition, memory, testimony and the like. It is natural to suppose that philosophy, in pursuit of understanding, will draw on a variety of cognitive capacities. In view of this, there is no reason to think that Xunxi and other self-cultivation philosophers fail to be genuine philosophers on account of their diverse methods of inquiry. Of course, insofar as self-cultivation philosophies aim at transformative knowledge, they will employ techniques in addition to the exercise of cognitive capacities. Knowing how to live well is a kind of practice and it requires more than intellectual understanding (just as, by comparison, knowing how to play the viola well is a practice that requires more than understanding musical notation or music theory). This is why Aristotle stressed practice in the form of habituation as crucial to attaining a life of virtue. Practice needs to be guided by philosophical understanding, but this understanding by itself is usually not sufficient for transformation. Hence, reason and other cognitive capacities are necessary but not sufficient for self-cultivation philosophies. With this in mind, let us now turn to the structure of self-cultivation philosophies. Though they are expressed in quite different styles of writing, I believe there is a common form that unites them and is exemplified in Xunzi. This form has four parts. First, there is an account of the typical default condition of human life, an analysis of the ways in which our lives are usually problematic in the absence of the

 See Ganeri (2019, p. 16).  See Aristotle (2002, p. 96 (I.3)) and Mill (2002, pp. 6–7 (Ch. 1, par. 5)). 25  Descartes (1993, pp. 26–7 (VI, 38–9)). 23 24

424

C. W. Gowans

proposed self-cultivation program. This may be called our existential starting point. For Xunzi, the pursuit of our natural self-centered desires gives rise to personal dissatisfaction and violent social conflict. More broadly, I would suggest, self-­ cultivation philosophies claim that, though at times our lives may be more or less happy, there are some pervasive features of the human condition that give rise to anxiety, fear, alienation, anger, greed, envy and the like. Different explanations may be given for these troublesome states, but there are recurrent themes. The explanation usually centers on disruptive desires and emotions that are rooted in some kind of ignorance about human nature and the world. Second, an ideal state of being is put forward as something that human beings are capable of achieving and that would alleviate the problematic features of our existential starting point. For Xunzi, this is the junzi and the sage. These figures are said to live in accord with Confucian moral virtues such as righteousness, ritual propriety, benevolence, filial piety, etc. and to have practical wisdom, equilibrium and joy (to a great extent in the case of the junzi and perfectly in the case of the sage). In general, in self-cultivation philosophies, the ideal state of being is sometimes described in rather brief and elusive terms, and it is sometimes suggested that it is quite difficult to attain. Once again, there is considerable diversity in the accounts of the ideal, but there are characteristic features. The ideal is usually said to include some kind of wisdom, to incorporate a psychological state of contentment (consisting of tranquility as well as joy), and to involve moral virtue. Though the ideal is commonly thought to be valuable in itself, the virtue dimension usually means that the ideal person benefits others as well as himself. Third, a set of practices are advocated as the means by which people can effect the transition from the existential starting point to the ideal state of being. These may involve a variety of philosophical, psychological, moral, therapeutic or spiritual exercises. For Xunzi, what is central is training the heart-mind so that it can understand the way of the sage kings, reflecting philosophically on classic texts that explain this way, especially the importance of the Confucian virtues, and practicing the rituals (sometimes including music) so as to achieve the emotional and behavioral dispositions of a virtuous life. Overall, in self-cultivation philosophies, there is quite a bit of diversity in the nature of these practices, but they usually involve a multifaceted set of exercises that are thought to be needed to achieve the transformative knowledge that issues in the ideal. These practices commonly involve the development of philosophical understanding as well as such things as moral training, forms of meditation, bodily practices and other kinds of self-discipline. It is often believed that these will require considerable time and effort to achieve the ideal (though there are disagreements about how much). Usually it is proposed that the exercises should be performed in a social context under the guidance of some master. Finally, there is presupposed in the first three features of self-cultivation philosophies some understanding of human nature or psychology (and sometimes a broader metaphysical account of the world we live in). This explains why human lives are commonly unsatisfactory (the starting point), what we are capable of so as to achieve a better life and why it would be better (the ideal), and how we could go

Self-Cultivation Philosophy as Fusion Philosophy: An Interpretation of Buddhist Moral…

425

about achieving this life (the practices). Of course, there are quite different accounts of this. In the early Confucian tradition, there was a basic division between Mencius’ view that we are born with moral sprouts that we can nurture to achieve the virtues and Xunzi’s view that we have no such sprouts but nonetheless have the ability to develop knowledge of the virtues and to transform our lives to be in accord with them. What they share in common is the belief that some account of human nature is important. Other self-cultivation philosophies have this belief as well though the specifics of their accounts vary widely.

3 Greek Philosophy: Epicurus If we start with the Chinese Warring States period and figures such as Xunzi, it is natural to see self-cultivation philosophies as a central form of philosophical activity. Of course, it might be thought that this is a distinctive and perhaps unusual feature of Chinese culture. However, as I have suggested, when we bring this perspective to other philosophical traditions we may find self-cultivation philosophies as well. Let us now turn to ancient Greek philosophy to see some ways in which this is true. I have already appealed to Aristotle in support of my understanding of philosophy as something that may have a practical aim and may employ diverse forms of cognition. However, it must be admitted that much of Aristotle’s philosophy emphasizes theory rather than practice (in fact, for Aristotle the best way of life is reflection), and that in this respect there is a contrast between classical Chinese philosophy and Aristotle’s philosophy (and indeed much of Greek philosophy). Those who comment on the practical orientation of Chinese philosophy often have such a contrast in mind, seeing Greek thought overall as comparatively more focused on theoretical concerns.26 My point here is that, though there may well be such contrasts on the whole, nonetheless there are self-cultivation philosophies in ancient Greek culture that are quite similar in important respects to those in Warring States Chinese culture. Here I leave aside to what extent Aristotle, and also Socrates and Plato, might be understood in terms of self-cultivation philosophy. It is in the Hellenistic period, and especially the Stoic and Epicurean philosophies, where this is most clearly true. My example will be Epicurus. Though in so many ways Xunzi and Epicurus are worlds apart, it is striking to what extent their philosophical projects converge. Epicurus is probably best known as a hedonist, but he has an unusual form of hedonism, and his overall philosophy conforms to the basic features of self-cultivation philosophy outlined in the last

26

 For example, see Lloyd (1990, p. 124).

426

C. W. Gowans

section.27 Epicurus lived in the fourth and third centuries B.C.E. (quite close to the time of Xunzi). Diogenes Laertius reports that he wrote numerous works, but these are almost entirely lost (only fragments remain). However, two of the works preserved by Diogenes, Letter to Menoeceus and Principal Doctrines, are valuable summaries of his philosophy.28 For our purpose, the place to begin is a passage in which Epicurus compares philosophy and medicine: Empty is the argument of the philosopher by which no human disease is healed; for just as there is no benefit in medicine if it does not drive out bodily diseases, so there is no benefit in philosophy if it does not drive out the disease of the soul.29

As Martha Nussbaum has pointed out, comparisons of philosophy and medicine were common among the ancient Greek philosophers.30 This is a clear sign of the importance of self-cultivation philosophy in the early stages of Western philosophy. Epicurus apparently developed an extensive metaphysics (a form of atomism) and epistemology (a kind of empiricism). But these theories were put in the service of enabling us to attain a good life. At the beginning of Menoeceus, Epicurus said that we should study and practice philosophy throughout our life in order to secure happiness and “the health of the soul.”31 He concluded the text with these words: “Practice these and the related precepts day and night, by yourself and with a like-­ minded friend, and you will never be disturbed either when awake or in sleep, and you will live as a god among men.”32 Menoeceus is clearly a manual of self-­ cultivation philosophy. For Epicurus, the existential starting point is that people’s lives are problematic because they are commonly dominated by fears, especially of death and the gods, and frustrated desires. “For a man is unhappy either because of fear or because of unlimited and groundless desire.”33 According to Epicurus, for the most part, the unfulfilled desires that frustrate us are the products of society and are neither natural nor necessary for happiness. Both the concern about our mortality and the dissatisfaction caused by desires are common features of the existential starting point in self-cultivation philosophies (both have parallels in Xunzi). As a hedonist, for Epicurus the ideal state of being is a life of pleasure and the absence of pain. But this is misleading because he thought the most valuable kind of pleasure is what he called katastematic (static) pleasure: being without some

 For interpretations of Epicurus or ancient Epicureanism that in various ways emphasize these features, though without stressing the phrase self-cultivation philosophy, see Cooper (2012, pp. 226–76), Hadot (2002, pp. 113–26), McEvilley (2002, pp. 611–21), Nussbaum (1994, ch. 4), Sorabji (2000) and Tsouna (2009). 28  See Inwood and Gerson (1997, pp. 28–36). 29  Inwood and Gerson (1997, p. 97) (a fragment quoted by Porphyry). 30  See Nussbaum (1994, ch. 1). 31  Inwood and Gerson (1997, p. 28 (122)). 32  Inwood and Gerson (1997, p. 31 (135)). 33  Inwood and Gerson (1997, p. 100) (a fragment quoted by Porphyry). 27

Self-Cultivation Philosophy as Fusion Philosophy: An Interpretation of Buddhist Moral…

427

need such as the need for water. Hence, the katastematic pleasures are “freedom from disturbance and freedom from suffering.”34 For this reason, in Menoeceus Epicurus says that “when we say that pleasure is the goal we do not mean the pleasures of the profligate or the pleasures of consumption...but rather the lack of pain in the body and disturbance in the soul.”35 In light of this, it has been suggested that Epicurus was more of a “tranquillist” than a hedonist.36 However, he also thought that such a life required friendship and the virtues commonly emphasized in Greek philosophy: justice, courage, temperance and practical wisdom. Epicurus had much to say about the practices that were needed to move people from the existential starting point to the ideal state of being.37 Central among these are memorizing, reflecting on and understanding Epicurean philosophy. Epicurus believed these allow us to see that there is no reason to fear death or the gods (in light of his atomism) and that the only important desires are the mostly easy-to-­ fulfill natural and necessary ones.38 But we also need to modify our habits so as to become “accustomed to simple, not extravagant, ways of life.”39 In addition, people need to learn how to appreciate pleasures in the present moment as well as how to find it in memories of the past and anticipations of the future. Moreover, in the Epicurean tradition, considerable attention was also given to moral improvement through following proper models, examination of conscience, confession, moral correction and the like. Finally, Epicurus based his self-cultivation program on an analysis of human nature and the nature of the world as a whole. As just noted, he thought his atomism undermined the fear of death (because there is nothing to fear since no part of us survives death) and of the gods (less obviously, because they have no concern with us). His empirical account of human nature also purported to explain why pleasure is the good, why our actual desires are often groundless and frustrate us, why we can avoid this by restricting ourselves to natural and necessary desires, and finally why we can undertake the self-cultivation program because we are autonomous agents.

4 Indian Philosophy: Buddhism When it is said that we do not find moral philosophy in Indian Buddhist thought what is usually meant is that we do not find explicit discussions of central topics in Western moral philosophy as it developed from its origins in ancient Greece up  Inwood and Gerson (1997, p. 43) (a fragment quoted by Diogenes).  Inwood and Gerson (1997, pp. 30–31 (131)). 36  See O’Keefe (2010, pp. 107 and 120). 37  See Tsouna (2009) for a nice summary of these in the Epicurean tradition. 38  It has been suggested that Epicurus did not encourage his followers to engage in philosophy themselves (see Nussbaum 1994, pp.  129–39 and Cooper 2012, pp.  274–6). For a compelling response to this objection, see Tsouna (2009, 263–5). 39  Inwood and Gerson (1997, p. 30 (131)). 34 35

428

C. W. Gowans

through contemporary debates. For example, we do not find systematic accounts of issues such as why we should be moral, the meaning and justification of basic moral principles, or whether morality is best understood in terms of consequentialism, virtue ethics or some other normative theory. However, if we pose the question raised by the perspective I have developed here—“Is there a self-cultivation philosophy in Indian Buddhist thought?”—the answer is clearly “yes”: there is moral philosophy in this sense. In fact, there are self-cultivation philosophies in several traditions of Indian philosophy such as yoga as articulated in Patañjali’s Yoga Sutra.40 But my focus here will be on Buddhism. Let us begin with the well-known text “The Shorter Discourse to Mālunkyāputta” (Cūḷamālunka Sutta) in the Majjhima Nikāya.41 Mālunkyāputta asks the Buddha why he does not state a position on ten “speculative views” (for example, about whether the world is eternal, infinite, etc.). The Buddha responds by giving a simile to explain why he does not discuss such views. Suppose a man wounded by a poison arrow was brought to a doctor by his friends and the man said he would not let the doctor treat him until he learned several facts about the attack such as the name and class of the person who shot the arrow, his complexion, where he lived, etc. The Buddha observed that the man would die if he insisted on knowing the answers to these questions before allowing the doctor to treat him. The point is that the man does not need the answers to these questions in order for the doctor to heal his wound. Likewise, the Buddha said, Mālunkyāputta does not need an account of the speculative views in order to benefit from the Buddha’s teaching. The Buddha explained he did not address these views because doing so would not have been beneficial to attaining enlightenment and overcoming suffering. Rather, he taught the Four Noble Truths because these are beneficial to attaining enlightenment and overcoming suffering. For our purpose, what is significant about this text is that the Buddha frames his basic teaching in terms of a comparison with medicine: just as the purpose of medicine is to heal the body so the purpose of his teaching is to enable us to overcome suffering. This is a clear indication—strikingly similar to what we have seen in China and Greece, especially the medical analogy passage in Epicurus—that the Buddha is putting forth a self-cultivation philosophy. The aim of his philosophy is to allow us to overcome suffering and he is concerned with philosophical positions only insofar as they contribute to this end. “Rights views” are important, but only insofar as they bear on this practical perspective.42 Moreover, there are other indications of the comparison with medicine. In the Majjhima Nikāya, enlightenment (Nibbāna) is sometimes depicted as a state of health, and the Buddha is portrayed as a doctor.43 In addition, much later, in the Visuddhimagga Buddhaghosa uses a

 See Soni (2010).  Ñāṇamoli and Bodhi (1995, pp. 533–36 (I 426–32)). 42  See the “Sammādiṭṭhi Sutta” in Ñāṇamoli and Bodhi (1995, pp. 132–44 (I 46–55)) (leaving aside complications raised by “no view” passages that appear in the Sutta Nipāta). 43  See Ñāṇamoli and Bodhi (1995, pp. 614–16 (I 510–12) and 867 (II 260)). 40 41

Self-Cultivation Philosophy as Fusion Philosophy: An Interpretation of Buddhist Moral…

429

medical simile to explain the Four Noble Truths: “the truth of suffering is like a disease; the truth of the origin is like the cause of the disease, the truth of cessation is like the cure of the disease, and the truth of the path is like the medicine.”44 The suggestion that the Buddha’s teaching is similar to medicine means that it is concerned with “knowing how” as much as with “knowing that,” with the transformative knowledge that enables us to change our lives so as to overcome suffering. In view of this, it is not surprising that the teaching of the Buddha is naturally understood according to the four-part structure of self-cultivation philosophies explained earlier. For the Buddha, the existential starting point is that suffering or unsatisfactoriness (dukkha) is a pervasive feature of our lives, as explained in the first two Noble Truths.45 In the First Noble Truth, suffering is characterized in a variety of ways: it concerns displeasure, frustrated desires, aging, illness, death, etc. But the basic problem, in my view, is not so much what happens to us as our attitude towards these phenomena. This is brought out in the Second Noble Truth that tells us that the origin of suffering is craving (taṇhā). Just as in Xunzi and Epicurus, desire is central to what is problematic in life, but the Buddha’s analysis of this is very different. In his broader teaching, the origin of craving is a specific delusion, the mistaken belief that we are selves. Hence, the “three unwholesome roots” are greed, hatred and delusion (the first are two contrary forms of craving, and the last centrally pertains to the self).46 This analysis tells us that what is problematic in our lives is the possessive attitude (attachment) that comes with the belief that we are selves, an attitude in which we interpret our experiences from the perspective of “I,” “me” and “mine.” This possessive orientation is the source of the craving that renders our lives unsatisfactory. The ideal state of being is Nibbāna (nirvana) as explained in the Third Noble Truth. This is the cessation of suffering brought about by the cessation of the belief that we are selves and hence the cessation of craving and the possessive attitude. Elsewhere in the Buddha’s teaching more positive characterizations of the enlightened person are suggested: he or she has attained wisdom, bliss, tranquility and virtue. Much of this is captured by the importance assigned to the four “divine abodes” (brahmavihāra): loving-kindness, compassion, appreciative joy and equanimity. The practices by which we move from suffering to the ideal state of being are explained in the Eightfold Path that make up the Fourth Noble Truth. The eight steps of the path are sometimes classified in three categories: virtue (right speech, action and livelihood), concentration (right effort, mindfulness and concentration) and

 Ñāṇamoli (1999, p.  520 (XVI 87)). I explain the comparison of medicine and philosophy in Buddhist and Hellenistic thought in more detail in Gowans (2010). 45  For example, as in standard formulations of the Four Noble Truths such as Bodhi (2000, vol. 2, pp. 1843–47 (V 420–24)) and Ñāṇamoli and Bodhi (1995, pp. 1097–1101 (III 248–52)). 46  For example, see Bodhi (2012, pp. 291–94 (I 201–205)). 44

430

C. W. Gowans

wisdom (right view and right intention).47 What is noteworthy here is that many kinds of training are thought to be required to achieve the transformative knowledge that constitutes enlightenment, that these include meditative disciplines (mindfulness and concentration), and that the different trainings are regarded as mutually supportive. For example, right view and wisdom are said to be “assisted by virtue, learning, discussion, serenity, and insight” (the last two are correlated with concentration and mindfulness).48 And we are told that “wisdom is purified by morality, and morality is purified by wisdom.”49 This suggests that wisdom, a full understanding of the Buddha’s teaching, requires both moral formation and meditative discipline. It is evident that philosophical understanding is an important part of the practices required to overcome suffering. These practices include grasping the philosophical basis of the overall teaching as well as arguments the Buddha gave in support of the no-self doctrine.50 However, it is also clear that intellectually accepting such arguments is not sufficient for the transformational knowledge we need to overcome suffering. Though “reflective acceptance” of the Buddha’s teaching is important, this teaching is also “unattainable by mere reasoning.”51 This is because the possessive attitude that gives rise to suffering, though it presupposes a false belief about self (and related doctrines such as impermanence and dependent origination), is also deeply rooted in our habitual ways of perceiving, thinking, feeling, choosing, acting and the like. Hence, only a broad array of practices, including both moral development and meditative exercises, can transform these habits and bring about a non-­ possessive way of living. These practices are needed to render us receptive to the truth and transformation by the truth as the Buddha saw it. It might be objected that the no-self teaching means that the Buddha did not have an account of human nature. But I believe this would be a mistake.52 The no-self teaching is an interpretation of our ordinary understanding of human nature and the Buddha obviously has much to say about characteristic features of human beings.  Ñāṇamoli and Bodhi (1995, p. 398 (I 301)). I take the Eightfold Path to be a summary of some of the more important practices required to attain enlightenment, not an exhaustive account of all the practices in the Buddha’s teaching. 48  Ñāṇamoli and Bodhi (1995, p. 390 (I 294)). 49  Walshe (1987, p. 131 (I 124)). 50  For example, see Bodhi (2000, vol. 1, pp. 901–903 (III 66–68)). Siderits analyzes this argument in (2007, ch. 3). 51  Ñāṇamoli and Bodhi (1995, pp. 782 (II 173) and 260 (I 167)) respectively. 52  It might also be claimed, from a Mahāyāna point of view, that emptiness, the claim that nothing has an inherent nature, also supports this objection. But I would question this as well. On account of limited space, I have focused here on early Buddhism mainly as expressed in the Pali Canon. However, the basic features of Buddhism I have drawn attention to, such as its psychological analyses, are also present in Mahāyāna Buddhism, and similar (though somewhat more complicated) points could be made there. It might also be objected that, on account of the no-self teaching, it is paradoxical to attribute a self-cultivation philosophy to Buddhism. But any account of Buddhist thought and practice faces this paradox (which is one motivation for the development of distinctions between conventional and ultimate truth). 47

Self-Cultivation Philosophy as Fusion Philosophy: An Interpretation of Buddhist Moral…

431

For example, some arguments for no-self employ the standard Buddhist psychology that identifies five aggregates (khandhas) of our nature as form (physical body), feeling, perception, volition and consciousness. There are also analyses of various factors that hinder enlightenment and so need to be removed as well as factors that facilitate enlightenment and so need to be cultivated.53 These draw on additional elements of Buddhist psychology. There are extensive discussions of these psychological states, understood as consistent with the no-self teaching, in the Abhidhamma literature and later Buddhist philosophical traditions. Hence, the fourth feature of self-cultivation philosophies, an understanding of human nature or psychology, is clearly a part of the Buddha’s program. As a self-cultivation philosophy, traditional Buddhist moral philosophy does not focus on the main themes in meta-ethics and normative ethics that have dominated much twentieth century and contemporary work in moral philosophy. Its concerns are closer to a number of topics in moral psychology that have gained prominence in recent years such as moral perception, attention, emotion, intention and desire as well as various character traits that involve these. For example, in early Buddhism there is much interest in the cultivation of the divine abodes—loving-kindness, compassion, appreciative joy and equanimity—mentioned earlier. Again, with the emergence of Mahāyāna Buddhism, there is an emphasis on the development of compassion and “perfections” such as generosity, patience and the like. The self-­ cultivation perspective enables us to see why these discussions are at the heart of Buddhist moral thought. It also allows us to understand how meditative practices such as mindfulness may be seen as forms of attention that are important for moral understanding. And it encourages us to read important texts such as Buddhaghosa’s Visuddhimagga and Śāntideva’s Bodhicaryāvatāra from their own point of view as guidelines in self-cultivation philosophy.54

5 Buddhist Self-Cultivation Philosophy: Past and Future I have suggested that we can think of fusion philosophy as a disposition to generate new perspectives by challenging standard understandings of disciplines and exploring alternative ways in which philosophy might be understood. Here I have proposed one way of doing this: by examining ancient Chinese, Greek and Indian philosophical traditions together, we can see diverse iterations of one form philosophy often took in the ancient world. Though it is certainly not the only form of philosophy, self-cultivation philosophy was an important and fascinating motif in the early development of philosophy. It is not a standard category in contemporary philosophical discourse, but I believe it is a valuable interpretive framework for  For example, see Ñāṇamoli and Bodhi (1995, pp. 151–54 (I 60–62)).  See Ñāṇamoli (1999) and Crosby and Skilton (1995). For valuable discussions of Buddhist philosophy along these lines, see Burton (2010), Dreyfus (1995), Finnigan (2019), Garfield (2015, ch. 9), Heim (2014), McRae (2015) and Wright (2009). 53 54

432

C. W. Gowans

understanding many ancient philosophical traditions. The figures I have highlighted as examples—Xunzi, Epicurus and Buddha—are in many respects deeply different from one another. Viewed from the perspective of much contemporary philosophy, or in isolation from each other, these figures and their philosophies might seem to be idiosyncrasies embedded in a very particular context or anomalies associated with a specific culture, historical period or religious preoccupation. But from the perspective of self-cultivation philosophy that I have sketched here they can be seen as various contributions to a distinctive kind of moral philosophy that developed across several major ancient civilizations. My focus in this paper has been historical: I have presented self-cultivation philosophy as a fruitful interpretive perspective for understanding much work in ancient philosophies. But I want to close by briefly reflecting on self-cultivation philosophy as a possible category for contemporary work in Buddhist philosophy: that is, not only for trying to understand Buddhist philosophy in the past, but for striving to develop a credible Buddhist philosophy in the present. Specifically, to what extent can classical Buddhist self-cultivation philosophies be adapted to contemporary circumstances? One implication of my argument, of course, is that this might benefit from drawing on the resources of other philosophical traditions such as those in Warring States China or the Hellenistic (and also Roman) periods of Mediterranean history.55 However, I also want to point in other directions. In my analysis, I stressed the comparison between philosophy and medicine in the development of ancient self-cultivation philosophies. An obvious issue is how plausible and productive this comparison might be for us.56 For example, to what extent is it reasonable to suppose that philosophical understanding could play a crucial part in making our lives better or making us better persons?57 In particular, could recent reconstructions of Buddhist philosophy—for instance as a consequentialist or virtue theory, or as showing how the no-self teaching could enable us to overcome suffering or justify universal compassion—play a central role in improving how we live our lives? A challenge for a contemporary Buddhist self-cultivation philosophy would be to show that they could. Related to this is the fact that we now know much more about medicine and, more importantly, we have developed forms of psychotherapy such as cognitive-behavioral therapy. These practices and the knowledge they presume may be relevant to assessing the claims of self-cultivation philosophies: the aspirations of these contemporary approaches to improving our lives are, at least in part, similar to those of the ancient self-cultivation philosophies (for example, in claiming to help us overcome various forms of dissatisfaction and anxiety). In fact, forms

 The great historical precedent for this, of course, is the development of Buddhism in China starting about the first century C.E. Both the emergence of Chan and the development of what is often called Neo-Confucianism reflect the interaction of Buddhism and native Chinese traditions such as Confucianism and Daoism. 56  I discuss this in Gowans (2010). See also Burton (2010) and Nussbaum (1994). 57  Bernard Williams raised this issue in a critique of Martha Nussbaum’s analysis of Hellenistic philosophy in Nussbaum (1994) (see Williams 1994). 55

Self-Cultivation Philosophy as Fusion Philosophy: An Interpretation of Buddhist Moral…

433

of mindfulness training based on the Buddhist tradition are now commonly considered to be one of the tools of cognitive-behavioral therapy.58 This suggests that a contemporary Buddhist philosophy that purports to be true to its ancient self-cultivation character would require fusion of a quite different sort than what I have been discussing. It would need to look to psychology, cognitive science, neuroscience, evolutionary biology, anthropology and other contemporary sciences as potential empirical resources for understanding and validating its claims, and perhaps for suggesting ways in which its practices might be enhanced or augmented. And, of course, in fact this has become an important dimension of some contemporary work in Buddhist moral philosophy.59 For instance, in light of what we know (or at least think we know or are in the process of coming to know) from such sciences, it is important to consider: whether there is reason to accept or question the Buddhist no-self teaching and Buddhist understandings of agency; whether there is evidence for or against Buddhist analyses of the cognitive and affective factors that inhibit and promote enlightenment; whether Buddhist meditation could be effective in developing our moral understanding and increasing compassion and loving-kindness (for instance, meditation practices such as mindfulness, those focusing on the divine abodes, and Śāntideva’s equality of and exchange of oneself and others); and whether overall Buddhism has a plausible understanding of moral cultivation and in particular whether we are capable of attaining Buddhist ideals such as universal compassion and the elimination of anger. There has already been much work done on these topics and such work would need to be at the heart of any contemporary Buddhist self-cultivation philosophy. It is hard to predict where this work will lead, but it is a natural adaptation of the ancient tradition in view of our increased understanding of human beings.

References Allinson, Robert E. 1989. Chuang-Tzu for Spiritual Transformation: An Analysis of the Inner Chapters. Albany: State University of New York Press. Ames, Roger T. 1985. The Common Ground of Self-Cultivation in Classical Taoism and Confucianism. Tsing Hua Journal of Chinese Studies 17: 65–97. Aristotle. 2002. Nicomachean Ethics. Trans. Christopher Rowe. Oxford: Oxford University Press. Bodhi, Bhikkhu, trans. 2000. The Connected Discourses of the Buddha: A New Translation of the Saṃyutta Nikāya, 2 vols. Boston: Wisdom Publications. ———, trans. 2012. The Numerical Discourses of the Buddha: A Translation of the Aṅguttara Nikāya. Boston: Wisdom Publications. Burton, David. 2010. Curing Diseases of Belief and Desire: Buddhist Philosophical Therapy. In Philosophy as Therapeia, Royal Institute of Philosophy Supplement, ed. Clare Carlisle and Jonardon Ganeri, vol. 66, 187–217. Cambridge: Cambridge University Press.

 For instance, see Tirch et al. (2016).  For example, see Coseru (2017), Davis (2017), Dunne (2019), and especially Flanagan (2011) and (2017). 58 59

434

C. W. Gowans

Cooper, John M. 2012. Pursuits of Wisdom: Six Ways of Life in Ancient Philosophy from Socrates to Plotinus. Princeton: Princeton University Press. Coseru, Christian. 2017. Breaking Good: Moral Agency, Neuroethics, and the Spontaneity of Compassion. In A Mirror is for Reflection: Understanding Buddhist Ethics, ed. Jake H. Davis, 109–128. New York: Oxford University Press. Crosby, Kate, and Andrew Skilton, trans. 1995. Śāntideva: The Bodhicaryāvatāra. Oxford: Oxford University Press. Davis, Jake H. 2017. ‘When You Know for Yourselves’: Mindfulness, Wisdom, and the Qualities of Heart. In A Mirror is for Reflection: Understanding Buddhist Ethics, ed. Jake H.  Davis, 223–235. New York: Oxford University Press. Descartes, René. 1993. Meditations on First Philosophy. 3rd ed. Trans. Donald A.  Cress. Indianapolis: Hackett Publishing Company. Dreyfus, Georges. 1995. Meditation as Ethical Activity. Journal of Buddhist Ethics 2: 28–54. Dunne, John D. 2019. Innate Human Connectivity and Śāntideva’s Cultivation of Compassion. In Readings of Śāntideva’s Guide to Bodhisattva Practice, ed. Jonathan C. Gold and Douglas S. Duckworth. New York: Columbia University Press. Fellezs, Kevin. 2011. Birds of Fire: Jazz, Rock, Funk, and the Creation of Fusion. Durham: Duke University Press. Fingarette, Herbert. 1998. Confucius: The Secular as Sacred. Prospect Heights: Waveland Press. Original Publication date 1972. Finnigan, Bronwyn. 2019. Śāntideva and the Moral Psychology of Fear. In Readings of Śāntideva’s Guide to Bodhisattva Practice, ed. Jonathan C. Gold and Douglas S. Duckworth. New York: Columbia University Press. Flanagan, Owen. 2011. The Bodhisattva’s Brain: Buddhism Naturalized. Cambridge: The MIT Press. ———. 2017. The Geography of Morals: Varieties of Moral Possibility. New  York: Oxford University Press. Ganeri, Jonardon. 2019. Epistemic Pluralism: From Systems to Stances. Journal of the American Philosophical Association 5: 1–21. Garfield, Jay L. 2015. Engaging Buddhism: Why It Matters to Philosophy. Oxford: Oxford University Press. Gowans, Christopher W. 2010. Medical Analogies in Buddhist and Hellenistic Thought: Tranquility and Anger. In Philosophy as Therapeia, Royal Institute of Philosophy Supplement, ed. Clare Carlisle and Jonardon Ganeri, vol. 66, 11–33. Cambridge: Cambridge University Press. ———. 2015. Buddhist Moral Philosophy: An Introduction. New York: Routledge. Graham, A.C. 1989. Disputers of the Tao: Philosophical Argument in Ancient China. Chicago: Open Court. Hadot, Pierre. 2002. What is Ancient Philosophy? Trans. Michael Chase. Cambridge: Harvard University Press. Heim, Maria. 2014. The Forerunner of All Things: Buddhaghosa on Mind, Intention, and Agency. New York: Oxford University Press. Hutton, Eric L., trans. 2014. Xunzi: The Complete Text. Princeton: Princeton University Press. ———., ed. 2016. Dao Companion to the Philosophy of Xunzi. Dordrecht: Springer. Inwood, Brad, and L.P.  Gerson, eds. and trans. 1997. Hellenistic Philosophy: Introductory Readings, 2nd ed. Indianapolis: Hackett Publishing Company. Ivanhoe, Philip J. 2000. Confucian Moral Self-Cultivation. 2nd ed. Indianapolis: Hackett Publishing Company. Keown, Danien. 2017. ‘It’s Ethics, Jim, but Not as We Know It’: Reflections on the Absence of Moral Philosophy in Buddhism. In A Mirror is for Reflection: Understanding Buddhist Ethics, ed. Jake H. Davis, 17–32. New York: Oxford University Press. Kline, T.C., III. 2006. The Therapy of Desire in Early Confucianism: Xunzi. Dao: A Journal of Comparative Philosophy 5: 235–246. Lloyd, G.E.R. 1990. Demystifying Mentalities. Cambridge: Cambridge University Press.

Self-Cultivation Philosophy as Fusion Philosophy: An Interpretation of Buddhist Moral…

435

McEvilley, Thomas. 2002. The Shape of Ancient Thought: Comparative Studies in Greek and Indian Philosophies. New York: Allsworth Press. McRae, Emily. 2015. Buddhist Therapies of the Emotions and the Psychology of Moral Improvement. History of Philosophy Quarterly 32: 101–122. Mill, John Stuart. 2002. Utilitarianism. Indianapolis: Bobbs-Merrill Company. Ñāṇamoli, Bhikkhu, trans. 1999. The Path of Purification (Visuddhimagga) by Bhadantācariya Buddhaghosa. Seattle: Buddhist Publication Society Pariyatta Editions (original publication date 1975). Ñāṇamoli, Bhikkhu, and Bodhi, Bhikkhu, trans. and eds. 1995. The Middle Length Discourses of the Buddha: A New Translation of the Majjhima Nikāya. Boston: Wisdom Publications. Nussbaum, Martha C. 1994. The Therapy of Desire: Theory and Practice in Hellenistic Ethics. Princeton: Princeton University Press. O’Keefe, Tim. 2010. Epicureanism. Berkeley: University of California Press. Parfit, Derek. 1984. Reasons and Persons. New York: Oxford University Press. Paul, L.A. 2014. Transformative Experience. Oxford: Oxford University Press. Raphals, Lisa. 2011. Embodied Virtue, Self-Cultivation, and Ethics. In Ethics in Early China: An Anthology, ed. Chris Fraser, Dan Robins, and Timothy O’Leary, 143–157. Hong Kong: Hong Kong University Press. ———. 2017. Chinese Philosophy and Chinese Medicine. In The Stanford Encyclopedia of Philosophy, ed. Edward N.  Zalta, Fall 2017 edition. https://plato.stanford.edu/archives/ fall2017/entries/chinese-­phil-­medicine/ Santana, Carlos. 2014. The Universal Tone: Bringing My Story to Light. New  York: Back Bay Books. Schofer, Jonathan W. 2000. Virtues in Xunzi’s Thought. In Virtue, Nature and Moral Agency in the Xunzi, ed. T.C. Kline III and Philip J. Ivanhoe, 69–88. Indianapolis: Hackett Publishing Company. Siderits, Mark. 2003. Personal Identity and Buddhist Philosophy: Empty Persons. Burlington: Ashgate Publishing Company. ———. 2007. Buddhism as Philosophy: An Introduction. Indianapolis: Hackett Publishing Company. ———. 2017. Comparison or Confluence in Philosophy? In The Oxford Handbook of Indian Philosophy, ed. Jonardon Ganeri, 75–89. New York: Oxford University Press. Slingerland, Edward. 2003. Effortless Action: Wu-Wei as Conceptual Metaphor and Spiritual Ideal in Early China. New York: Oxford University Press. Soni, Jayandra. 2010. Patañjali’s Yoga as Therapeia. In Philosophy as Therapeia, ed. Clare Carlisle and Jonardon Ganeri, 219–232. Cambridge: Cambridge University Press. Sorabji, Richard. 2000. Emotion and Peace of Mind: From Stoic Agitation to Christian Temptation. New York: Oxford University Press. Stalnaker, Aaron. 2006. Overcoming Our Evil: Human Nature and Spiritual Exercises in Xunzi and Augustine, Moral Traditions Series. Washington, DC: Georgetown University Press. ———. 2016. Xunzi on Self-Cultivation. In Dao Companion to the Philosophy of Xunzi, ed. E.L. Hutton, 35–65. Dordrecht: Springer. Tirch, Dennis, Laura R.  Silberstein, and Russell L.  Kolts. 2016. Buddhist Psychology and Cognitive-Behavioral Therapy: A Clinician’s Guide. New York: Guilford Press. Tsouna, Voula. 2009. Epicurean Therapeutic Strategies. In The Cambridge Companion to Epicureanism, ed. James Warren, 249–265. Cambridge: Cambridge University Press. Van Norden, Bryan W. 2017. Taking Back Philosophy: A Multicultural Manifesto. New  York: Columbia University Press. Walshe, Maurice, trans. 1987. The Long Discourses of the Buddha: A Translation of the Dīgha Nikāya. Boston: Wisdom Publications. Wang, Robin. 2010. The Virtuous Body at Work: The Ethical Life as Qi 氣 in Motion. Dao 10: 339–351.

436

C. W. Gowans

Wang, Robin R. 2012. Yinyang: The Way of Heaven and Earth in Chinese Thought and Culture. New York: Cambridge University Press. Williams, Bernard. 1985. Ethics and the Limits of Philosophy. Cambridge: Harvard University Press. ———. 1994. Do Not Disturb: Review of Martha Nussbaum’s the Therapy of Desire. London Review of Books 16 (20): 25–26. Wright, Dale S. 2009. The Six Perfections: Buddhism and the Cultivation of Character. New York: Oxford University Press.

Can We Know Whether Śāntideva Was a Consequentialist? Charles Goodman

Can we describe the ethical views of premodern Buddhist authors, without distorting them, using the terms and concepts employed in contemporary discussions of philosophical ethics? If we can, just how should we do so? Mark Siderits was one of the first authors to propose that we try to understand the normative views of the South Asian Buddhist tradition considered as a whole,1 and of Śāntideva in particular,2 as forms of consequentialism. Since his pioneering work, the discussion has advanced considerably, and scholars have raised a number of questions about, and objections against, consequentialist interpretations of Buddhist ethics. This paper defends a consequentialist interpretation of Śāntideva in particular, offering replies to the most important of these questions and objections. If my arguments are successful, they reveal that Siderits’ pioneering articles about Buddhist ethics were quite close to the mark, at least so far as Śāntideva is concerned. At the very center of the interpretation of Śāntideva put forward by Siderits and defended both by him and by myself, we find the arguments presented at Bodhicaryāvatāra (BCA) VIII.90–103.3 Philosophers who have reflected on these matters seem to have somewhat divergent assessments of how strong these arguments are; and if they are very bad arguments, we probably shouldn’t attribute them to Śāntideva. But the first of these, which appeals to the claim that I am not morally special, closely resembles a powerful argument for utilitarianism that is  Siderits (2007).  Siderits (2000). 3  See Vaidya and Tripathi (1988, pp. 161–66); Crosby and Skilton (1995, pp. 96–97). 1 2

C. Goodman (*) SUNY Binghamton, Department of Philosophy and Department of Asian and Asian American Studies, Binghamton, NY, USA e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_22

437

438

C. Goodman

advanced by prominent defenders of that view today. As for the second line of reasoning in the passage, known as the Ownerless Suffering Argument, its validity has been widely questioned. Stephen Harris’ critique of the argument makes some important points, but as I will show, can be effectively answered. As a member of the class of “arguments from below,” though, the Ownerless Suffering Argument also faces a serious substantive challenge that has been examined and debated by such contemporary philosophers as Derek Parfit and Mark Johnston. I offer some possible responses that defenders of the Ownerless Suffering Argument might use to blunt the force of this challenge. If any of these responses succeed, that will bolster a claim that both Siderits and I want to make: that Śāntideva’s works are important today not merely for their historical influence and aesthetic beauty, but also because they have significant philosophical contributions to offer to contemporary conversations about ethics. Let us first examine the interpretive debate about how to understand the overall nature of Śāntideva’s normative views. Though there is considerable scholarly disagreement about this matter to be dealt with, one quite important point seems to be widely accepted among the scholars working on Śāntideva’s ethics today. Śāntideva’s views about how to live and what to do are structured around three emotions which, taken together, correspond quite closely to what English-speaking ethicists of the eighteenth and nineteenth centuries called “impartial benevolence.” Building on this consensus, we might go further and assert, as I have before, that Śāntideva was a utilitarian; indeed, that he may quite possibly have been the first person in the history of the world to write a clear statement of utilitarianism. Yet this bolder claim is still very controversial. There is some distance between reading Śāntideva as commited to impartial benevolence and reading him as committed to utilitarianism. The ongoing controversy about the interpretation of Śāntideva’s ethics has brought into focus a series of interpretive questions that occupy that intermediate distance. Many of the objections I shall consider against the type of intepretive strategy favored by myself and Siderits can be found in the valuable work of Stephen Harris, the most important critic of the consequentialist intepretive strategy. In articulating one of the most crucial of his objections, Harris relies – as I have as well – on Shelly Kagan’s distinction between foundational and factoral levels of moral reflection.4 Here the factors of moral reflection are those considerations that should be taken into account in practical contexts while deciding what to do. The foundational level concerns those considerations that can be used to provide the most basic type of justification for moral norms. Harris and I agree in interpreting Śāntideva as holding that, at least for some advanced practitioners, the factoral level of moral engagement should be dominated by impartial benevolence. If the foundational level is also based on impartial benevolence, then Śāntideva must be a utilitarian. But Harris protests, with reason, that we cannot justifiably infer directly from the factoral level to the foundational level. A deontologist might claim that we have a duty to be impartially benevolent; a virtue

 He cites Kagan (1998, p. 190), for this.

4

Can We Know Whether Śāntideva Was a Consequentialist?

439

ethicist might see impartial benevolence as the height of human perfection; and even an egoist could hold that, because of the workings of karma, cultivating impartial benevolence is the right way to live, because it will be in my interest. If we look at the matter this way, it seems that, from the mere fact that he recommends impartial benevolence, we can infer almost nothing about the deep theoretical structure of Śāntideva’s ethics. Should we, then, just give up on the project of reconstructing his views about ethical theory, and return our focus to his factoral-level recommendations? Considerable purchase can be gained on these very difficult questions by examining a short but rich and complex passage at ŚS 145–46: As the bodhisattva Sārthavāha said in the Sūtra on Chanting the Dharma Together, “Blessed One, when a bodhisattva longs for Awakening first of all for all sentient beings, not for himself … this, Blessed One, is chanting the Dharma together.” It is only by giving things up that his own welfare is accomplished. Nevertheless, for fear of losing benefits for sentient beings, he does not place his own burden on unworthy sentient beings. But where no benefits to sentient beings will be lost, what difference does it make if the welfare of the world is promoted by him or by somebody else? Suppose that he fails to discard what is wholesome of his own in order to bring about what is wholesome for other bodhisattvas. Well, if he fears the suffering of the lower realms for himself, what others fear is also suffering. If he is indifferent, thinking “That suffering has nothing to do with me,” then as the sūtras say, he undergoes a downfall. (Goodman 2016, p. 141)

Here Śāntideva seems to regard it as mandatory, under certain kinds of circumstances, for a bodhisattva to sacrifice “what is wholesome of his own” (Skt. sva-­ kuśalam) in order to produce wholesome states in others. And this requirement is justified on fundamentally consequentialist grounds: namely, the equal moral importance of others’ suffering as compared to my own. A crucial feature of this passage is that the bodhisattva is being called upon to sacrifice not just material things, nor even the goodness or good karma (Skt. puṇya) that is seen as the cause of obtaining these things, but something much closer to what we would call virtue. The passage therefore raises a serious concern about interpretations of Śāntideva as a virtue ethicist. In some situations, virtue ethicists might urge me to sacrifice my material possessions, or my blood, or my arms, or my life; but they could never claim that morality requires me to make myself less virtuous on the whole, as this would directly contradict the foundation of their normative view. We will return to this issue. Another key lesson of the passage at ŚS 145–46 will take much longer to explain. In the first paper I published on this issue,5 I rested considerable argumentative weight on the claim that Mahāyāna ethics and consequentialism are both committed to agent-neutrality, and that this constitutes a key similarity between these two traditions of normative reflection. Agent-neutrality seems to be a helpful issue for discussion to focus on, since agent-neutrality is in some ways a clearer, more specific  Goodman (2008).

5

440

C. Goodman

and more tractable concept than that of consequentialism itself. Due to the industrious efforts of “consequentializers,” we now know that a very wide variety of ethical considerations and theoretical structures can be recast in superficially consequentialist form.6 This literature has drained much of the distinctive content out of the general concept of consequentialism. Agent-neutrality is another matter. This term and its opposite, agent-relativity, have been given exceptionally clear and now widely accepted7 definitions by Derek Parfit in Reasons and Persons. For Parfit, a theory is agent-neutral if it gives all agents the same, shared moral aims, and agent-relative if it assigns different moral aims to different agents.8 If we apply this distinction to Buddhist views, its application seems glaringly obvious. A normative theory based on the Way of the Disciples (Skt. Śrāvaka-yāna) would give each agent her own, unique aim: that she become an Arhat and transcend the cycle of birth and death. And a normative theory based on the Mahāyāna would give every agent in the world the very same aim. It’s the one expressed in BCA III.33: that all sentient beings someday attain Buddhahood, and that meanwhile, they be happy. So the first view seems to be agent-relative, and the second is transparently agent-neutral. Agent-neutrality has strong and possibly disturbing implications for our moral obligations in certain types of cases where we must respond to the wrongdoing of others. Suppose, for instance, that a gang of burglars has spent weeks systematically preparing a robbery spree; they plan to break into dozens of apartments tonight and steal whatever they can. Suppose also that there is exactly one way in which I can stop them. If I now steal their crowbar, a tool that they legally own, this will disrupt their plans enough to stop them from stealing anything tonight and give the police enough time to apprehend them before they can regroup. An agent-relative theory, such as a strict form of deontology, could give me the aim that I not steal anything, and so forbid me from intervening in this way. By contrast, imagine an agent-­neutral theory that gives everyone the aim that as few things as possible be stolen. Stealing the crowbar seems like the most effective way available to minimize the total number of thefts that occur, and so unless there are additional considerations that bear on the decision, the agent-neutral theory will imply that I should steal it. If we start from agent-neutrality, and then introduce the kinds of egalitarian background assumptions that strike modern Westerners as obviously correct, we can argue powerfully for another thesis that I will call “agent-indifference.” Suppose that a particular action, one which will have significant harmful and/or beneficial consequences for sentient beings, could be carried out by any one of several agents. Obviously it does matter whether a heart surgeon or a philosopher is the one who carries out heart surgery, since this will affect the outcome. But what if the expected result will be the same regardless of which of several agents carries out the action?

 See, e.g., Portmore (2011).  In particular, by Harris (2015, p. 15). 8  Parfit (1984, p. 27). 6 7

Can We Know Whether Śāntideva Was a Consequentialist?

441

I will use “agent-indifference” to mean the thesis that, where the overall benefits and harms to sentient beings will be the same, it does not matter, morally, which agent performs the action. Barbra Clayton has plausibly argued9 that, within Śāntideva’s normative perspective, there may be some exceptions to the principle that I’m calling “agent-­ indifference.” Citing passages from both the BCA and the ŚS in support, she writes: These verses suggest that one who has taken the bodhisattva vow is morally more significant than one who has not because the bodhisattva has taken on the task of aiding all sentient beings. Thus to interfere with the development of a bodhisattva is more problematic than interfering with a non-bodhisattva, and it is worse for a bodhisattva to fail than for an ordinary being because if the bodhisattva falters, literally everyone is deprived of benefit, and it is karmically very harmful for the bodhisattva. As such, it is worse to harm or hinder them. This implies that bodhisattvas as moral agents are “worth more” than ordinary beings and their actions and actions with regard to them have more weighty consequences. Such moral privilege seems to undermine the agent-neutrality that is said to characterize consequentialism.10

Clayton’s interpretive claims about the special status and significance of bodhisattvas strike me as correct. As Clayton herself carefully shows, though, the reasoning behind these recommendations is actually consequentialist.11 Obstructing bodhisattvas is especially morally problematic because doing so creates obstacles to the ability of those bodhisattvas to create benefits for others; and therefore the foundational source of Śāntideva’s position appeals to the welfare of sentient beings. So Clayton’s conclusion is that Śāntideva’s ethical view is consequentialist, but not agent-neutral. Though Clayton has done us a useful service in pointing out this topic, I would argue that the passages do not give us any reason to doubt that Śāntideva accepts agent-neutrality, defined as the view that morality gives everyone a shared set of common aims. Consider a modification of the case of the gang of burglars. Suppose that I, the agent who could steal the crowbar, am a student at the police academy who hopes to become a police officer and apprehend burglars. If I now steal the crowbar, I will likely be expelled from the police academy and be unable to make any further contribution to crime prevention, so that in the long run, more burglaries will eventually occur. In this modified case, the agent-neutral theory will surely imply that I should not steal the crowbar, even though I could thereby prevent more thefts in the short run. In view of this kind of example, I suggest that we should interpret Śāntideva as holding a thesis I’ll call restricted agent-indifference. When it comes to beneficial actions that help others, who performs them is of no intrinsic significance; the bodhisattva should adopt that course of action that leads to the greatest benefit for the world, whether she herself produces that benefit or not. This claim, or something

 In Clayton (2009).  Clayton (2009, p. 21). 11  Clayton (2009, p. 21). 9

10

442

C. Goodman

very close to it, is explicitly stated at ŚS 145: “what difference does it make if the welfare of the world is promoted by him or by somebody else?” On the other hand, when it comes to harmful actions, the effects of negative karma from those actions on the unfolding of the bodhisattva’s spiritual career can mean that actions that would be beneficial in the short run will be harmful from a more long-term perspective. On this reading, which agent performs the action still does not matter intrinsically; but there is, at least sometimes, a hidden but very important instrumental reason for bodhisattvas to keep their hands clean, even at the expense of allowing others’ hands to get dirty. This view of restricted agent-indifference might allow Śāntideva to avoid accepting some of the stranger consequences of classical utilitiarianism. In a famous discussion, Bernard Williams proposed the example of an unemployed chemist named George whose advisor offers to get him a job making chemical and biological weapons. When he protests that he is morally opposed to such weapons, his advisor informs him that there is another chemist, who happens to be disturbingly enthusiastic about weapons of mass destruction, and who would otherwise be in line for the position.12 Thus if George decides to take the position, the overall consequences are likely to be less bad for those at risk of harm from chemical and biological weapons than they would be if he refuses. As Williams argues, a Western utilitarian such as Mill or Sidgwick is clearly committed to the conclusion that George should accept the job. And this conclusion, for Williams, is deeply disturbing, because it means that the utilitarian theory is demanding that George should sacrifice his personal moral integrity to the greater good; and we may be attracted to the view that morality should never ask us for a sacrifice of that particular type. As may be obvious, however, Śāntideva may have available to him a possible way out. If we stipulate that George is a bodhisattva whose training would be severely impaired by having to engage in so karmically damaging an activity as the development of weapons of mass destruction, it is possible that during the cosmic unfolding of George’s spiritual career, the consequences of his beneficial actions for sentient beings may be important enough to outweigh the stakes in this case. That would allow Śāntideva to say that George the bodhisattva may have a good enough reason to refuse the job. Here it may be relevant to consider one of the strangest stories mentioned in Śāntideva’s works: the chapter on Jñānavatī from the Samādhi-rāja-sūtra. Śāntideva refers to, and quotes briefly from, this narrative in the context of an extended discussion of the propriety of meat-eating.13 The chapter tells us of a monk named Bhūtamati who is the last living transmitter of a tradition of meditative absorption, and who is seriously ill. His illness can be cured only by eating meat. Others offer

12 13

 Smart and Williams (1973, pp. 97–98).  See SS 134. The discussion as a whole is found at Goodman (2016, pp. 128–31).

Can We Know Whether Śāntideva Was a Consequentialist?

443

Bhūtamati a piece of cooked meat which, unbeknownst to the monk, is human flesh, freely offered by a princess named Jñānavatī. He eats the meat and is cured.14 Many aspects of this remarkable story could be worth considering, but the one most relevant here is the fact that those around Bhūtamati do not allow him to realize that he is engaging in cannibalism until after the fact. In a way, they protect him from having to take on the karma of knowingly and intentionally eating human flesh. If we assume for the sake of argument that there really are some people who have greatly advanced along a path of mental purification that will eventually lead to highly beneficial abilities, perhaps the rest of us could have strong reasons to protect those people from situations that would present them with tragic choices that could result in psychological damage, even at considerable cost to other values. The main purpose of these remarks has been to argue that there is no good reason to give up the claim that Śāntideva accepts agent-neutrality. The passages examined by Clayton raise issues about agent-indifference, but do not represent any serious problem for consequentialist interpretations of Śāntideva, since the basis of the exceptions to unrestricted agent-indifference is able to fit within a utilitarian framework. I now turn to another worry that has been raised by several authors: that the aspects of Śāntideva’s ethics that sound like utilitarianism may not be intended for, or applicable to, everyone. Thus Harris points out, following Michael Barnhart, that Śāntideva’s exhortations to practice universal and impartial benevolence are specifically intended for an audience of bodhisattvas. Of course, if everyone ought to become a bodhisattva, then these passages are, in some sense, applicable to everyone. But should we attribute the latter view to Śāntideva? Drawing on Nattier’s excellent work on the Inquiry of Ugra, Harris points out that there are indications that early Mahāyānists did not expect everyone to become a bodhisattva, even in the very long term.15 So Harris writes that “To strengthen his position, Goodman also owes us textual support for the position that Mahāyāna texts claim that all beings should become bodhisattvas.”16 It’s always nice to be able to pay people what you owe them. ŚS 33 is one of many passages with the same message, as follows: “May all sentient beings walk in happiness, pursuing all aspects of the bodhisattva’s way of life.”17 True, Śāntideva cites the Inquiry of Ugra many times in the ŚS, but that doesn’t mean we should seriously wonder whether he agrees with that text about the issue now in question. Harris fails to reflect on the significance of the fact that the Inquiry of Ugra is an extremely early sūtra. By the time of Śāntideva, more or less the entire Indian Mahāyāna tradition had adopted a form of the Lotus Sūtra’s doctrine of bodhisattva  See Dutt (1954, pp. 481–82).  Nattier (2003). 16  Harris (2015, p. 257). 17  Goodman (2016, p. 35). And see, as a close parallel, BCA X.1: “Through my good [that arises] from considering the Bodhicaryāvatāra, may all people become ornaments of the way of life that leads to Awakening (Skt. bodhi-caryā.)” Vaidya and Tripathi (1988, p.  289). Translation by the author. 14 15

444

C. Goodman

universalism: that everyone ought to become, and perhaps even, eventually will become, a bodhisattva. So although it is true in practice that Śāntideva is speaking only to a restricted audience––like every author everywhere––in theory, everyone should someday reach a point at which these words could be meaningful to them. Let us return, then, to the issue of the cogency of a virtue ethics interpretation of Śāntideva. Matthew Kapstein, who attended a short talk I gave on these issues, offered an interesting response to my use of ŚS 145–46 to undermine such interpretations. Kapstein’s reading of Śāntideva begins with the plausible interpretive claim that bodhicitta is a kind of sovereign virtue: it both sums up and has greater importance than all the other Buddhist qualities of character.18 Moreover, virtue ethicists certainly could recommend that someone should sacrifice a less important virtue for the sake of developing some more important virtue, thereby becoming more virtuous on the whole. Perhaps when Śāntideva claims that the bodhisattva commitment sometimes requires us to sacrifice our own wholesome states for the sake of the virtue of others, part of his meaning is that by making such an altruistic sacrifice of virtue, we will strengthen our bodhicitta and therefore become, overall, more virtuous rather than less. Given this objection, it is clear what it would take to have a genuinely decisive test case between the consequentialist and virtue ethics interpretations. Suppose that a bodhisattva were to confront a situation in which she could take an action that would result in her entirely forgetting her bodhicitta and thus being deprived of this virtue for billions of years, but that this same action would also result in ten other sentient beings arousing bodhicitta and following the Mahāyāna path. Stipulate that the consequences of performing this action would be objectively and impersonally better than the consequences of omitting it. Should the bodhisattva carry out this action, or not? Śāntideva never explicitly considers such a possibility; indeed, it is plausible to hypothesize that, on his view, forgetting one’s bodhicitta would never actually lead to better consequences than retaining and cultivating it. But given Kapstein’s invocation of bodhicitta as a sovereign virtue, this is the sort of hypothetical scenario that we would need in order to be confident in being able to distinguish the two interpretations. Though the text does not settle our issue explicitly, we can give a powerful argument that bears on how it should be settled, consistently with the overall philosophical commitments expressed by Śāntideva. BCA VIII.95–96 read: When happiness is dear to myself and others equally, what is so special about me that I strive after happiness only for myself? When fear and suffering are disliked by me and others equally, what is so special about me, so that I protect myself and not others?19  Extensive textual evidence could be cited in support of this claim. See, e.g., ŚS 9, at Goodman (2016, p. 11). 19  See Vaidya and Tripathi (1988, p. 163). This is my translation, but I follow Crosby and Skilton in several respects. 18

Can We Know Whether Śāntideva Was a Consequentialist?

445

If there is nothing special about me, then why should my bodhicitta be so important in my own deliberations that I treat it as if it were more practically significant than the bodhicitta of ten other sentient beings? Any cogent response to this point would, it seems, equally undercut the reasoning just quoted. It follows that if these two verses are seriously meant – if Śāntideva actually accepts this reasoning, at least at the conventional level, rather than offering it as a mere exercise in skillful means that should not be taken seriously as reasoning – then any interpretation of his ethics as egoism or as virtue ethics can be ruled out. The argument for altruism tells powerfully in favor of the consequentialist reading. Some authors who have worked in this area seem to think that the argument at BCA VIII.95–96 is very weak, weak enough indeed to provide us with a reason of interpretive charity not to read Śāntideva as committed to it. Therefore it seems relevant to point out that extremely similar arguments are taken quite seriously by philosophers today. These two verses closely parallel the core of Philip Pettit’s argument for consequentialism in his justly celebrated contribution to Three Methods of Ethics. Pettit’s essay is worth a much more thorough discussion than I can provide here. I shall merely note the following passage, which plays a crucial role in his argument: The universalizability constraint, as I understand it, cannot be plausibly rejected. To reject it would be to argue that some individual agents or people or places or whatever are so special that while something involving them is right to do, or is right for them to do, it is not right in general for that sort of thing to be done, or for that sort of individual to do it. But to take this line is to say that there is nothing persuasive to be said in moral exchange between people who differ in their attachment to the individual in question.20

Josh Greene’s widely read, semi-popular book in defense of utilitarianism, Moral Tribes, contains a similar passage: To me, I’m special. But other people see themselves as special just as I do. Therefore, I’m not really special, because even if I’m special, I’m not especially special. There is nothing that makes my interests objectively more important than the interests of others.21

Thus, close relatives of Śāntideva’s argument are being made by prominent consequentialist philosophers today. The fact that a form of words in an ancient text sounds like an argument that philosophers today would be prepared to make, though, does not entail with certainty that it is the same argument or even that it is intended as an argument at all. As scholars interested in Śāntideva seem to agree, the BCA is a poem, a philosophical treatise, and a meditation manual, all at the same time. We can point to specific individual passages which unambiguously have the primary function of moving our emotions,22 convincing us of a thesis,23 or instructing us how to direct our attention

 Pettit (1997, p. 142).  Greene (2013, p. 200). 22  Take, say, BCA II.45, Crosby and Skilton (1995, p. 18). 23  Is there any other way to read BCA IX.141? Crosby and Skilton (1995, p. 130). 20 21

446

C. Goodman

so as to progress towards Awakening.24 But there are cases in which we may not be sure, about a specific verse or passage, which combination of these functions it may have. In particular, we may worry that what appears to us to be an interesting and important philosophical argument may never have been intended to be taken in that way, but may merely constitute instructions for a meditation practice that involves imagining something to be true that Buddhists of Śāntideva’s time would in other contexts readily have conceded to be false. And we may worry that a passage we are inclined to read as having an internal logical structure may actually contain no arguments or internal inferential relations at all, but may simply be a series of reflections arranged in the order that the author thought would be most psychologically compelling. Harris cogently raises the first of these two worries about the argument for altruism; additionally, and remarkably, he introduces the second worry as well. About BCA VIII.95–96, Harris writes that An opponent could raise obvious objections against the normative interpretation of the verses I have suggested above. For instance, even a Buddhist accepting the doctrine of no-­ self (anātman) might point out that my current and future collections of physical and mental states are closely connected and claim that this justifies prioritizing my own welfare, even though the other person desires happiness as much as I do. In Indian works of metaphysics and epistemology, objections to the position an author sets forth are developed in detail before being responded to, and the fact that Śāntideva does not do so here suggests that his primary motivation may be psychological transformation, rather than normative theorizing.25

This objection does not survive confrontation with the canonical commentary of Prajñākaramati. The hypothetical opponent makes an argument from absence: Śāntideva does not respond to an obvious objection, so he is not making an argument. But in fact Prajñākaramati does consider this very objection: Perhaps it will be said: “While there is no self or the like, still there is a single series, and likewise the collection of many things such as hands and feet is a single body. In this case such things as warding off harm and the like as one’s own will be restricted to the pair that is suitably linked in this world and the next. Hence your ‘for there is no difference’ [between the case of another person and the case of the hand and foot] has an unproven reason, and the previous reason [“because it is suffering”] is inconclusive.26

This passage settles at least this much: Prajñākaramati does not think that VIII.95–96 are merely a meditative exercise. He thinks these verses are intended to provide rational reasons for altruism, reasons against which opponents might raise objections, claiming that they are unproven or inconclusive. Furthermore, Prajñākaramati thinks that Śāntideva intended VIII.101 as a response to the objection just explained. As the commentary states, To this objection he says,

 BCA V.54 is a very clear example; Crosby and Skilton (1995, p. 38).  Harris (2015, p. 6). 26  Goodman and Siderits (2016, p. 246). 24 25

Can We Know Whether Śāntideva Was a Consequentialist?

447

101. The series and the collection, like a queue, and army and so on, are unreal. The one who owns the suffering does not exist. Therefore, whose will it be? The series does not exist as a single ultimately real entity. But this is just a stream-like succession of resultant moments in the relation of effect and cause, for nothing distinct from that is apprehended… This is only nominally existent. So attachment to this is not right … Thus since it lacks an ultimately real object, the thought is mistaken. Alternatively this means that it does not hold up under analysis. Hence, since there is nothing whatsoever like a self that could be an owner, there is nothing to which suffering is connected. Hence “whose will this suffering be that is thought to be one’s own?” The meaning is that there is no one at all.27

There are two issues here that should be kept carefully apart. We can ask whether Prajñākaramati is right to think that verse 101 is a response to the objection from the close connections between present and future time-slices of the same series; and we can ask whether, if it is, the response is any good. As to the first issue, if there is any kind of scholarly case to be made that Prajñākaramati misunderstands the verses on which he comments, no one has yet made it. As to the second issue, though, Harris has made his position clear: he thinks it would be a bad response. In an article published in 2011, he has attacked the reasoning of BCA VIII.101–103  in quite strong terms. Harris examines this issue from several perspectives and in relation to the interpretations proposed by several writers, including both Siderits and John Pettit. In each case he raises the same principal objection against it, which takes the form of a dilemma. If we are operating in a context in which we see common-sense moral beliefs as presumptively correct, then the argument surely fails. But once we endorse anātman and rely on it to reject selfishness, thereby decisively breaking with common sense, then why should we continue to accept our ordinary intuition that anyone’s future suffering, including our own, matters at all or should be prevented? On this horn of the dilemma, Harris says, altruism does not follow: Here, we can remember, Śāntideva has presented us with a choice: either all suffering must be eliminated, or none of it. Again, we can ask, why should we be required to choose the first option? Complete altruism and total apathy are alike in being ap-parently rational states for an equanimous being to adopt.28

Concluding from these considerations that “all configurations of this passage as an argument fail to be convincing,”29 Harris once again proposes that we consider reading the passage, not as a rational argument, but only as meditation instructions. To me this seems like a grave underestimation of Śāntideva’s argument. If we insist that the conclusion of the argument of BCA VIII.90–103 must be that altruism is the uniquely rational disposition, we indeed face a problem. But is entirely consistent with Harris’ dilemma objection that Śāntideva has offered a valid argument for a disjunctive conclusion: Since selfishness is irrational, there being no self, the  Goodman and Siderits (2016, pp. 246–47).  Harris (2011, p. 116). 29  Harris (2011, p. 118). 27 28

448

C. Goodman

most rational approach to life must be either a universal and unrestricted altruism, or total normative nihilism. But any utilitarian would be overjoyed to reach such a conclusion, another form of which would be the claim that if anything matters at all, utilitarianism is true. Forced to choose between utilitarianism on the one hand and, on the other, a nihilism so radical as to imply that I have no reason whatsoever to brush my own teeth, there are few ethicists indeed who would not opt for utilitarianism. If Śāntideva’s argument can establish the disjunctive conclusion, then although it would not quite go so far as to establish altruism beyond all challenge, it would resolve the “dualism of practical reason” which besets Sidgwick’s Methods of Ethics,30 driving egoism and virtue ethics from the field and leaving utilitarianism as the only normative view left. Still more can be said on behalf of Śāntideva’s argument. As various forms of mental suffering arise in my subjective awareness, I directly experience them as bad. If I pretend to reject this, I deny my own experience. Granted this point, I can run a simple argument in the style of Śāntideva, as follows: Future suffering is bad, since it is suffering, like this present suffering. Even if it is our attachment to a sense of self that normally makes us care about our own future suffering, that fact will not stop this simple argument from being cogent. I for one find it utterly convincing. If I have him correctly, Harris maintains that even if we accepted the simple argument, that would not be enough to get us to altruism. Suppose we agree, he says, that The phenomenal feel of suffering is negative and as such it is to be removed. Nevertheless, the opponent may object that simply pointing out the fact that a mental state is negative does not obviously entail the normative conclusion that we ought to prevent or remove it. Of course, generally we are motivated to remove these negative mental states in ourselves, those we love, and even strangers, but arguably this may be the case only because we falsely believe ourselves and others to be enduring persons.31

I cannot agree with the argument presented in these remarks. Suppose we accepted a severely restricted form of motivational internalism: If I would never, under any circumstances, be motivated to prevent x from happening, then I do not believe that x is bad. It would then follow that, if recognizing no-self destroyed all motivation to prevent suffering  – certainly a coherent and metaphysically possible empirical hypothesis – it would have to do so by, among other things, removing the belief that suffering is bad. The motivation to prevent suffering that is an inherent aspect of the view that suffering is bad might not be particularly strong: under many circumstances, perhaps it would not be strong enough to lead to effective altruistic action. But assuming motivational internalism, if the actual practice of meditation destroys my emotional confusion and self-cherishing, leaving behind an awareness that suffering is bad, then the motivation flowing from this awareness will be left alone to work. In the absence of any obstacles that could limit its operation, that awareness would be enough to motivate compassionate action.

30 31

 Sidgwick (1981, pp. 507–9 et al.).  Harris (2011, p. 108).

Can We Know Whether Śāntideva Was a Consequentialist?

449

Now say that we do not assume motivational internalism. Well, the issue at hand was not really about motivation in the first place. What we want to know is whether the Buddhist teaching of no self makes altruism the only rational approach to life. So consider the principle Singer appeals to in his famous article “Famine, Affluence, and Morality”: that “if it is in our power to prevent something bad from happening, without thereby sacrificing anything of comparable moral importance, we ought, morally, to do it.”32 The obviousness of this principle, and the immense difficulty of finding a reasonable way to deny it, help to explain the immense influence of the article in which it appears. It might be true, empirically, that whatever degree of motivation we presently have to put this principle into practice somehow depends on our sense of being temporally extended selves. Even if this were true, it would be, in a sense, irrelevant. The question is whether the truth of Singer’s principle depends on assumptions about our being temporally extended selves. And it is very far from obvious that it does. Certainly, anyone arguing that the principle does so depend would need to provide an argument in support of that claim; and Harris does not. I conclude that Harris’ objections against the reasoning of BCA VIII.90–103 are unconvincing. There is, however, another kind of worry about the Ownerless Suffering Argument, and one that raises deep and difficult philosophical issues. Śāntideva’s argument can be understood as having the form of an “Argument from Below.” An argument from below starts from the premise that a certain phenomenon just consists in entities, events or processes at a lower level. The argument then licenses us to conclude that the higher-level phenomenon has no value or significance of its own, over and above the significance of what constitutes it. In the most relevant example, if the continued existence of a person just consists in the occurrence of and relations between various physical and mental events, then the significance of that person’s continued existence is nothing more than the significance of those events and the relationships between them. There are important reasons to worry about the validity of arguments from below. One of these was first pointed out by Mark Johnston,33 and then brought to the attention of scholars of Buddhist philosophy by Roy Perrett in an article in Philosophy East and West.34 In Reasons and Persons, Parfit had claimed that arguments from below are valid in general. Johnston persuaded him to qualify this claim by pointing out that, if physicalism is true, mental events and processes just consist in physical events and processes at the subatomic level. But individual events and processes at the subatomic level pretty clearly don’t have any intrinsic normative significance. So if we apply an argument from below to this kind of materialist view, we get the conclusion that people’s mental lives have no intrinsic normative significance. And there’s hardly any distance at all between that conclusion and an unrestricted nihilism about morality.

 Singer (1972).  In Johnston (1997). 34  Perrett (2002). 32 33

450

C. Goodman

The conversation between Parfit and Johnston did not end there; a further and highly illuminating exchange took place between them about the viability of these arguments. By carefully examining this disagreement, we may be able to learn much about the prospects for the Ownerless Suffering Argument today. Parfit and Johnston hold similar views about many relevant issues. They agree that our ordinary, intuitive conception of our own existence and identity is mistaken at a deep level. They agree that the available scientific evidence strongly favors the conclusion that souls, Cartesian egos, or other spiritual substances do not exist. And they agree in saying that, when we arrive at an accurate understanding of the metaphysics of our own existence, such understanding will provide rational support for caring more than most of us do now about the interests and well-being of others. They disagree about whether that support for altruism can be provided through appeal to arguments from below: Parfit continued to defend these arguments, and Johnston to critique them. We should be careful to note, though, that although it falls into the class of arguments from below, the conclusion of Śāntideva’s Ownerless Suffering Argument is significantly more radical than any view to which Parfit commits himself. That’s because Parfit leaves the door open to the possibility that the relations among the mental and physical events that make up people have normative significance in and of themselves. According to what Parfit calls the Moderate Claim, Relation R, which consists in a combination of psychological connectedness and psychological continuity, does have some normative significance. According to the Extreme Claim, such relations have no significance in and of themselves; only the character of the mental events actually matters. So the Ownerless Suffering Argument is the result of accepting the Extreme Claim and then applying an argument from below to humans. Parfit does not profess to be able to settle the disagreement between the Extreme Claim and the Moderate Claim. He does mention several writers, including Butler, Sidgwick and Wiggins, to whom he tentatively attributes the view that if Reductionism about personal identity is true, the Extreme Claim follows.35 I think that this is correct. When we reflect carefully on Relation R, we see that it doesn’t really matter in and of itself. One reason is the arbitrariness hiding in its structure. Psychological continuity obtains if there is a chain of connections each member of which is at least half as strong as the connections that normally obtain between within a person’s life between one day and the next. Why should “at least half” be the right figure? Further, suppose that, due to dramatic technological advances, I came to know that this body would most likely survive for a thousand years. Should I want the personality in this body to continue to change, learn and develop over time, hopefully in the direction of greater knowledge, wisdom and compassion? Or should I hope that the personality in this body will eventually become stagnant and static, so that I never stop being psychologically connected to this body in the future? If I

35

 Parfit (1984, p. 307).

Can We Know Whether Śāntideva Was a Consequentialist?

451

really have good reason to care about Relation R, I should hope for the second possibility, but that seems very implausible to me. It’s likely that these arguments won’t be convincing to the majority of philosophers, who want to be able to preserve at least some fragments of a common-sense undersanding of practical reason. Certainly, in a full vindication of the Ownerless Suffering Argument, we would need to provide convincing reasons to adopt the Extreme Claim rather than the Moderate Claim. But, of course, that discussion is of little significance if arguments from below are invalid as applied to people. In a prize essay published online in 2007, Parfit proposed a way to distinguish the argument from physicalism to nihilism from the kinds of argument from below that he wants to endorse. There Parfit conceded that the mere fact that certain phenomena just consist in certain others is not sufficient to license a valid argument from below. And yet some such arguments do appear to be valid. Suppose there are a small number of trees close together on the top of a hill. Now suppose I learn that, in English, this group of trees can be referred to as a copse. Parfit claims – and he seems to be right to say so – that the copse has no value or significance over and above the value or significance of the individual trees and their relations to each other. Given that I already know about the trees, if I learn that a copse also exists, this is not, for Parfit, a new fact about reality, but merely a fact about our language, with no independent normative significance. For Parfit, what makes the argument from below possible in this case is that the fact of the copse’s existence is, in a certain sense, “merely conceptual.”36 Given what we mean by “copse,” it is true, as a matter of logical necessity, that there is nothing to a copse but some trees and their relations to each other. Parfit thinks that it follows that the value of a copse is nothing over and above the value of the trees and their relations. In his terminology, to accept the validity of an inference like this is to accept “Realism About Importance.” Mental processes are disanalogous in important respects. Physicalists today usually concede that space aliens who were physically very different from Earth life forms could still have mental lives. Most also agree that a sufficiently advanced silicon-based computer could also support mental processes. Drawing on this agreed basis of multiple realizability, Richard Boyd writes this: [T]he version of materialism best supported by available evidence entails that mental states admit sufficient plasticity in the way in which they are realized that it is logically possible for mental states to be nonphysically realized, even though in the actual world all mental phenomena are physically realized. By plasticity of a type of event, state, or process I understand its capacity to be realized in more than one way …37

But if, in some distant possible worlds, mental processes are realized in a non-­ physical way, then it is not logically necessary that they are no more than certain specific physical processes. Therefore, it is perfectly coherent to accept the

36 37

 Parfit (2007, p. 36).  Boyd (1980, p. 87). See also Chalmers (1996, p. 148).

452

C. Goodman

argument from below as applied to copses and reject the argument as applied to mental processes, thus avoiding moral nihilism. Into which category, though, should we put people? Parfit repeatedly applies Realism about Importance to reject the view that personal identity is what matters. So he holds that it is a conceptual truth that a person is nothing more than just a series of physical and mental events. The argument from below succeeds in the case of people because it is conceptually impossible for there to be people who are not selfless. Johnston begins his response to Parfit’s claims by accepting much of this apparatus; indeed, he describes the copse example as “nicely chosen.”38 And he accepts a version of Parfit’s Realism about Importance that appeals to the a priori character of our knowledge that a copse is a small group of trees and the a posteriori character of our knowledge that materialism is true. And then Johnston deftly turns Parfit’s own argument against him. For in Reasons and Persons, Parfit had claimed that the non-reductionist view might have been true. His example goes as follows: suppose that we discovered convincing evidence of the truth of reincarnation. That, Parfit thinks, would give us strong reasons to conclude that the soul exists, thereby refuting Reductionism.39 So our knowledge that Reductionism is true is empirical; moreover, the claim that a person is nothing over and above a collection of mental and physical processes is contingent, since a person might instead have been a soul that was the owner of some mental and physical processes. It would follow that arguments from below cannot apply to people. Johnston’s conclusion is that we can admit the doctrine of anatta, in the sense that there are no persisting souls or selves worth caring about, and yet still allow that personal identity and the persistence of individual personalities matter.40

There is something rather odd about this way of putting the point. Johnston’s discussion emphasizes the Buddhist historical roots of the denial of any self worth caring about, the claim he shares with Parfit and calls anatta. What is remarkable is that the Buddhist tradition is committed to reincarnation, precisely the empirical phenomenon that, Parfit claims, would, if it existed, prove the existence of the soul.41 So either the great philosophers in the Buddhist tradition were simply too stupid to recognize the conflict between reincarnation and no-self, or something has gone seriously wrong with Johnston’s argument. Historical Buddhist thinkers were able to claim that reincarnation and no-self were compatible because they were dualists of a quite unfamiliar kind. While unequivocally rejecting the existence of spiritual substances, they also claimed that mental processes were non-physical. In this context, Buddhist philosophers  Johnston (2010, p. 312).  Parfit (1984, pp. 227–228). 40  Johnston (2010, p. 315). 41  Johnston himself calls attention, at pp. 314–5, to the possibility that out-of-body post-mortem experiences could be shown empirically to provide information that could not have been obtained in any other way. Yet some Buddhists also believe in experiences of this kind. 38 39

Can We Know Whether Śāntideva Was a Consequentialist?

453

compared reincarnation to the transmission of flame from one candle to another. That is, they understood reincarnation as the continuation of a process rather than the continued existence of a substance. Their process understanding of persons allowed Indian Buddhists to accept that the wheel of existence worked in practice roughly as their Hindu neighbors thought it did, while simultaneously rejecting the soul that was the metaphysical keystone of the Hindu religious system. Buddhist philosophers tried to motivate this remarkable combination of views with a series of a priori arguments against the existence of souls. According to them, substance dualism was self-contradictory, because nothing could possibly have all the properties and functions that the soul was supposed to have. Indeed, the notion of substance in general was seen as deeply philosophically confused. Belief in a soul was regarded by these Buddhists as an especially problematic and especially important instance of a rationally unjustifiable mental tendency to reify, and thereby to misunderstand not only how things actually exist, but even what forms of existence are possible at all. At least one of these a priori arguments can be found in the Bodhicaryāvatāra itself. Śāntideva writes: If the Self is eternal and without thought processes, then it is evidently inactive, like space. Even in contact with other conditioning factors, what activity can there be of something which is unchanging? What part does something play in an action if, at the time of the action, it remains exactly as it was prior to it? If the relationship is that the action is part of it, then which of the two is the cause of the other?42 The key steps in this reasoning are the claims that an eternal substance must be unchanging, and an unchanging entity cannot participate in causal relations. Buddhists held that something unchanging, if it were to produce an effect, would have to produce the same effect at every single moment at which it existed. If it produced different effects at different times, it would have to change. But in order to be the agent of actions, a soul would have to cause different effects at different times. Now, of course, the soul might be in contact with other contributing factors. But in order to register the presence of those factors, so as to adjust its contribution, the soul would have to be able to change its internal state depending on which factors were present; and an unchanging entity, by definition, can’t do that. This argument against a permanent self or soul occurs in many Buddhist texts. In the Soul chapter of the Tattvasaṃgraha, Śāntarakṣita repeats it ad nauseam, using it against almost all the positions he criticizes. Buddhists used this argument in a theological context as well, to refute the idea of a permanent God. It closely resembles the arguments in analytic philosophy of religion that question whether a timeless, unchanging, simple God would be able to act in a world of time. If He can’t change his internal state, how can He know what time it is? If He can’t know what time it is, how can He choose to act differently at some times than at others?

42

 Śāntideva, Bodhicaryāvatāra VI.29–30, following Crosby and Skilton p. 52.

454

C. Goodman

You may or may not be impressed with this a priori argument. I find it quite interesting and at least somewhat compelling, but I don’t expect everyone to agree. In the same way, I don’t know what it would cost you, in terms of philosophical commitments that you would have to take on board or throw overboard, to embrace the argument in these verses. The price might be quite high. But if you pay it, you get the Ownerless Suffering Argument for free. If the a priori argument succeeds, then the soul is metaphysically impossible; so it will turn out to be a conceptual truth that a person simply is a series of mental states together with whatever constitutes them, just as it is a conceptual truth that a copse just is some trees. Therefore, persons will have no intrinsic significance over and above the intrinsic significance of the mental states that make them up and the relations among those mental states, just as the value of the copse is no more than the value of the trees and the relations among them.43 As Siderits has informed me, and as could be inferred from the tenor of some of his writings, he is not an advocate of a priori knowledge. As a result, he is likely to be unwilling to accept the argument I have just explained. As it happens, though, there may be an alternative route to the same conclusion, one that could be more congenial to a philosopher with Siderits’ epistemological commitments. This alternative route begins with a well-known passage in Chapter IX of Vasubandhu’s Abhidharma-kośa: [Vasubandhu:] The object of the term “I” and of this identification of the self is the aggregates. [Opponent:] How do you know? [Vasubandhu:] Because of people’s affection for them. 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.44

Here Vasubandhu’s point concerns both the psychological process called ahaṃkāra in Sanskrit, and the meaning of the term “I” in ordinary usage. The term ahaṃkāra, though shared in common by Buddhists and at least some of their opponents, is interpreted quite differently by each side. Whereas advocates of the soul understand the ahaṃkāra as an identification of what the self is, one that may be either correct or incorrect, Buddhists see it as a psychological process that actually constructs an inescapably erroneous sense of self. For Vasubandhu, the philosophical advocate of a soul badly misrepresents the ordinary common sense to which he would like to appeal as against the markedly counterintuitive Buddhist teaching of no self. Asked a question like “Please tell me something about yourself,” very few of us would reply with such assertions as “I am invisible,” “I am permanent and unchanging,” or “I am simple, with no parts.” Yet

 And the intrinsic significance of whatever implements the mental states, if it has any such significance. 44  Goodman (2009, pp. 303–4). 43

Can We Know Whether Śāntideva Was a Consequentialist?

455

these are qualities of the soul as postulated by the Nyāya-Vaiśeṣika tradition, and so, assuming the truth of that tradition’s views about the self, should be correct answers to the question. If Vasubandhu is correct in claiming that the ahaṃkāra “has the aggregates as its object” – a claim which, it should be noticed, even many Buddhists would reject – then Johnston’s claim that the existence of a soul, though probably not actual, is in some sense possible, will not pose the kind of threat to arguments from below that he supposes it does. For if Vasubandhu’s argument succeeds, a soul, even if it were to exist, would not actually be the basis of our personal identity. This claim is closely related to Locke’s point, at Essay Concerning Human Understanding II.27.10,45 that when we consider hypothetical cases in which the memories, personality traits, or other psychological characteristics of an individual are transferred from one soul to another, most of us have the strong intuition that identity goes with the bundle of characteristics, not with the same spiritual substance. These considerations lead to the idea that even though it is metaphysically possible, and perhaps even in some weak sense epistemically possible, that a soul might be involved in the continued existence and identity of a person, a version of Reductionism might nevertheless follow from our own self-understanding, so that most or all forms of non-­ Reductionism are either incoherent on their own terms or cannot give us what we unavoidably want from a theory of personal identity. If they can be defended, these conclusions would be enough to vindicate the Ownerless Suffering Argument, even without appeal to the distinction between a priori and a posteriori ways of knowing. I have defended the interpretation that I share with Siderits of Śāntideva as a consequentialist ethical theorist, and the arguments for altruism in the BCA that play such a central role in that interpretation, from a variety of objections. But let me conclude by noting one final set of problems that I have not tried to resolve in this essay. Śāntideva was, obviously, a Mādhyamika; yet the Ownerless Suffering Argument seems to be formulated from an Abhidharma perspective. How can we understand the relation between these two forms of Buddhist philosophy within Śāntideva’s thought? The issues raised by this question are immensely difficult. I have tried to engage them to some degree in my contribution to a collaborative volume on Buddhist ethics on which I worked closely with Siderits and with several other remarkable scholars.46 But I do not claim to have done more than take some preliminary steps in the direction of resolving them. If only Śāntideva were an Ābhidhārmika, there would be few if any real obstacles to understanding him as fully committed to the two arguments for altruism that he so brilliantly stated. But as things stand, the difficulties involved in seeing how the arguments for altruism could possibly fit into a Madhyamaka worldview constitute the most important remaining reasons for doubting whether he truly was committed to a consequentialist view of ethics.

45 46

 See Nidditch (1979, pp. 335–336).  The Cowherds (2016, ch. 8).

456

C. Goodman

References Boyd, Richard. 1980. “Materialism without Reductionism: What Physicalism Does Not Entail.” In Ned Block, ed. Readings in the Philosophy of Psychology, vol. 1 pp. 67–106. Chalmers, David. 1996. The Conscious Mind: In Search of a Fundamental Theory. New York: Oxford U. Press. Clayton, Barbra. 2009. Śāntideva, Virtue, and Consequentialism. In Destroying Māra Forever: Buddhist Ethics Essays in Honor of Damien Keown, ed. John Powers and Charles Prebish. Ithaca: Snow Lion. Crosby, Kate, and Andrew Skilton, trans. 1995. Śāntideva. The Bodhicaryāvatāra. New  York: Oxford World’s Classics. Dutt, Nalinaksha. 1954. Gilgit Manuscripts. Vol. II, part 3. Calcutta: Calcutta Oriental Press, Ltd. Goodman, Charles. 2008. Consequentialism, Agent-Neutrality, and Mahāyāna Ethics. Philosophy East and West 58 (1): 17–35. ———, trans. 2009. Vasubandhu’s Abhidharmakośa: The Critique of the Soul. In Buddhist Philosophy: Essential Readings, ed. William Edelglass and Jay Garfield, 297–308. New York: Oxford University Press. ———, trans. 2016. The Training Anthology of Śāntideva: A Translation of the Śikṣā-samuccaya. New York: Oxford University Press. Goodman, Charles, and Mark Siderits, trans. 2016. Bodhicaryāvatāra-pañjikā VIII.90–103 by Prajñākaramati, Commenting on Śāntideva. In Moonpaths: Ethics and Emptiness, by the Cowherds, 241–248. Oxford University Press. Greene, Joshua. 2013. Moral Tribes: Emotion, Reason, and the Gap Between Us and Them. New York: Penguin. Harris, Stephen. 2011. Does Anātman Rationally Entail Altruism? On Bodhicaryāvatāra VIII.101–03. Journal of Buddhist Ethics 18: 93–123. ———. 2015. On the Classification of Śāntideva’s Ethics in the Bodhicaryāvatāra. Philosophy East and West 65 (1): 249–275. Johnston, Mark. 1997. In Human Concerns Without Superlative Selves, ed. Jonathan Dancy. Reading Parfit. Johnston, Mark. 2010. Surviving Death. Princeton: Princeton U. Press. Kagan, Shelly. 1998. Normative Ethics. Boulder: Westview Press. Nattier, Jan, trans. 2003. A Few Good Men: The Bodhisattva Path According to the Inquiry of Ugra (Ugraparipṛcchā). Delhi: Motilal Banarsidass (First published 2003; Indian reprint, 2007). Nidditch, Peter, ed. 1979. John Locke. An Essay Concerning Human Understanding. Oxford: Clarendon Press. Parfit, Derek. 1984. Reasons and Persons. New York: Oxford University Press. ———. 2007. “Is Personal Identity What Matters?” Available at https://www.stafforini.com/docs/ parfit_is_personal_identity_what_matters.pdf Perrett, Roy. 2002. Personal Identity, Minimalism, and Madhyamaka. Philosophy East and West 52 (3): 373–385. Pettit, Philip. 1997. The Consequentialist Perspective. In Three Methods of Ethics, ed. Marcia Baron, Philip Pettit, and Michael Slote, 92–174. Malden: Blackwell. Portmore, Douglas. 2011. Commonsense Consequentialism: Wherein Morality Meets Rationality. New York: Oxford University Press. Siderits, Mark. 2000. The Reality of Altruism: Reconstructing Śāntideva. Philosophy East and West 50 (3): 412–424. ———. 2007. Buddhist Reductionism and the Structure of Buddhist Ethics. In Indian Ethics: Classical Traditions and Contemporary Challenges, ed. P. Bilimoria, J. Prabhu, and R. Sharma, vol. 1, 283–296. Burlington: Ashgate. Reprinted in Westerhoff, Jan, ed. 2016. Mark Siderits. Studies in Buddhist Philosophy, pp. 263–276. Oxford: Oxford University Press. Sidgwick, H. 1981. The Methods of Ethics. 7th ed. Indianapolis: Hackett (First published 1907).

Can We Know Whether Śāntideva Was a Consequentialist?

457

Singer, Peter. 1972. Famine, Affluence, and Morality. Philosophy and Public Affairs 1 (1): 229–243. Smart, J.J.C., and Bernard Williams. 1973. Utilitarianism: For and Against. New York: Cambridge University Press. The Cowherds. 2016. Moonpaths: Ethics and Emptiness. New York: Oxford University. Press. Vaidya, P.L., and Sridhar Tripathi, eds. 1988. Bodhicaryāvatāra of Śāntideva with the Commentary Pañjikā of Prajñākaramati. 2nd ed. Darbhanga: Mithila Institute.

Selfhood and Resentment Rick Repetti

1 Preliminary Claims and Disclaimers I am not attempting to articulate “the” Buddhist view of agency or free will. Some object that my view is not even properly understood to count as “a” Buddhist perspective, e.g., Meyers (2018). However, the Pudgalavādins endorsed “a” Buddhist view of the self, and constituted the largest minority of the early Indian Buddhist community, despite their endorsement of a doctrine rejected by most other Buddhist schools. To say that they were not true Buddhists would be to commit a Buddhist version of the No True Scotsman fallacy, the “No True Buddhist” fallacy. Siderits (2017), for example, described a ‘Buddhist Reductionist’ view of free will as “a” possible Buddhist view—a view a Buddhist could adopt given conceptual resources in Buddhism—although he does not assert it, nor claim any Buddhists ever held it. Likewise, my view is one a Buddhist may adopt, given conceptual resources in Buddhism.

2 Preliminary Justifications Because Buddhist practice is framed as an attempt to bring about the realization of no-self, meditative instruction shapes practitioners’ expectations, practices, and experiences accordingly. Rather than reveal its absence, however, Buddhist meditation can increase agency, and thus selfhood.

R. Repetti (*) CUNY/Kingsborough Community College, Department of History, Philosophy, and Political Science, Brooklyn, NY, USA e-mail: [email protected] © Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6_23

459

460

R. Repetti

Agency and selfhood are under attack in contemporary scientific research and simultaneously threatened by the explosion of interest in secular mindfulness. Critics of “McMindfulness” (Purser and Loy 2013)1 assert that mindfulness instruction encourages a nonjudgmental stance toward experience, diminishing socio-­ political agency. Referring to the current polarized state of affairs, Cozort (2017) notes worries associated with a rising tide of political libertarianism, individualism, and egoism associated with Ayn Rand, emphasizing that “the Buddha … refuted the underlying premise of libertarianism, namely, that we are autonomous moral agents” (xvi). Some, like Pereboom (2001), Caruso (2018), and Harris (2013), think we are better off without belief in free will and ultimate moral responsibility, predicting that the no-self view will engender greater compassion. I’m unconvinced. Peter Strawson (1962) argued that the truth of determinism would not undermine our normative practices because they are too entrenched; Galen Strawson (2017) argued that Buddhist meditation can help determinists alter their responses. I argued (2017b) that Buddhism may make it possible for individuals to gradually accept a no-self view. However, this may not be desirable. If one believes she lacks selfhood or agency, she may be more likely to be manipulated, for there is nothing in her psychology to defend. As the adage has it, if you stand for nothing, you’ll fall for anything. If there’s no you, there can be no standing of any kind. What would be the point? Absent belief in agency, studies (Martin et al. 2017) suggest an individual may become slave to her strongest passions. Group think often trumps individual thinking in all but those with a strong sense of agency. In the worst-case scenario, coming to believe one lacks agency might destroy one’s agency, in which case the no-self doctrine would be the cause of psychic suicide for those who practice meditation in the hopes of bringing about the experience of a no-self—arguably a form of philosophical zombiehood. Thus, it is prudent to view with caution any claim that enlightenment consists in the realization that there is no agent-self. Only an agent-­ self has reason to engage in the network of reactive attitudes that constitute the lived societal framework of normativity. Peter Strawson thought the necessity of this framework would survive putative news of the truth of determinism. Would the opposite hold for belief in the self?

3 Agency and Selfhood Are we, or are we not, agents? We experience ourselves, for the most part, as agents. Harris (2013) claims that the phenomenology of experience reveals that we do not actually experience ourselves as agents. This may be true in his case, although his track record of successful agency suggests otherwise, but it is not true in my

 For a collection of criticisms, see Purser et  al. (2016); for a reply to such criticisms, see Repetti (2016). 1

Selfhood and Resentment

461

experience.2 My earliest memories involve agency that has grown over the years, particulary in connection with my long-term meditation practice, ironically. We begin flailing our limbs without understanding they’re ours or what moves them, as a kaleidoscope of sensations floods our consciousness. As associations and patterns emerge with repetition, they eventually form perceptual objects integrated in ways random sensations are not: that face unites eyes, nose, and mouth, and moves while the room doesn’t. That face is a thing, not a fleeting sensation, and it makes that sound. Intuitively, the degree of conscious attention, focused by a desire to understand, makes the difference in terms of how quickly an infant learns. I recall, at age three, when adults were sitting around our kitchen table, and I knew enough to know they were speaking, despite that their speech was too complex to grasp, barring an occasional word. I suspect an earlier time when I knew their speech had meaning, though I knew none of it. I surmise that my intent to decipher language played a role in the language-acquisition process, exhibiting a kind of proto-agential activity in which curiosity-intentions played a causal role in my learning. Gradually, we learn those are our limbs and they move by our willings: my hand grips my mother’s finger when I squeeze my hand. This is a milestone in the emergence of agency: the ability to move my body by my efforts, to crawl, grab, bite, swallow, scream, close eyelids, and so on. Bodily control is the earliest, most basic, fundamental element of the embodied sense of agency, which is why we raise our hands when prompted to show we have free will. Around age 4 volitional bodily control characterizes our conception of free will: studies suggest we think free will is being able to do what we want to do; by age 6, we think free will is being able to not do what we want, presumably having noticed the consequences of simply doing whatever we want (Kushnir et al. 2015).3 This ability to control whether one acts on one’s desire or refrains from doing so is more sophisticated than the ability to act on desires, what may be described as free won’t. Being able to control whether we enact our desires is arguably what characterizes our evolving sense of ourselves as agents,4 and our senses of ourselves become more differentiated based on the different types of choices we make and actions we perform as we mature: we become different kinds of people through our choices, e.g., impulsive or prudential, cautious or daring, selfish or altruistic, etc.5 But the volitional somatic sense of control—a rudimentary sense of embodied free will insofar as we can move our bodies volitionally (Brent 2017)6 and a rudimentary

 See Repetti (2019) for an in-depth critique of Harris’s argument.  This interestingly parallels the shift from pre-conventional to conventional developmental stages in moral psychology. 4  Frankfurt (1971) identifies personhood with having a metavolitional hierarchy such that, ideally, we enact only those desires we approve. 5  Frankfurt (1988) argues that those volitions we identify with and those we dissociate from (“externalize”) characterize the process of self-formation and personal identity. 6  See also Brent (2018) for an application of this reasoning against the Buddhist no-self doctrine. 2 3

462

R. Repetti

sense of free won’t insofar as we can refrain from doing so—forms the embodied core of our sense of ourselves as agent-selves. The infant mystery of control over bodily movement is rediscovered, however, during walking meditation, which raises the question: from whence do walking volitions arise, how do they manifest, and what connects them with bodily muscular efforts? A mature practitioner of walking meditation, focused on sensorimotor activity at the somatic/kinesthetic level, finds it puzzling trying to locate walking volitions. She slows down to make walking more mindful, feeling the heel lift, the ball of the foot lose contact with the ground, the suspended leg moving forward, the heel touching the ground, but those sensations differ from the phenomenology of intention that moves the leg. The experience is clearly that of voluntary movement, despite the opacity of intention and the sometimes-robotic sense of effortless walking. Repetitive motions can quickly take on this character, but this is not the same as experiencing such movements as involuntary. We know how to walk from childhood, but do not know how it is that we do, despite the neurophysiological explanation. This is complicated by the Taoist idea of wu wei, “doing without doing”, or flow, that state of perfect attunement expert athletes and others claim to attain when their skills so perfectly match the challenges at hand that they say they are “in the zone” (Csikszentmihalyi 1990). It is difficult to explicate the differences between flow walking, walking meditation, power walking, ordinary walking, and mindless walking, although degree of attention and intention intuitively play causal roles in the differences. Enlightenment allegedly consists in the realization that misunderstanding agency is the primary confusion precipitating all suffering. But it seems agency is the better part of the explanation for all voluntary behavior—walking and refraining, at will, along with everything else we do and that seems to matter most to us. If agency is an illusion, then does enlightenment involve agentless action?7 Recalling the infant’s experimental attempts to discover how to control its limbs, correlated with its limited or absent sense of any embodied agency, however, one wonders how an enlightened being controls his limbs if he lacks any sense of embodied agency? If he is in a permanent state of flow, is his behavior no longer voluntary? That’s counterintuitive. The Tibetan Buddhist philosopher, sage, and meditation master Tsongkhapa advocates adopting the agential stance regarding oneself—despite the no-self doctrine—and the agentless stance regarding others (so as to minimize negative reactive attitudes towards them), noting that the Buddhist path prescribes ways to increase control over oneself, even over powerful emotions (McRae 2017). But if we are not autonomous agent-selves, can we do so, and, if not, why bother trying? The no-self doctrine is not only puzzling because it represents a perspective beyond the grasp of unenlightened mortals. It is puzzling for considerations of logical consistency.

 See Repetti (2019) for an in-depth exploration of this possibility.

7

Selfhood and Resentment

463

4 Nobody Home? What would it mean if there was no agency or agent-self. If there is no agent in here, for me, you, and everyone else, then we are all some sort of philosophical quasi-­ zombies. (Full philosophical zombies are otherwise identical but lack consciousness.) If we possess consciousness, but lack agent-selfhood, then the light of consciousness would be on, but nobody would be home. Consciousness would be causally superfluous, an appendage serving no purpose.8 If there is no causal relationship between consciousness and the movements of our bodies, the speech that comes out of our mouths, or anything else we do, then everything that matters to us is illusory. We might think we can grasp this based on certain experiences we have that resemble it, e.g., being in the zone. But it is difficult to understand enlightenment as a case of nobody home, of awareness with no subject that directs actions, but merely notices them. Nāgārjuna asserted the interdependence of the agent, agency, and action, in line with the Buddhist doctrine of dependent origination, the interdependency of all causes and conditions. But being at one with action does not entail one no longer exists. Can there be speech without a speaker, even for someone so skilled at speech that they are at one with it? How could speech occur without intentional agents on both ends, speaker and listener? The Buddha presumably intended to convey his wisdom, and devoted the bulk of his enlightened life to doing so, to his hearing audience members. If he did not intend it, why should anyone listen to him? His speech would then be equivalent to involuntary sneezes, coughs, etc. Nobody speaks to nobody listening? The very concept of rationality, similarly, arguably makes no sense absent the intentionality of the reasoner. Being persuaded by a sound argument cannot happen in a world in which there is nobody to persuade or be persuaded. Similar arguments undermine the idea that rationality can occur in a deterministic world: seeing a point, accepting an argument, realizing something is a fallacy or a valid argument— all these cannot be meaningful in a framework in which the only explanatory money, so to speak, is cashed in deterministic terms, e.g., neural firings that lack intentionality. Neural firings do not operate in accordance with principles of rationality. Buddhists are divided about whether dependent origination is deterministic, but even if it is not, the same objections seem to apply to a not-fully-deterministic version of the doctrine. An increasingly popular attempt to solve this problem, which Siderits (2017), Harvey (2017), Meyers (2017), and others have developed, deploys the two truths doctrine, which—simplifying greatly—differentiates between conventional truth, what ordinary folks believe about ordinary objects and things, and ultimate truth, which enlightened beings grasp about how those ordinary things are empty of the

 For a critique of this “consciousness epiphenomenalism”, see Repetti (2019).

8

464

R. Repetti

metaphysically substantive character or nature that ordinary beings ascribe to them.9 As Nagasena the Buddhist monk famously argued with King Milinda in the Milinda Pañha, the chariot the king rode is not more than the collection of its parts, nor identical to that collection, but the term “chariot” is merely a pragmatic, convenient, conventional designation of the object. Likewise, “Nagasena” is just a convenient designator for the collection of parts that serves a practical purpose. At the level of ultimate metaphysical truth, there is neither a chariot nor a Nagasena, but both exist at the conventional level. Conventional language is true at the conventional, pragmatic level, but ultimate reality lacks the sorts of objects described conventionally, which conventional understanding erroneously construes as metaphysically substantive. In previous writings, I developed this dialethic strategy in an attempt to work out the meaning of the theory, proposing that as agency increases along the path, at the conventional level, the sense of an independent agent-self decreases, and vanishes— according to Buddhist doctrine—upon enlightenment, at which point the sage enjoys the perception of ultimately agentless truth (Repetti 2010b). Here, however, I am considering the opposite: that ‘worldlings’—simplifying, the Buddhist term for ordinary beings—effectively lack a real self or agency, but meditation virtuosos cultivate a real agent-self. For it appears that effective agency definitively increases along the Buddhist path, sometimes to titanic levels, and certain of the Buddha’s arguments and claims may be taken to suggest that there is real agency and possibly a real agent-self for meditation virtuosos. I have experienced what may be described as undifferentiated meditative trance states, what the Buddhists call “jhanas”, and which they describe as approaching or approximating nirvāṇa. But, as Nozick (1981) rightly objects, maybe that is just what it is like when the cognitive apparatus is damped down to being analogous to a soundless stereo: the stereo is on, the speakers are live, the turntable is spinning, but the stylus is not touching the vinyl. Absent any supramundane ideology, these experiences might be meaningless. The mere fact that the sense of self is, so to speak, off-line, inactive, or merely tacit, is no argument against its reality. If it were, then dreamless sleep, and even dreams lacking an episodic/narrative center, would prove the non-existence of the self. But, clearly, they do not. Likewise, agents so in the flow that they feel no self/other boundary does not thereby disprove the existence of a self/other boundary, any more than someone who closes their eyelids disproves the existence of the visible world. Indeed, these expert agents began as arm-flailing infants, but thousands of hours of training fed their cultivation of skill that no longer requires being looped through the agent self-consciously. Skill, as Aristotle, the Buddha, and Confucius agree, is a function of habituation. But that doesn’t erase the agency required to formulate those expert levels of skill, nor entail that their exercise is involuntary.

 That is the Mahāyāna view; some early Buddhists see ultimate reality as the micro-level at which momentary atomistic psychophysical tropes are the ultimately real building blocks of conventional reality (Siderits 2017). 9

Selfhood and Resentment

465

Just as a backward-looking tracing analysis enables us to attribute negligent responsibility to someone who is so intoxicated that they are non-compos mentis when they commit some otherwise blameworthy action if they may reasonably be held responsible for having become non-compos mentis, say, by intentionally drinking too much alcohol precisely to give them the courage and the excuse for their behavior, we can employ a backward-looking tracing analysis to attribute praise to someone so skillful that they spontaneously commit praiseworthy actions if they may reasonably be held responsible for having become so skilled that their behavior is always spontaneously praiseworthy (Repetti 2010a). This seems to hold even if their spontaneous skill is so perfected that they are psychologically similar to someone who is non-compos mentis—that is, if being able to reflect on alternative courses of action, weigh consequences, recognize the optimal course of action under the circumstances, make a conscious resolve and the necessary effort to enact it are requirements for being compos mentis, and are absent in the expert agent’s case. Assuming perfectly spontaneous ideal behavior is possible for the enlightened, does their agency disappear? If my reasoning about learning to ambulate makes sense, conscious voluntary effort plays a key role in the cultivation and maintenance of embodied agency. Thus, it is difficult to imagine voluntary behavior without agential input from consciousness, or at least dispositional agential input when one is in a flow state: should things unexpectedly veer off course (say, due to the deceptive trickery of ingenious manipulative interveners), the agent remains able to consciously, voluntarily intervene in the otherwise spontaneous activity. If she is unable to regain voluntary control, then this allegedly enlightened state is not desirable. If agency disappears, is there any difference between plants, the tides, the weather, and enlightened behavior? Without the tension between yin and yang, so to speak, without an organism/environment differentiation or duality, can there be any experience at all? Appeals to mystical oneness or nonduality as transconceptual “yogic perception”—and thus beyond the grasp of the rational mind, not unlike appeals to God’s inscrutable ways—here beg the question, which is whether such ideas make sense conceptually. Do these individuals still make ordinary choices whenever confronted by completely novel circumstances, or are they like highly fluent speakers of a language who know its entire lexicon, who are presumably able to speak coherently without reflectively deliberating about grammar or syntax? Have they attained such a level of skill that there is no longer any conscious attention, effort, or other element of agency ever required, no matter what circumstances in which they find themselves? While that seems theoretically possible, it seems equally implausible to think such a level of spontaneity is not somehow guided by consciousness, whether in linguistic or non-linguistic domains of action, even if the relationship between intention and action is automated. Another way of expressing this implausibility is that it is one thing to be able to attain that level of skill in one domain, or even in a few, versus being able to attain it in every domain of human activity. The concept of an omni-savant is extremely implausible. But the Buddha is depicted as such, and arguably though nobody is home.

466

R. Repetti

Nirvāṇa is described as the unconditioned, whereas everything in samsara is conditioned. Maybe the attainment of nirvāṇa establishes a psychological threshold which, once attained, renders everything spontaneously harmonious, being permanently in the zone, so to speak, which is functionally equivalent to there being no self/other boundary, thus no psychological basis for a possible mismatch between the enlightened organism and its environment. However implausible, if we assume that, arguendo, then the unconditioned still presents a puzzle resembling the mind-­ body problem: If mind is non-physical, how can it causally interact with physical bodies? Likewise, if nirvāṇa is unconditioned, how can it interact with everything else that is conditioned? How can it raise my arm, as opposed to yours, if there is no me/you distinction? It is unclear how the Buddha, in an unconditioned mental state, engages with any conditioned phenomena at all: How does he move, teach, raise his hand in a gesture? Does some sort of gestalt constituted by the organism/environment field simply manifest through the Buddha’s body, the way clouds become dense, form water molecules, and just rain? That’s fanciful mystical thinking—that is, it’s implausible. What possible evidence could there be for it? It’s one thing to utter abstract doctrines like emptiness, nonduality, and interdependence, but anyone not already committed to such ideology requires evidence strong enough to ground such claims. Within the context of the no-self discussion, relative to the ubiquitous experience of agency and the dependence on the reality of agency presupposed by all other things that matter to us, the premise on which to ground the conclusion that there is no self cannot be more dubious than that conclusion itself, otherwise that begs the question. Thus, arguably, resting the no-self doctrine on the idea that the Buddha experienced and embodied it, which is difficult to even understand, or resting it on other, more esoteric doctrines, amounts to begging the question. Nor can the phenomenology of momentary trance states serve as grounds for the no-self claim, for reasons noted above: eclipsing the experience of self does not entail its non-existence in dreamless sleep, meditation, or otherwise. The Buddha not only performs actions masterfully, but—crucially—claims he can think whatever thought he wants to think, not think any thought he wants not to think, have any resolve he wants to have, not have any resolve he wants not to have, and, by extension, have or not have any emotion, attentional object, perception, or other mental state he wants to have or does not want to have, respectively (Aṅguttara Nikāya II.36–7). Call this “the Buddha’s mental freedom claim”. How can the Buddha state that he possesses such self-mastery, if there is nobody home? It seems incoherent to think the Buddha’s mental freedom claim emerged from environmental conditions constituting his organism/environment field/gestalt, moving his vocal chords to utter such a claim of self-mastery. If there is no self, then why doesn’t the universe speak equally through me, as opposed to speaking through him, and why is it that I lack the Buddha’s mental freedom, if there is no metaphysical difference between us, both of us lacking any really existing organism/environment boundary? Apologists claim that the illusion of separateness explains it, but who has the illusion? Nobody?

Selfhood and Resentment

467

The impression of a being completely merged with its environment is at odds with the Buddha’s mental freedom claim, which depicts an ability to control what goes on inside the organism side of the organism/environment field that greatly surpasses the strongest ability that proponents of libertarian free will advocate, the agent-causal ability to do otherwise, standardly considered implausible by its critics because it flirts with Cartesian dualism. However, Buddhists and Hindus alike allege the attainment not only of mental freedom, but of titanic psychic powers along the meditative path that are far more incredible. But rejecting agent-causality while accepting the sort of titanic agency implied by the Buddha’s mental freedom claim and the like is to apply a double standard—special pleading, to the effect that the Buddha is special indeed.

5 Freedom of the Mind I find it useful to compare the Buddha’s mental freedom claim with a similar, but far less ambitious claim, made by Frankfurt. Frankfurt (1971) distinguishes freedom of action from freedom of the will. Freedom of action is being able to enact desires, as when a horse wants to run and does. Recall 4-year-olds being able to do what they want, and 6-year-olds being able to not do what they want. Freedom of the will, for Frankfurt, consists in having the sort of effective desires one wants to have, effective desires being the ones that issue in action, and not having desires issue in action that one does not want to issue in action. Freedom of the will is having the sort of will one wants to have. The Buddha is able to have the resolve he wants to have, which, in Frankfurt’s language, is having the will one wants to have—free will. The Buddha thus satisfies Frankfurt’s criteria for free will. But Frankfurt’s analysis is restricted to a simple harmony between one’s desires and metadesires, whereas the Buddha’s control far exceeds mere meta-volitional harmony, and includes metamental control over all mental states: cognitive, conative, affective, and any somatic states that may be brought under voluntary control. If the Buddha’s mental freedom claim is correct, it implies “freedom of the mind”, having the sort of mental states one wants to have, which includes freedom of the will or volitions, but also of thought, emotion, attention, and anything we can bring under voluntary control. Can a non-agent-self exert such titanic, masterful self-control? That’s prima facie implausible, if not incoherent. The Buddha never stated there’s no self, technically, only (in the Anattalakkhaṇa Sutta) that this or that constituent (or aggregate of constituents) of ourselves is not the self, because if it were, then we could have it of that constituent that it be as we wish it to be, namely, free of suffering; but we cannot exert that sort of control over our constituents, so they are not the self. Let’s call this the “no-control/no-self” argument. However, this argument implies self-control is a criterion for the self. If lacking self-control implies no self, then it follows logically that possessing self-­ control implies a self. However, whereas the average person lacks such control, which implies the average person lacks a self, the Buddhist path prescribes how they

468

R. Repetti

can attain self-control, implicitly attaining selfhood. Such reasoning implies a spectrum between non-control/non-self and control/self. Crucially, the Buddha’s mental freedom claim asserts that he possesses the control that in the no-control/no-self argument he implies is necessary for a self, but that ordinary folks lack. This implies that while ordinary folks lack a self, the Buddha has one, since he can control the aggregates constituting him: he can have the sort of mind, thoughts, volitions, feelings, etc. he wishes to have, namely, that they be free of suffering. (The Buddha’s Third Noble Truth is that nirvāṇa—mental freedom—is the cure for suffering, and his Fourth Noble Truth, the Eightfold Path, prescribes the path that leads to the end of suffering.) The Buddha satisfies his own implicit criteria for a self; we generally don’t. The most coherent account of real existence in Buddhism is the causal-­ontological principle: something is real if it has causal powers. The Buddha’s mental freedom claim implies he has causal control over his mental states, and the no-control/no-­ self argument implies such control entails a self. Implicitly, then, the Buddha has a real self. Self-controlling agency is arguably real on this line of reasoning, however previously unnoticed and inconsistent with other interpretations of Buddhism. There are other grounds in early Buddhism for resisting the idea that the Buddha thought there was literally no agent-self or agency. The Buddha rejected the ideas of inevitabilist causation by fate, gods, karma, matter, and chance (Repetti 2017b). These are admittedly not identical to the concepts constituting contemporary free will skepticism (namely, determinism, indeterminism, and luck), but they are cognate forms of free will skepticism that he ridiculed when he asked the fatalists whether their legs walked themselves. By analogy, the idea that mouths speak themselves is equally ridiculous. That the potter, pot, and pot-making are interdependent, as Nāgārjuna insisted, does not eliminate the reality of any of these triadic nodes; likewise, that the agent, agency, and action are interdependent does not eliminate their reality. Again, this relates to the Buddhist analogue of the dualist’s mind-body problem: explaining how an unconditioned mind can engage with conditioned phenomena. One possible answer, which I have explored in previous writings (Repetti 2010b, 2017c), is that the meditation virtuoso increases causal/functional agency and simultaneously decreases the metaphysically substantive conception of the self. However, it is not clear whether this less-substantively-construed self is no self at all, or just a more flexible, fluid, free type of self. What the proper account of the enlightened understanding of the self is, I suggest, is an open question, at least for anyone who is not enlightened, which likely includes almost everyone who has written on the subject. The vague, abstract, and/ or otherwise enigmatic statements made about it by the allegedly enlightened leave its interpretation significantly underdetermined. Because some of the divergent doctrines that have evolved throughout the history of Buddhism may reasonably be construed as having been sensitive to the sorts of influences that meme theory suggests, it seems reasonable to consider the possibility that the literalist version of the no-self doctrine is something that evolved for purposes of skillful means, however unintentionally. But the fact that the pronouncements about the self, from the

Selfhood and Resentment

469

Buddha through the early Buddhist sages, are collectively indeterminate itself counts as some reason to think their vagueness may have been intentional.

6 The Pudgalavādins, Perspectivalism, and Stealth Skillful Means The Pudgalavādins affirmed the indeterminate reality of the person, which they thought was the only way to steer between the horns of the dilemma between the eternalism of the unchanging immaterial ātman of then-prevalent Vedic culture and the nihilism of post-mortem annihilation asserted by then-materialist Carvakas, as well as the only way to explain the core Buddhist doctrines of karma and rebirth. The fact that the Pudgalavādins’ view is no longer held by a living community of Buddhists is irrelevant to its assessment, but the fact that they existed provides a historical precedent for the view that what may be called the “yes-self” doctrine is not necessarily unorthodox, contrary to popular opinion. But even if the Pudgalavādin view was unorthodox, it must be recalled that Buddhism, Jainism, Carvaka, and a host of other doctrines are unorthodox relative to the dominant Vedic philosophies of India within which Buddhism arose. The relationship between orthodoxy and truth is like the relationships between law and truth, culture and truth, belief and truth: we hope they converge, but they frequently don’t. Although one person’s meditation experiences are merely anecdotal, they count for something if they are fairly stable over multiple decades of practice. Thus, in my experience as a multiple-decades practitioner, it appears to me that my meditation practice has increased the causal/functional effectiveness of my agential abilities, on the one hand, and modified my self-conception, such that I realize that my pre-­ philosophical conception of myself as an immaterial, unconditioned, transcendent spirit has been replaced by a sense of myself as an embedded, interdependent, conditioned being. However, I have not only not lost the sense that I exist as a center of causal efficacy in the world, but I have become increasingly convinced of it, despite the difficulty of explicating the nature of the self in analytically precise terms. Arguably, the reason both Śāntideva and Tsongkhapa encourage Buddhist practitioners to adopt what Breyer (2013) calls “Buddhist perspectivalism”—the idea that they should view others’ behavior impersonally, but their own personally—is that advanced practitioners might develop such a (potentially unskillful) sense of agency that an emphasis on the literalist interpretation of the no-self doctrine promises to serve as the best soteriological antidote. Following Dawkins’ main idea about memes as cultural analogues of genes, the ideas that work tend to catch on, independently of their probable or actual truth. Perspectivalism arguably glosses over the implicit contradiction between one being an agent and others not being agents. Because Śāntideva, who did try to address this conflict, may be thought of as trying to dance around this contradiction, that is reason enough to wonder why. Recall that in most Indian contemplative

470

R. Repetti

traditions, it is believed that as one cultivates greater meditative skill, one also cultivates psychic powers. Thus, on the one hand, Buddhists and Hindus alike believe that meditative skill brings with it tremendous mental acuity, mental powers, and even psychic abilities; on the other hand, just as increased skill in weight lifting and martial arts training are likely to increase practitioners’ egos, so too heightened mental powers threaten to do the same thing, which is undesirable on the path. The trick is to get aspirants to become extremely effective practitioners without thereby bolstering their egos. That meditation virtuosos cultivate titanic levels of agency only makes the need for such a strategy all the more pressing. The cultivation of psychic powers prior to conquering the flame of ego-craving, analogous to the cultivation of great scientific and technological ability by our species prior to our moral and spiritual evolution, poses a threat. Using psychic powers in the service of ego-cravings is dangerous. Thus, along with advocating adopting the agent stance for purposes of self-control, it makes sense to advocate the view that enlightenment consists in the realization of no-self for purposes of diminishing the likelihood of enhancing ego-craving as one enhances functionally-effective agency. One who is always on guard against ego-enhancing behaviors is less likely to enable them if she thinks the ultimate goal is the dissolution of the illusory ego, as compared to practitioners who believe some sort of self, however indeterminate or ineffable, persists through lifetimes and bears karma. On the latter conception, that of some sort of self that persists through lifetimes and bears karma, one has solid reasons to cultivate good karma for one’s own sake, as opposed to for the sake of attaining enlightenment, much less the attainment of enlightenment for the sake of all sentient beings, which latter aspiration defines the saintly bodhisattva vow. Just as it is conceivable the literalist no-self meme may have caught on due to its soteriological efficacy within the Buddhist world, so too it is conceivable the bodhisattva ideal—of aiming not at one’s own enlightenment, but at approaching enlightenment for the sake of all sentient beings—may have caught on. Enlightenment for the sake of the being who seeks it is arguably the greatest act of self-interest, if not selfishness, insofar as nirvāṇa is construed as the summum bonum in Buddhism, the greatest good, the greatest bliss, and the greatest mental state. Aiming for it arguably enhances the sense of self, self-grasping, and the propensity toward self-gratification that is soteriologically counter-productive. Thus, just as the literalist interpretation of the no-self doctrine may have evolved as an effective antidote strategy to dampen the ego-enhancing effects of increased agency along the path, so too the bodhisattva ideal may have evolved as an effective antidote strategy to dampen the ego-enhancing effects of increased concern with one’s own attainment of nirvāṇa. However, just as one has solid reasons to cultivate good karma for one’s own sake, as opposed to for the sake of attaining enlightenment, if one adopts any sort of self-conception other than the no-self doctrine, so too one has solid reasons to cultivate good karma for oneself coincident with seeking the attainment of enlightenment for the sake of all sentient beings if one adopts any sort of self-conception other than the literal no-self doctrine. For if it is true that the bodhisattva aspiration is more altruistic, benevolent, and noble, and thus engenders more and better good

Selfhood and Resentment

471

karma for those who adopt it, then there is great self-interest in adopting it, paradoxically and somewhat ironically. One (sometimes disputed) interpretation of the bodhisattva vow is to postpone the full attainment of one’s own enlightenment until all other sentient beings attain nirvāṇa. This is a Buddhist version of Jesus’s claim that he who is first shall be last, and vice versa. Somewhat humorously, one wonders what will happen when there are only bodhisattvas left: which of them will attain enlightenment last? Will the one who wants to literally be the last actually be the one with the greatest desire to be the noblest being? Wouldn’t that disqualify him? Humor aside, another Christian idea comes to mind here: St. Augustine, who prayed to God to make him chaste, but not yet (so he could enjoy some more sin). The bodhisattva has the best of all possible worlds, if Buddhist metaphysics about karma and rebirth is accurate, for she gets to cultivate the best karma possible and reap its rewards maximally, staying in the world of form as long as possible, while enjoying nirvāṇa-approximating mental states maximally, just enough to maximize happiness but not enough to disappear into paranirvāṇa (the final nirvāṇa experienced when one has no more karma left to cause another incarnation). Whereas Mahāyāna Buddhists nearly unanimously adopt the bodhisattva ideal and use it as the basis for comparatively depicting the earlier Buddhist path as an inferior path, arguably it is the bodhisattvas who implicitly wish to prolong their joyful experiences in increasingly better realms within samsara, the cycle of rebirth, not unlike Augustine. Equally arguably, such a preference only makes sense if one implicitly adopts some version of a self-conception other than the literalist no-self doctrine, notwithstanding doctrinal verbal pronouncements to the contrary. Returning to the issue of perspectivalism, which suggests that from one perspective it is useful to adopt a view of oneself as agential and from another as unreal, Tsongkhapa may or may not have been trying to resolve this conflict by reminding practitioners that because of their practices they are able to control themselves in the face of the negative behavior of ordinary beings who lack such skills entirely. But whatever his motive in making the distinction, his explanation of the difference between practitioners engaged in mind-training and worldlings pushed and pulled about by their otherwise uncontrolled mental states works to explain the perspectival asymmetry (McRae 2017), and I would highlight the crucial underlying fact that the difference is one of causal control, which makes it a real difference, consistent with the idea that worldlings lack a self but virtuosos possess one. It clearly supports my general claim that the meditative path increases agency, in ways that verify the Buddha’s mental freedom claim, and thus, implicitly, agency and selfhood. Buddhists distinguish between the lived, conventional experience of ordinary beings, and what is ultimately true about that experience—that everything is momentary, interdependent, and empty of any independent, substantial nature. The reality doesn’t change upon enlightenment; only our understanding and experience of it does, akin to the way our understanding and experience of a mirage changes upon realization of the fact that it is a mirage and how it is produced (Garfield 2015). Thus, the ordinary person arguably experiences some sort of self, personhood, and free will, just as in the world of ordinary experience there are such things as carry-on luggage, economy seating, and pretzels, but upon attaining

472

R. Repetti

enlightenment our understanding of these renders them analogous to the mirage. This is another form of perspectivalism: from the conventional perspective, there are agents; from the ultimate, enlightened perspective, there are not. But Tsongkhapa’s perspectivalism is reversed: the ordinary person lacks agency, whereas the Buddhist virtuoso possesses it. Analysis of the Buddha’s claims likewise. Can we reconcile these opposing forms of perspectivalism? The Pudgalavādins may have been trying to do just that. Breyer’s perspectivalism is about Tsongkhapa’s distinction between adopting the non-agential stance toward worldlings and the agential stance towards oneself (by meditation virtuosos), whereas the perspectivalism regarding the distinction between the unenlightened view of ordinary objects and persons at the conventional level and the emptiness of all things at the ultimate level is the traditional Buddhist two truths doctrine. If both forms of perspectivalism are cogent, each may be considered to involve a set of two truths. Each set of truths contains elements at some odds with each other within that set, but both sets are also at some odds with each other set. Conceivably, advanced meditative trance states do not reveal that there is no self, but simply render it inactive, thus invisible. Consider this thought experiment and its dilemma. Imagine attaining nirvāṇa: it either involves awareness or it doesn’t. If it does not involve awareness, how can one know about it? If it does involve awareness, doesn’t awareness imply a subject experiencing it? Isn’t it self-illuminating? If not, who is it that is allegedly aware that it does not exist as a subject of awareness? It is difficult to imagine what awareness of no-self could even mean.10 Perhaps that is another reason why the Pudgalavādins thought there was an indeterminate, ineffable self.

7 Self-Development or Self-Destruction? A more problematic idea, suggested earlier, is the possibility that there is a real self, but meditative practice of bringing about mental states that lack a self eventually dissolve the previously real self, bringing about an act of psychic suicide. As implied—unintentionally—by Cozort’s quote above, a philosophy that fosters belief in the literalist no-self doctrine is thus arguably the greatest enemy of individualism, autonomy, and personhood, and the entire universe of meaning that rests on such foundational assumptions. So, rather than revering Buddhism’s long tradition of saintly sages and aspirational selflessness narratives, perhaps we ought to take them all with a grain of skeptical salt before surrendering our autonomous agency in the name of faith in the esoterically enigmatic.

 For an excellent collection of essays on whether awareness implies a self, see Siderits et  al. (2013).  For a comprehensive collection of essays on divergent Buddhist  views of free will and agency, see Repetti (2017a).  10

Selfhood and Resentment

473

One thing is clear, however: meditative training cultivates an effective sense of agency and selfhood that increases through practice, contrary to the idea that selfhood decreases. Reactivity decreases, mindlessness decreases, and the extent to which one’s first-order desires and emotions have a life of their own decreases, but the sense of being able to control desires, emotions, thoughts, attention, and the like increases. Agential efficacy clearly increases along the meditative path, and agential efficacy is autonomy, self-control. Given the causal-ontological principle, that agency is real. Paradoxically, the meditation virtuoso exhibits the maximal level of self-controlling powers—maximal autonomy—that she probably decreasingly has reason to deploy, given her evolving dispositions, whereas ordinary worldlings have little if any control over their own mental states, and thus little if any genuine autonomy or selfhood, as is likely the case with most less evolved animals.

8 Conclusion This line of reasoning admittedly runs counter to the widespread belief that Buddhist meditation leads to the realization that there is no self. Again, it could, given the belief that it does, but that is arguably highly undesirable, and to the extent that practitioners approximate the blotting out of their agency, the worse off they likely will be. The worldling likely thinks she has a self, but arguably lacks one, and the virtuoso likely thinks she lacks a self, but arguably has one‚ the presence or absence being a function of control. To go against Buddhist doctrine and widespread contemporary thought and adopt this “yes-self” doctrine, ironically, one must possess a robust sense of self. Taking these considerations seriously could prevent the gullible from devoting decades of disciplined practice trying to bring about psychic suicide that could conceivably convert them into philosophical quasi-zombies. Such matters are not best left to groupthink—whether from pre-modern Buddhism or contemporary thought. Given the reality and importance of effective agency and selfhood, resentment and all the other reactive attitudes and normative practices (that Peter Strawson thought would survive the putative discovery that determinism holds and entails the unreality of free will) remain rationally appropriate. But not because they are so entrenched that removing their foundation would not lead to their revision. Rather, on this analysis and interpretation of the way Buddhist meditation increases agency and selfhood, the foundations of normativity found in the reality of agency and selfhood remain solid, thanks to the causal-ontological principle, the Buddha’s own mental freedom claim, and the implicit criterion for selfhood contained in his no-­ control/no-self argument. Purser and Loy (2013) and Purser (2019) argue that “McMindfulness”, the widespread embrace of secular mindfulness meditation practices in the West, tends to spread the message that the causes of stress are to be dealt with by adjusting one’s internal responses to them (via stress-managing, nonjudgmental meditative practices), rather than by attempting to change societal, institutional structures that

474

R. Repetti

perpetuate exploitation and injustice, and that this has the devious effect of protecting the status quo, rendering workers more complicit and productive, and feeding capitalist disparities. That’s one possibility, but Buddhism itself is arguably just as guilty, as are most religions (Repetti 2016). Another possibility is that the increase in agency is empowering, but only if practitioners realize the difference between practicing to dissolve the agent-self and to cultivate it. Becoming Nobody is a film (Catto 2019) about Ram Dass’s life journey toward the spiritual goal of transcending ego. Whether “becoming nobody” makes one a better Buddhist, a more docile, productive consumer, or an agent-self virtuoso depends on the person, the practice, and the perspective. Just as Nozick (1981) argued that absent any supramundane metaphysics the meditative experience might be meaningless, the presence or absence of a no-self versus a yes-self theoretical framework informing meditative practice might make all the difference between the dissolution of the agent-self and the cultivation of agency and selfhood to the virtuoso level, respectively. It may very well be that belief in the yes-self view is a necessary condition for agency and selfhood, and that in this case choice of belief itself is a matter of free will.

References Brent, M. 2017. Agent-causation as a solution to the problem of action. Canadian Journal of Philosophy 47 (5): 656–673. ———. 2018. Confessions of a deluded Westerner. Journal of Buddhist Ethics 25: 687–713. Breyer, D. 2013. Freedom with a Buddhist face. Sophia 52 (2): 359–379. Caruso, G. 2018. Skepticism about moral responsibility. In Stanford encyclopedia of philosophy. https://plato.stanford.edu/archives/spr2018/entries/skepticism-­moral-­responsibility. Catto, J. 2019. Becoming nobody: everybody’s busy being somebody. Film. Love Serve Remember Foundation with Google Empathy Lab, Ojai. Cozort, D. 2017. Foreword. In Buddhist perspectives on free will: agentless agency? ed. R. Repetti, xvi–xvii. London: Routledge. Csikszentmihalyi, M. 1990. Flow: the psychology of optimal experience. New York: Harper & Row. Frankfurt, H. 1971. Freedom of the will and the concept of the person. Journal of Philosophy 68: 5–20. ———. 1988. Identification and externality. In The importance of what we care about, ed. H. Frankfurt, 58–68. Cambridge: Cambridge University Press. Garfield, J. 2015. Engaging Buddhism: why it matters to philosophy. New  York: Oxford University Press. Harris, S. 2013. Free will. New York: Free Press. Harvey, P. 2017. Psychological versus metaphysical agents: a Theravāda Buddhist view of free will and moral responsibility. In Buddhist perspectives on free will: agentless agency? ed. R. Repetti, 158–169. London: Routledge. Kushnir, T., A. Gopnik, N. Chemyak, et al. 2015. Developing intuitions about free will in children ages four and six. Cognition 138: 79–101. Martin, N., D.  Rigoni, and K.  Vohs. 2017. Free will beliefs predict attitudes toward unethical behavior and criminal punishment. Proceedings of the National Academy of Sciences 114 (28): 7325–7330.

Selfhood and Resentment

475

McRae, E. 2017. Emotions and choice: lessons from Tsongkhapa. In Buddhist perspectives on free will: agentless agency? ed. R. Repetti, 170–181. London: Routledge. Meyers, K. 2017. Grasping snakes: reflections on free will, samadhi, and dharmas. In Buddhist perspectives on free will: agentless agency? ed. R. Repetti, 182–192. London: Routledge. ———. 2018. Talking past each other? Reply to Rick Repetti. Journal of Buddhist Ethics 25: 961–987. Nozick, R. 1981. Philosophical explanations. Cambridge: Cambridge University Press. Pereboom, D. 2001. Living without free will. Cambridge: Cambridge University Press. Purser, R. 2019. McMindfulness: how mindfulness became the new capitalist spirituality. London: Repeater Books. Purser, R., and D. Loy. 2013. Beyond McMindfulness. Huff. Post, July 1, 2013. www.huffingtonpost.com/ron-­purser/beyond-­mcmindfulness_b_3519289.html. Accessed 1 Oct 2019. Purser, R., D. Forbes, and A. Burke, eds. 2016. Handbook of mindfulness: culture, context and social engagement. New York: Springer. Repetti, R. 2010a. The counterfactual theory of free will: a genuinely deterministic form of soft determinism. Saarbrücken: LAP Lambert Academic. ———. 2010b. Meditation and mental freedom: a Buddhist theory of free will. Journal of Buddhist Ethics 17: 165–212. ———. 2016. Meditation matters: replies to the anti-McMindfulness bandwagon! In Handbook of mindfulness: culture, context and social engagement, ed. R. Purser, D. Forbes, and A. Burke, 473–493. New York: Springer. ———., ed. 2017a. Buddhist perspectives on free will: agentless agency? London: Routledge. ———. 2017b. Why there ought to be a Buddhist theory of free will. In Buddhist perspectives on free will: agentless agency? ed. R. Repetti, 22–33. London: Routledge. ———. 2017c. Agentless agency: the soft compatibilist argument from Buddhist meditation, mind-mastery, evitabilism, and mental freedom. In Buddhist perspectives on free will: agentless agency? ed. R. Repetti, 193–206. London: Routledge. ———. 2019. Buddhism, meditation, and free will: a theory of mental freedom. London: Routledge. Siderits, M. 2017. Buddhist paleocompatibilism. In Buddhist perspectives on free will: agentless agency? ed. R. Repetti, 133–147. London: Routledge. Siderits, M., E. Thompson, and D. Zahavi, eds. 2013. Self, no self? Perspectives from analytical, phenomenological, and Indian traditions. New York: Oxford University Press. Strawson, P. 1962. Freedom and resentment. Proceedings of the British Academy 48: 1–25. Strawson, G. 2017. Free will and the sense of self. In Buddhist perspectives on free will: agentless agency? ed. R. Repetti, 72–83. London: Routledge.

Publications of Mark Siderits

1. Books: Monographs 1976 The Formlessness of the Good: Toward a Buddhist Theory of Value. Dissertation, Yale University. 1991 Indian Philosophy of Language: Studies in Selected Issues, Studies in Linguistics and Philosophy Monograph 46. Dordrecht: Kluwer. 2003 Personal Identity and Buddhist Philosophy: Empty Persons. Ashgate (revised 2nd edition, 2015). 2007 Buddhism as Philosophy: An Introduction. Hackett Pub. Co. (second edition, 2021). 2022 How Things Are: An Introduction to Buddhist Metaphysics. Oxford University Press. 2022 The Buddha’s Teachings As Philosophy. Hackett Pub. Co.

Edited Collections 2010 (with Evan Thompson and Dan Zahavi) Self, No Self? Perspectives From Analytical, Phenomenological, and Indian Traditions. Oxford University Press. 2011 (with Tom Tillemans and Arindam Chakrabarti) Apoha: Buddhist Nominalism and Human Cognition. Columbia University Press. 2011 (with Georges Dreyfus, Georges, Bronwyn Finnigan, Jay Garfield, Guy Newland, Graham Priest, Koji Tanaka, Sonam Thakchoe, Tom Tillemans and Jan Westerhoff) Moonshadows. Conventional Truth in Buddhist Philosophy. Oxford University Press. 2016 Studies in Buddhist Philosophy. Edited with an Introduction by Jan Westerhoff (and the assistance of Christopher V. Jones). Oxford University Press. 2020 (with Ching Keng and John Spackman) Buddhist Philosophy of Consciousness:Tradition and Dialogue. Brill: Rodopi.

© Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6

477

478

Publications of Mark Siderits

Translations 2013 (with Shoryu Katsura) Nāgārjuna’s Middle Way: The Mūlamadhyamakakārikā. Boston: Wisdom Publications.

2. Articles in referred journals and magazines 1976 Zeno and Nāgārjuna on motion (with J.D. O’Brien). Philosophy East and West 26:281–299. 1978 Why “Wittgenstein and Zen”? (in Japanese). Risu (August):75–92. 1979 A note on the early Buddhist theory of truth. Philosophy East and West 29 (4):491–499. 1980 The Madhyamaka critique of epistemology. I. Journal of Indian Philosophy 8 (4):307–335. 1981 The Madhyamaka critique of epistemology. II. Journal of Indian Philosophy 9 (2):121–160. 1982 More things in heaven and earth. Journal of Indian Philosophy 10 (2):187–208. 1985 Word meaning, sentence meaning, and apoha. Journal of Indian Philosophy 13 (2):133–151. 1985 Was Śāntarakṣita a “Positivist”? In Buddhist Logic and Epistemology, edited by B. K. Matilal. Dortrecht: D. Reidel, 184–197. 1986 The sense-reference distinction in Indian philosophy of language. Synthese 69 (1): 81–106. 1987 Beyond compatibilism: A Buddhist approach to freedom and determinism. American Philosophical Quarterly 24 (2):149–159. 1988 Ehring on Parfit’s relation R. Analysis 48 (January):29–32. 1988 Nāgārjuna as anti-realist. Journal of Indian Philosophy 16 (4):311–325. 1996 Do persons supervene on skandhas? Journal of Indian Philosophy and Religion 1:55–76. 1997 Buddhist reductionism. Philosophy East and West 47 (4):455–478. 1998 Relativism, objectivity and comparative philosophy. Journal of Indian Philosophy and Religion 3:1–15. 2000 Altruism and reality: An exchange (with Paul Williams). Philosophy East and West 50:412–459. 2000 Madhyamaka on naturalized epistemology. In Concepts of Knowledge East and West. Calcutta: Ramakrishna Mission Institute of Culture, 262–276. 2001 Buddhism and techno-physicalism: Is the eightfold path a program? Philosophy East and West 51 (3):307–314. 2003 Deductive, inductive, both or neither? Journal of Indian Philosophy 31 (1/3):303–321. 2003 The soteriological significance of emptiness. Contemporary Buddhism 4 (1):9–23. 2004 Causation in early Madhyamaka. Journal of Indian Philosophy 32 (4):393–419. 2004 Perceiving particulars: A Buddhist defense. Philosophy East and West 54 (3):367–382. 2005 Freedom, caring and Buddhist philosophy. Contemporary Buddhism 6 (2):87–116. 2006 (with Katsura, Shoryu) Mūlarnadhamakakãrika I–X. Journal of Indian and Tibetan Studies (Indogakuchibettogaku Kenkyi) 9 (10):129–184. 2008 Contradiction in Buddhist argumentation. Argumentation 22 (1):125–133. 2008 Paleo-compatibilism and Buddhist reductionism. Sophia 47 (1):29–42.

Publications of Mark Siderits

479

2011 (with Katsura, Shoryu and Yoshimizu, Kiyotaka) editors’ preface. Journal of Indian Philosophy 39 (4–5):351–352. 2012 Note to self. The Philosophers’ Magazine 56 (56):104–105. 2013 Buddhist paleocompatibilism. Philosophy East and West 63 (1):73–87. 2013 (with Garfield, Jay). Defending the semantic interpretation: A reply to Ferraro. Journal of Indian Philosophy 41 (6):655–664. 2013 Determinism, responsibility, and Asian philosophy. Philosophy East and West 63 (1):1–3. 2014 Causation, ‘Humean’ causation and emptiness. Journal of Indian Philosophy 42 (4):433–449. 2016 Response to Levine. Journal of World Philosophies 1 (1):128–130. 2019 The prapañca paradox. Journal of Indian Philosophy 47 (4):645–659. 2020 Buddhist non-conceptualism: Building a smart border wall. Philosophy East and West 70 (3):615–637. 2021 Born believer? Comparative Philosophy 12 (1).

3. Chapters in edited volumes and conference proceedings 1985 The Prabhākara Mīmāṃsa Theory of Related Designation. In Analytical Philosophy in Comparative Perspective, edited by B. K. Matilal and J. L. Shaw. D. Reidel, 253–297. 1989 Thinking on Empty: Madhyamaka Anti–realism and Canons of Rationality. In Rationality in Question: On Eastern and Western Views of Rationality, edited by S. Biderman and B.-A. Scharfstein. Leiden: Brill, 231–249. 1994 Matilal on Nāgārjuna. In Relativism, Suffering and Beyond: Essays in Memory of Bimal Krishna Matilal, edited by P. Bilimoria and J. N. Mohanty. Delhi: Oxford University Press, 69–92. 1997 Distinguishing the Mādhyamika from the Advaitin: A Field Guide. In Essays in Indian Philosophy, edited by S.R. Saha. Calcutta: Jadavpur University Press, 69–92. 1999 Apohavada, Nominalism and Resemblance Theories. In Proceedings of the Third International Dharmaktrti Conference, edited by Shoryu Katsura. Vienna: Verlag der Osterreichische Akademie der Wissenschaften, 341–348. 2000 Nyaya Realism, Buddhist Critique. In The Empirical and the Transcendental, edited by Bina Gupta. Rowman and Littlefield, 219–231. 2004 Buddhist Reductionism and Error Theory. In Proceedings of the Second Tokyo Conference on Argumentation. Tokyo: Japan Debate Association. 2006 Buddhist Reductionism and the Structure of Buddhist Ethics. In Indian Ethics: Classical and Contemporary Challenges, edited by P. Bilimoria, J. Prabhu and R. Sharma. Aldershot: Ashgate. 2006 Buddhist Nominalism and Desert Ornithology. In Universals, Concepts and Qualities: New Essays on the Meanings of Predicates, edited by Arindam Chakrabarti and Peter Strawson. Aldershot: Ashgate, 91–103. 2006 Apohavāda. In Philosophical Concepts Relevant to Sciences in Indian Tradition, edited by P.K. Sen. Vol. III, History of Indian Science, Philosophy and Culture, D.P. Chattopadhyaya, General Editor. Delhi: Motilal Banarsidass. 2009 Is Reductionism Expressible? In Pointing at the Moon: Buddhism, Logic, Analytic Philosophy, edited by Mario D’Amato, Jay L. Garfield and Tom J. F. Tillemans. Oxford University Press, 57–69.

480

Publications of Mark Siderits

2010 Is Everything Connected to Everything Else? What the Gopis Know. In Moonshadows: Conventional Truth in Buddhist Philosophy, edited by J. Garfield. Oxford University Press, 167–180. 2010 Buddhas as Zombies: A Buddhist Reduction of Subjectivity. In Self, No Self?: Perspectives From Analytical, Phenomenological, and Indian Traditions, edited by Mark Siderits, Evan Thompson and Dan Zahavi. Oxford University Press. 2020 Some Sceptical Doubts About “Buddhist Scepticism”. In Buddhism and Scepticism: Historical, Philosophical, and Comparative Perspectives, edited by Oren Hanner. Freiburg: Bochum, 21–35.

4. Encyclopedia articles 1997 Nāgārjuna. Routledge Encyclopedia of Philosophy. London: Routledge. 2005 Madhyamaka Buddhism. Encyclopedia of Philosophy (2nd edition). Farmington Hills: Macmillan Reference. 2019 Siderits, Mark, Buddha. The Stanford Encyclopedia of Philosophy, edited by Edward N. Zalta (Spring 2019 edition), https://plato.stanford.edu/archives/spr2019/entries/ buddha/.

5. Anthologized articles 2021 Non-Self, Empty Persons. In Philosophy of Mind: Classical and Contemporary Readings, edited by David Chalmers (2nd edition). Oxford University Press.

6. Book ­reviews 1979 Kemeth K. S. Ch’en, The Chinese Transformation of Buddhism. Journal of Chinese Philosophy 6 (1):111–113. 1980 David J. Kalupahana, Causality: The Central Philosophy of Buddhism. Journal of Indian Philosophy 8:191. 1998 Arindam Chakrabarti, Epistemology, Meaning and Metaphysics After Matilal. Philosophy East and West 48 (3):503–513. 2000 The Reality of Altruism: Reconstructing Śāntideva. Review of Altruism and Reality: Studies in the Philosophy of the Bodhicāryāvatara by Paul Williams. Philosophy East and West 50 (3):412–424. 2005 William S. Waldron, The Buddhist Unconscious: The Ālaya-vijñāna in the Context of Indian Buddhist Thought. Philosophy East and West 55 (2):358–363. 2009 Chakravarthi Ram-Prasad, Indian Philosophy and the Consequences of Knowledge: Themes in Ethics, Metaphysics and Soteriology. Ars Disputandi 9. 2010 Jan Westerhoff, Nāgārjuna’s Madhyamaka: A Philosophical Introduction. Mind 119 (475):864–867. 2013 Dan Arnold, Brains, Buddhas, and Believing: The Problem of Intentionality in Classical Buddhist and Cognitive-Scientific Philosophy of Mind. International Journal for Philosophy of Religion 74 (2):237–241.

Publications of Mark Siderits

481

2015 Stephen Cross, Schopenhauer’s Encounter with Indian Thought: Representation and Will and Their Indian Parallels. Journal of the Indian Council of Philosophical Research 32 (2):273–278. 2018 Joerg Tuske, Indian Epistemology and Metaphysics. Dialectica 72 (3):479–484. 2018 The Nyāya-sūtra: Selections with Early Commentaries. Translated by Matthew Dasti and Stephen Phillips. Philosophy East and West 68 (3):1–3. 2021 Arindam Chakrabarti, Realisms Interlinked: Objects, Subjects, and Other Subjects. Philosophical Review 130 (3):467–471. 2021 The Bloomsbury Research Handbook of Indian Philosophy of Language, edited by Alessandro Graheli. Philosophy East and West 71 (1):1–5.

Index

A Abductive inference, 110 Abhidharma, 6, 9, 13, 20–22, 27, 30, 40, 43, 52, 54, 99, 105–107, 109–111, 113, 114, 116–119, 121, 166–175, 177, 180, 181, 186, 191, 192, 231–233, 235, 241, 260, 264, 267–274, 300, 345–349, 351, 352, 354, 356, 359, 360, 396, 404, 455 Abhidharmakośa (AK), 92, 170, 174, 175, 233, 299, 300 Abhidharmakośabhāṣya (AKBh), 43, 46, 47, 52, 87, 92, 94, 95, 102, 111–113, 115, 117, 170, 174, 175, 219, 339, 346, 347, 351, 352, 359, 395, 401 Ābhidharmikas, 111, 117, 167, 169, 171, 173–175, 177, 178, 181, 187, 350, 352 Abhinavagupta, 200 Absence, 2, 37, 70, 72, 104, 131, 147, 157, 171, 191, 196, 201, 207, 221, 234, 237, 262, 283, 284, 292, 331, 334, 335, 337–339, 423, 426, 446, 448, 459, 473, 474 Absolute nothingness, 3 Acquaintance, 54 Actuality, 74, 219, 290, 291 Advaita Vedānta, 8, 255–256 Affect-biased attention, 180, 205 Affective bias, 197, 204 Afflicted mind, see kliṣṭa-manas Afflictive emotions, 401 Agency, 4, 13, 36, 50–57, 112, 118, 122, 157, 195, 277, 396, 397, 406, 433, 459–474 Agent, 5, 22, 39, 46, 50, 51, 56, 90, 92, 94, 112, 113, 117, 122, 136, 146, 153, 172,

218, 237, 244, 245, 277, 350, 358, 391, 392, 395–397, 399, 400, 406, 427, 440–442, 445, 453, 460, 461, 463–465, 468–470, 472 Agent causation, 5, 50, 51 Agentless action, 462 Agent-neutrality, 439–441, 443 Aggregates, 13, 27, 46, 47, 49–53, 57, 92–95, 105, 106, 109, 112, 114–116, 118, 126, 128, 219, 268–270, 273, 277, 317, 349, 353, 395, 431, 454, 455, 467, 468 ahaṃkāra, 454, 455 ālambana, 173, 218 See also Perceptual object ālayavijñāna, 74, 80, 81, 480 ākāra, 126, 172, 173, 186, 187, 190, 192 See also Aspect; Forms; Mental image; Representation Albahari, M., 53 Anālayo, Bhikkhu, 141, 142 Anapanasati, 24 anātman, 125–127, 130, 136, 446, 447 See also Non-self; No-self Aṅguttara Nikāya, 466 Annas, Julia, 391 Annihilationism, 292, 469 Anscombe, G.E.M., 51, 190, 390 antarābhava, 137 Anti-realism, 4, 7, 8, 13, 161, 209, 235, 237, 241, 261, 262, 276, 283–293, 300, 479 Antirealist stance, 53 Anti-reflexivity principle, 22 anumāna, 327–342 Anusmarana-vikalpa, 174, 177, 191

© Springer Nature Switzerland AG 2023 C. Coseru (ed.), Reasons and Empty Persons: Mind, Metaphysics, and Morality, Sophia Studies in Cross-cultural Philosophy of Traditions and Cultures 36, https://doi.org/10.1007/978-3-031-13995-6

483

484 Anvaya-vyatirekin, 334 Aśokāvadāna, 130, 132–136 Apoha, 12, 165, 179, 181–192, 196, 197, 200, 201, 205, 208–210, 477, 478 See also Nominalism apohavāda, 183, 479 Apperception, 52, 400 Appropriation, 218, 270, 273, 401 See also upādāna aretē, 422 Argumentum ad ignorantiam, 9 Arhat, 129, 132, 133, 388, 407, 440 Aristotle, 55, 70, 99, 247, 283, 320, 390, 392, 393, 395, 396, 398, 399, 422, 423, 425, 464 Arnold, Dan, 13, 142, 143, 192, 259–279, 480 ārya-śrāvaka, 287 asaṃjñika-sattvāḥ, 123 Asaṅga, 54, 198, 405 Aspect, 4, 5, 7, 11, 23, 24, 28, 29, 31–33, 50, 51, 53, 54, 57, 79, 110, 168, 169, 172, 184, 186, 202, 203, 208, 223, 242, 244, 304, 331, 342, 351, 372, 388, 389, 391, 419, 443, 448 intentional, 54 subjective, 29, 32, 52, 54, 55 See also ākāra ātman, 11, 117, 118, 125, 329, 331, 332, 339, 469 Atomism, 426, 427 attakāmo, 88 Attention, 2, 10, 19, 31, 40, 52, 53, 63, 70, 88, 91, 109–112, 126, 132, 135, 168, 180, 192, 195, 202–206, 208, 218, 225, 244, 304, 317, 339, 355, 364, 388, 399, 401, 402, 405, 407, 410, 418, 423, 427, 430, 431, 445, 449, 452, 461, 462, 465, 467, 473 Attentional inhibition, 202, 205 Attentional orienting, 180 See also manaskāra Attentional selection, 202, 205 Augustine, 471 Austin, J.L., 100 Autonomous agency, 397, 472 Autonomous self, 122 Averroës, 79 avyākṛta, 365 Awareness, 14, 20, 24–27, 30–32, 51, 54–57, 80–83, 88, 89, 99, 104, 124, 128, 154, 167–175, 177, 178, 180, 181, 183, 186, 189, 191, 197–207, 255, 266, 268, 291, 396, 423, 448, 463, 472

Index nonjudgmental, 24, 25 subsequent, 172, 199, 201, 202, 206 Ayer, A.J., 308 B Barresi, John, 78 Belief, 8, 9, 26, 41, 44, 83, 88, 89, 101, 107, 117, 122, 124, 128, 136, 141, 142, 152–157, 162, 214, 236, 237, 270, 329, 359, 370, 378, 379, 401, 411, 425, 429, 430, 447, 448, 453, 460, 469, 472–474 Belief-independence, 101 Berkeley, George, 67, 71, 89, 92, 170, 199 Bermúdez, José Luis J, 56, 194 Bhāsarvajna, 283, 291 bhāva, 270, 287, 377, 378, 380 bhavaṅga-citta, 14 Bhāviveka, 252, 253, 367 bhūtatā, 219 bhūyo-darśana, 333 Binding problem, 126 Blind-sight, 23, 111 Block, Ned, 124, 176 Bodhicaryāvatāra, 394, 431, 437, 443, 453, 480 Bodhisattva, 14, 199, 388, 405–408, 439, 441–444, 470, 471 Boltzmann brains, 158–161 Boltzmann successor, 158–161 Book of Changes, 301 Boundary, organism/environment, 466 Boundary, self-other, 464, 466 Brahman, 255, 256 Brahmasūtrabhāṣya, 255 Brains-in-vats, 123 Brain-transplant, 137 Brentano, Franz, 54, 55 Brumbaugh, Robert S., 3 Buddha, 3, 4, 10, 32, 35, 42, 57, 58, 63–85, 87, 88, 90–95, 100, 115, 116, 122–124, 135, 142, 168–170, 197–199, 236, 248, 250, 262, 266, 283, 286, 287, 295, 299, 303, 304, 338, 356, 359, 398, 400, 401, 404–407, 429–430, 432, 460, 463–469, 480 Buddhaghosa, 43, 91, 141, 178, 388, 399, 404, 405, 407, 428, 431 Buddha-nature, 123, 124 Buddhist exceptionalism, 26 Buddhist nominalism, 165, 181, 209, 477, 479 Buddhist Personalism, 13, 46, 49–51, 57, 270 See also Pudgalavāda

Index Buddhist Personalists, 46–48, 51, 52, 56, 57 See also Pudgalavādins Buddhist Reductionism, 7, 13, 35, 36, 42–51, 53, 57, 109, 111, 116, 117, 231–236, 245, 345–360, 478, 479 Bundle theory, 68 Burton, David, 275, 276, 377, 431, 432 C Campbell, John, 91 Candrakīrti, 94, 96, 236, 240, 252–254, 261, 265–278, 363, 367 Cantor, Georg, 315, 317–320 Caraka-saṃhitā, 328, 329, 332 Cardona, George, 338, 340 Carnap, Rudolf, 36 Carpenter, Amber, 14, 51, 53, 178, 272, 273, 300, 387–411 Cartesian Ego, 37, 39, 57, 450 Cartesianism, 311 Cartesian theater, 21, 31 Cārvakā, 19, 469 See also Materialism; Physicalism Cassam, Quassim, 54, 91 Categories, 36, 41, 45, 76, 101, 108, 176–180, 183, 184, 187, 192, 207–210, 213, 237, 261, 270–272, 290–292, 310, 312, 388, 389, 391, 395, 399–404, 406, 408, 418, 429, 431, 432, 452 Category mistake, 310, 312 catuṣkoṭi, 249, 250, 295, 296, 303, 363–365, 367, 373, 375, 380–382 Causal continua, 219 Causal determination, 41, 42 Causal efficacy, 111, 112, 172, 186, 187, 213–217, 219, 221, 222, 225, 289, 469 Causal-efficacy argument, 105, 114 Causal nihilism, 216, 222, 224 Causal-ontological principle, 468, 473 Causal overdetermination, 48, 49 Causal relations, 29, 30, 42, 52, 112, 159, 160, 172, 211, 212, 214, 216, 217, 219, 222, 226, 235, 273, 286, 453, 463 Causal theory, 9, 213 Causation, 5, 9, 12, 49–51, 68, 92, 125, 144, 171, 172, 211–226, 235, 338, 367, 468, 478, 479 Causation, downward, 49 Cause, 41, 45, 49, 52, 87–89, 104, 113, 172, 188, 190, 214–217, 220, 236, 244, 338, 339, 399, 410, 453, 471

485 Cessation, 25, 85, 93, 142, 156, 160, 213, 214, 219–222, 224–226, 279, 288, 292, 338, 404, 429 cetanā, 109–111, 180, 396 Chadha, Monima, 13, 105–138 Chakrabarti, Arindam, 2, 12, 13, 165, 181, 191, 210, 283–293, 477, 479–481 Chalmers, David, 19, 21, 135, 176, 451, 480 Chariot analogy, 111 Charmides, 398 Chatterjee, Amita, 208, 209 Chinese Buddhism, 296, 301, 302 Chāndogya Upaniṣad, 255 Chomsky, Noam, 309, 313, 323 citta, 80, 109 citta-santāna, 14, 80 Clarke, Desmond, 64, 69, 75 Clark, Maudemarie, 97 Clayton, Barbara, 391, 392, 441, 443 Clement of Alexandria, 90 cogitatio, 64, 66, 67 Cognition, 4, 8–13, 22, 23, 27, 31, 55, 98, 103, 104, 112, 121–123, 166–192, 194–202, 206, 208, 209, 233, 244, 287, 291, 327, 332, 333, 339, 350, 352, 358, 359, 398, 409, 410, 425 Cognitive closure, 308 Cognitive illusion, 154 Cognitive science, 13, 23, 26, 121, 122, 125, 165–210, 433 Coherentism, 245 Collins, Steven, 40, 43, 53, 94 Combination problem, 55 Comparative philosophy, 417, 478, 479 Compassion, 3, 10, 22, 76, 118, 121, 398, 403, 411, 429, 431–433, 448, 450, 460 Compatibilism, 6, 15, 478 Conceivability principle, 47, 51 Concept, demonstrative, 195–197 Conceptual construction, 107, 114, 147, 185, 200, 275 Conceptual dependence, 100, 237, 244 Conceptualization, 23, 25, 109, 166, 173, 174, 179, 185, 194, 196, 199–201, 204–206, 210, 253, 376 Confluence approach, 1, 5–13 Confluence philosophy, 1–14, 418 Confucianism, 432 Conscious episode, 31, 55, 57 Conscious experience, structural dynamics of, 42

486 Consciousness, 9, 19, 36, 66, 95, 122, 146, 170, 225, 276, 283, 311, 431, 461 double aspect view of, 29 nature of, 19, 32, 55, 124, 125 pre-reflectively self-aware, 31 reflexive, 20, 22, 30, 31, 53, 55–57 Consciousness-process, 67–68, 71–74, 84, 85 Consciousness-stream, 14 Consequentialism, 161, 391, 392, 428, 437, 439–441, 445 Constructing consciousness, 146 Constructional hierarchy, 145 Constructions, 92, 107, 114, 143–149, 166, 173–175, 177, 179, 183–185, 190, 197, 198, 200, 201, 275, 373, 402, 407 Constructivism, Goodmanian, 143, 145, 149, 161 Containers, 145, 295–304 Continuity, 10, 13, 38–41, 57, 58, 89, 91, 125, 137, 138, 141–162, 187, 450 relation, 41, 143, 450 spatio-temporal, 40 See also Psychological connectedness; Psychological continuity Contradiction, 250–252, 254, 260, 261, 263, 264, 289, 295, 297, 303, 304, 317, 323, 345, 350, 357, 365–367, 376, 380, 469, 478 Conventionalism, 240, 241, 245 Conventional truth, 101, 151, 211, 213, 219, 233, 234, 236, 238, 240, 241, 248, 253, 259, 261, 262, 264–267, 273, 346, 349, 351–353, 355, 357–360, 364, 371, 372, 376, 379, 381, 382, 463, 477, 480 Copernicus, Nicolaus, 8 Corpse-man, 124, 128, 130–138 Correspondence theory, 240, 318 Coseru, Christian, 1–14, 35–58, 85, 126, 137, 300 Cosmology, Buddhist, 19, 27 Cox, Collett, 100, 107, 346 Crane, Tim, 49, 196 Cross-modal binding, 125 Csikszentmihalyi, Mihaly, 462 Cultural alterity, 123 D D’Amato, Mario, 348, 479 dāna-pāramitā, 127 Daodejing, 301 Dasein, 261, 273 Dasti, Matthew, 166, 330, 481

Index Dawkins, Richard, 149, 150, 469 Da zhidu lun, 127–138 Death, 37, 68, 90, 93, 117, 123, 137, 141–162, 220, 283, 285, 365, 366, 404, 426, 427, 429, 440 Deflationism, 20–26, 240 Deguchi, Yasuo, 253, 254 de la Vallée Poussin, Louis, 52, 92, 100, 170, 286 Dennett, Daniel, 20, 21, 24, 31, 124, 137, 145, 176 Deontology, 161, 440 Dependence, mereological, 237, 238 Dependent origination, 27, 114, 117, 213–215, 218–226, 236, 243, 248, 263, 270, 338, 382, 430, 463 Dereification, 204, 205 Descartes, René, 4, 14, 19, 55, 63–85, 123, 310, 311, 423 de se judgement, 88 Determinate perception, 37 See also savikalpaka pratyakṣa Determinism, 4, 124, 159, 460, 468, 473, 478, 479 Devasarman, 46 Devoid of conceptualization/mental construction, see kalpanāpoḍha Dhammajoti, Bhikku, K.L., 169–174 Dhammapāda, 390, 393 Dharma, 12, 98, 107, 109, 110, 132, 171, 172, 236, 243, 248, 286, 398, 404–406, 439 Dharmakīrti, 13, 19, 27, 30, 98, 101, 109, 110, 143, 166, 175, 179, 182–192, 196, 200, 201, 204, 208, 209, 245, 252, 266, 328, 342 dharmatā, 199, 219 Diachronic unity, 125 Dialetheism, 13, 251, 252, 295 Differentialist, 287, 288 Dignāga, Diṅnāga, 8, 12, 30, 53–55, 58, 99, 110, 166, 170, 171, 181–183, 252, 299, 328, 330–332, 334–337, 339–342 Diogenes Laertius, 426 Diotima, 399 Direct perception, 172, 173, 207, 211 See also savikalpaka pratyakṣa Disposition, 24, 50–52, 189, 208, 222, 253, 255, 313, 339, 393, 395, 396, 401, 420, 424, 431, 447, 473 Divyāvadāna, 130, 134 Dōgen, 303, 304 Domain principle, 297–303 Donnellan, Keith, 371

Index Double negation elimination, 365 Doyle, Sir Arthur Conan, 356, 357 dravya-sat, foundational, 95, 96, 98, 267 dravyataḥ, 94, 346 Dreyfus, Georges, 13, 19–33, 126, 179, 183, 185, 187, 192, 195, 266, 403, 431, 477 dṛṣṭānta, 329, 331 Dualism, 19, 20, 27, 49, 63, 64, 124, 142, 199, 200, 209, 210, 390, 448, 453, 467 Dualism, nature of, 124 Duerlinger, James, 40, 47, 92, 94–96, 268, 347 dukkha/duḥkha, 299, 429 See also Suffering Dummett, Michael, 89, 283, 284, 291, 308 Dunne, John, 98, 166, 179, 184–187, 189–191, 197, 199, 201, 202, 204, 205, 245 Dynamic model, 52 E Ego, 11, 39, 56, 470, 474 See also Cartesian Ego Egoism, 89, 445, 448, 460 Egological, 53, 54 Eliminative reductionism, 37 Eliminativism, 21, 48, 108, 116, 119, 260, 374 Eliminativist naturalism, 142 Eltschinger, Vincent, 117, 182 Embodied, 9–11, 36, 44, 121, 122, 143, 153, 168, 202, 208, 284, 311, 312, 419, 461, 462, 465, 466 Embodied cognition, 121, 122, 208 Emergence, 89, 396, 431, 432, 461 Emergentism, 49 Emergent phenomenon, 122 Emotion, 37, 75, 90, 180, 195, 202, 203, 208, 307, 401, 402, 406, 411, 424, 431, 438, 445, 448, 462, 466, 467, 473 Empathy, 28 Empiricism, 9, 426 Emptiness, 3, 6, 7, 9, 10, 32, 91, 117, 145, 151, 212, 214, 219, 220, 223, 225, 231, 235–237, 244, 248, 255, 263, 296–299, 303, 304, 323, 363, 364, 373, 375–382, 400, 430, 466, 472, 478, 479 Enactive, 121, 122, 168, 207–210 Enlightenment, 6, 10, 25, 88, 262, 264, 393, 395, 405, 428, 430, 431, 433, 460, 462–464, 470–472 Entities non-material, 27 partite, 43, 44 Epicurus, 425–429, 432

487 Episodic memory, 41 Epistemic instruments, 3 See also pramāṇa Epistemological pessimism, 154 Erroneous, 101, 107, 185, 186, 197, 198, 209, 210, 358, 454 Essential nature, 397 Eternalism, 95, 224, 226, 286, 288, 469 eudaimonia, 389, 391, 396 Evan, Gareth, 101, 193, 311 Event causation, 5 Everest, 9 Evidence, 5, 7, 12, 50, 75, 90, 129–131, 134–136, 153, 156, 157, 180, 202, 205, 252, 254, 284, 295, 327, 328, 331, 341, 375, 377, 380, 381, 433, 444, 450–452, 466 Evil demon, 123 Excellence, 261, 392, 393, 397, 422 See also aretē Existent/existence, 4, 24, 36, 64, 91, 105, 136, 141, 171, 220, 231, 248, 264, 284, 297, 321, 331, 346, 364, 388, 421, 449, 464 conventionally real, 260, 267, 277 (see also samvṛti-sat) ultimately real, 263, 264, 267, 270, 272, 273, 278 (see also prajñapti-sat) Experience, 6, 19, 37, 66, 90, 109, 121, 143, 165, 223, 239, 262, 290, 321, 333, 347, 394, 422, 448, 459 Expressibility, 345–360 Externalism, 4, 8, 237, 358 Extreme Claim, 24, 450, 451 Extrinsic nature, 215–219, 223, 226, 235, 377 F Fallibilism, 232 Fictionalism, 345–360 First-person, 47, 51, 111, 125, 126, 153, 156, 157, 194, 438 experience, 126 perspective, 26, 28–30, 46, 125, 152, 156, 157, 194 First-personal self-givenness, 125 Flanagan, Owen, 50, 143, 162, 301, 433 Flourishing, 390, 397 See also eudaimonia Foot, Philippa, 390 For-me-ness, 56 Forms, 4, 32, 40, 63, 98, 109, 127, 142, 169, 214, 235, 250, 268, 284, 296, 309, 328, 346, 363, 391, 417, 437, 460

488 Foundationalism, 8, 239, 245, 265 Franco, Eli, 142 Frankfurt, Harry, 207, 461, 467 Frauwallner, Erich, 340, 411 Freedom, 4, 5, 52, 87, 214, 388, 427, 466–469, 471, 473, 478 Freedom of the will, 467 Free will, 51, 122, 124, 396, 459–461, 467, 468, 471–474 Free will, embodied, 461 Frege, Gottlob, 89, 290, 309–311 Freud, Sigmund, 152 Functionalism, 142 Further fact view, 40, 50, 56 Fusion philosophy, 167, 168, 245, 247, 417–433 G Ganeri, Jonardon, 13, 40, 53, 87–104, 110, 118, 126, 130, 137, 168, 169, 233, 273, 347, 423 Garfield, Jay L., 13, 165, 167, 200, 206, 244, 247–257, 265, 267, 296–300, 303, 348, 363–365, 371, 374–378, 382, 393, 431, 477, 479, 480 Garrett, Brian 51 Geach, Peter, 290, 291 Gedankenexperiment, 123 Geertz, Clifford, 275, 278 German idealism, 3 Givenness, 56, 125 Global availability, 20, 22, 110 God, 64–66, 73, 89, 123, 124, 141, 250, 309, 311, 329, 403, 410, 426, 427, 453, 468, 471 Gold, Jonathan, 111, 112, 114, 199, 289 Gombrich, Richard, 88, 93, 393 Goodman, Charles, 169, 345, 348, 352–355, 391, 392, 437–455 Goodman, Nelson, 13, 14, 143–149 Gorgias, 389, 402 Gowans, Christopher W. 14, 417–433 grāhaka[ākāra], 167, 198, 199 See also Grasper, Subjective aspect grāhya[ākāra], 167, 198, 199 See also Grasped, Objective aspect Grammarians, 113, 271, 277, 338 Grasped, 167, 170, 172, 184, 199–201, 204, 262, 264, 332, 398 Grasper, 167, 198–201, 204–206 Greene, Josh, 445 Greiffenhagen, Christian, 14, 307–323

Index Griffiths, Paul J., 166, 197, 198 Grounding, 9, 53, 55, 162, 232, 242, 272, 276, 347, 353 Guerrero, Laura, 13, 245, 345–360 H Habitual tendencies, 52 Hadot, Pierre, 404, 426 Hallisey, Charles, 389, 394, 406 Hansen, Anne, 389, 403, 406 Hard problem, of consciousness, 19–33 Harivarman, 46 Harries, Karsten, 3 Harris, Stephen, 438, 440, 443, 446–449, 460, 461 Harvey, P., 53, 389, 401, 463 Hattori, Masaaki, 54, 171, 181, 183 Heart-mind, 420, 424 Heart Sūtra, 255 Hegel, G.W.F., 195, 252 Heidegger, Martin, 3, 54, 261, 266, 273, 310, 311 Heim, Maria, 389, 393, 405, 406, 431 hetu, 12, 328–330, 332 See also Cause Hilbert, David, 317 hiri, 110, 111 Hobbes, Thomas, 64, 67, 78 Hoffman, Donald, 149, 150 Horst, Steven, 242–244 Hough, Sheridan, v, 54 Huayan, 295, 301, 302 Humean causation, 212 Hume, David, 56, 64, 67, 68, 73, 90, 93, 96, 104, 212, 225 Huntington, C.W., 101, 276 Husserl, Edmund, 27, 55, 125, 321 Hypnosis, 24 I Idealism ethical, 408 German, 3 impersonal, 387, 408 Identification, type-, 175, 183, 185, 187, 188, 196 Identity, 35–37, 40–42, 45, 47, 49, 56, 57, 70, 77, 78, 84, 90–92, 97, 118, 127, 132, 137, 296, 301, 302, 342, 345, 355, 356, 382, 390, 450, 455 See also Personal identity

Index Identity conditions, 35 Illocutionary negation, 367, 372, 373 Illusion, 21, 89, 96, 97, 100, 101, 122, 125, 137, 154, 156, 200, 201, 255, 278, 285, 462, 466 Illusionism, about consciousness, 21 Imagination, 66, 75, 83, 123, 155, 166, 173, 197, 198, 358, 359 I-making, 53, 54 Immunity to error through misidentification, 126 Impartial benevolence, 438, 439, 443 Impersonal description thesis, 38, 39, 91 Inada, Kenneth, 3 Inclosure schema, 297–300, 303 Incommensurability, 124, 308 Independent existence, 4, 13, 364, 376, 380, 381 Indeterminate perception (nirvikalpaka pratyakṣa), 469 See also Indirect perception, nirvikalpaka pratyakṣa Indexical, 53, 55 Indirect perception, nirvikalpaka pratyakṣa, 27 Individuality, 52, 90, 93 Infinite, 14, 65, 144, 148, 187, 249, 302, 307–323, 340, 373, 428 Infinite regress, 148, 189, 217, 254, 373 Infinity, 14, 145, 308, 309, 311–323 Infinity, mathematical, 316–319, 321, 323 Innate ideas, 72, 75, 423 Insentient object, 53, 123 Insubstantiality, of a person, 157 Intelligibility, 11, 269, 272, 276, 279, 308, 351, 388, 409 Intention, 29, 30, 41, 51, 109–111, 113, 160, 180, 192, 284, 323, 366, 393, 394, 396, 406, 430, 431, 461, 462, 465 Intentional action, 52, 53, 396 Intentional aspect, 54 Intentional content, 52 Intentional descriptions, 22 Intentionality, 54, 57, 167, 192, 200, 202, 358, 463, 480 Intentionality, subject-object, 200, 202 Interface theory, 149–158 Intermediate existence, 137 Internalism, 4, 8, 448, 449 Intersubjective, 54, 158, 353 Intertheoretic relations, 36 Intertraditional comparisons, 63 Intrinsic nature, 13, 83, 211–220, 223–226, 235, 237, 241, 264, 346, 364, 373, 377, 380

489 Intrinsic properties, 48, 49, 182, 212 Introspective awareness, 54 Intuition, 2, 21, 37, 56, 124, 137, 138, 155, 158, 168, 194, 196, 197, 207, 262, 265, 266, 269, 273, 279, 290, 395, 423, 447, 455 Inwood, Brad, 426, 427 Ipseity, 125 Irrealism, 149 I-sensing, 53 Ivanhoe, Philip J., 301, 419 J Jackson, Frank, 124 Jayanta Bhaṭṭa, 104 Jesus, 471 Jinglu yixiang, 138 jñāna, 165 See also Cognition Jñānaśrīmitra, 342 Johnston, Mark, 270, 379, 438, 449, 450, 452, 455 Judgment, 21–25, 51, 179, 274, 421, 460, 473 junzi, 420, 424 Justification, 7–9, 153, 162, 195, 213, 263, 270, 428, 438, 459–460 K Kagan, Shelly, 438 kalpanā, 166, 184, 197, 198 See also Conceptual construction kalpanāpoḍha, 166 Kamalaśīla, 213, 219, 221, 288 Kant, Immanuel, 8, 68, 194–198, 207, 392, 423 Kapstein, Matthew T., 92, 96, 102, 113, 115, 126, 264, 267, 268, 444 karma, 19, 66, 92, 112, 115, 118, 128, 160, 162, 214, 393, 395, 398, 439, 442, 443, 468–471 Karmic conditionality, 160, 162 karuṇā, 3, 393, 405 katastematic (static) pleasure, 426, 427 Kathāvatthu, 9, 46 Katsura, Shōryū, 3, 12, 190, 231, 236, 237, 249, 263, 277, 279, 283, 327–342, 364, 365, 478, 479 Kellner, Birgit, 172, 186 Keng, Ching, 11, 477 Keown, Damien, 391, 395, 417 Keśavamiśra, 333–337 khandha, 93, 94

490 King Milinda, 43, 50, 106, 346, 464 kliṣṭa-manas, 54, 126 See also Afflicted mind Kriegel, Uriah, 56 Kripke, Saul, 270 Kuhn, Thomas, 308, 322 Kumārajīva, 130, 131, 138 L Lamotte, Étienne, 54, 127, 130 Laurence, Stephen, 167, 179, 184, 193, 196 Leary, Mark, 89 Lebenswelt, 321 Lewis, David, 107, 145, 161 Li, 301, 302, 419 Libertarianism, 460 Lichtenberg, G.C., 63, 65, 83–85 Life-continuum mind, 14 Life-world, 321 See also Lebenswelt Limits of language, 4, 11 Locke, John, 4, 55, 56, 64, 67, 68, 72, 455 Lopez, Donald, 122, 123, 173 Lycan, William, 364, 369, 370, 373 M Machery, Edouard, 178, 179 MacIntyre, Alasdair, 390, 392, 396, 397 Mackenzie, Matthew, 13, 231–245 Madhyamaka, 1, 4, 5, 7, 8, 12–14, 20, 26–30, 32, 33, 43, 99, 101, 117, 143, 145, 197, 213, 231, 235–242, 244, 245, 247–257, 259–279, 295, 297, 300, 302, 363, 397, 455, 478–480 Madhyamakāvatāra, 101, 253, 266, 269–271, 276–278 Mādhyamika, 8, 10, 29, 101, 149, 211–226, 231, 232, 235–237, 242–244, 248, 251, 254, 455, 479 mahābhūta, 99 Mahāsīlava-jātaka, 135, 136 Mahāvibhāṣā, 346 Mahāyāna, 134, 250, 262, 274, 323, 354, 392, 399, 430, 431, 439, 440, 443, 444, 464, 471 MajjhimaNikāya (M), 169, 338, 404, 428 Mamertus, 67 Manas, 54, 126 See also Mind manaskāra, 180 mano-vijñāna, 126

Index See also Introspective awareness Margolis, Eric, 167, 179, 184, 193, 196 Martin, Ray, 78 Materialism, 19, 20, 161, 451, 452 Matilal, Bimal Krishna, 8, 96, 98, 99, 166, 198, 271, 277, 391, 478480 Matrix, 123, 202, 205 Matter, 2, 7, 19, 41, 42, 53, 55, 58, 65, 74, 75, 77–81, 88, 91, 96, 99, 100, 102, 108, 112, 124, 144, 145, 147, 148, 152, 153, 158–160, 165, 196, 206, 211–213, 222, 224, 225, 233, 234, 238, 249–252, 254–256, 265, 266, 272, 273, 275, 278, 279, 285, 287, 310, 341, 356, 357, 373, 390, 395, 402, 437–442, 447, 448, 450–452, 462, 463, 465, 466, 468, 473, 474 Matthen, Mohan, 209 McDowell, John, 195, 390, 393, 397 McMindfulness, 460, 473 McRae, Emily, 431, 462, 471 Meditation, 24, 26, 84, 111, 128, 167, 175, 202–206, 211, 304, 403–408, 423, 424, 433, 445–448, 459–462, 464, 466, 468–470, 472, 473 Meditation, open monitoring, 202, 204–206 Meditation virtuoso, 464, 468, 470, 472, 473 Meditative practice, 2, 20, 26, 32, 404, 407, 431, 472–474 Meillassoux, Quentin, 148 Memory, 41, 73, 75, 80, 102, 103, 111–115, 125, 127, 143, 149, 166, 172, 177, 178, 180, 191, 194, 196, 204, 287, 291–293, 423, 479 autobiographic, 204 episodic, 41 intellectual, 72, 75 long-term, 179 semantic, 41 Meno, 403, 409 Mental causation, 12, 51, 211–226 Mental construction, 166, 177, 179, 190, 197, 198, 201 Mental continuity, 13, 143, 150–152, 154–159, 162 Mental continuity, Buddhist theory of, 157, 159 Mental image, 178, 191, 197, 203, 218 Mental state, self-referential, 51, 53 Mental stream, 8, 57, 126, 137, 156 Mereological nihilism, 6, 44, 48, 49, 105–111, 116, 347

Index Mereological realism, 49, 349 Mereology, 6, 44, 48, 49, 51, 57, 105–111, 116, 148, 236–238, 261, 271, 273, 298, 347–350, 354 Merricks, Trenton, 48, 97 Metalinguistic analysis, 45 Metaphysics, 4, 5, 8, 11–14, 30, 43, 46, 47, 51, 52, 55, 64, 78, 80, 99, 103, 137, 169, 172, 181, 243, 284, 286, 295, 301, 323, 388, 395, 396, 400, 401, 408, 410, 421, 426, 446, 450, 471, 474, 477, 480, 481 descriptive, 103 realist perspective on, 29 revisionary, 396 Meta-volitional harmony, 467 Metta, 393, 405 Metzinger, Thomas, 10, 137, 146, 149, 194, 207 Meyers, Karin, 50, 395, 459, 463 Micropsychism, 55 Middle way, 95, 104, 215, 255, 264, 283, 478 Milindapañha, 43, 50, 346, 396 Mill, John Stuart, 392, 423, 442 Mind, 9, 20, 36, 63, 90, 107, 121, 141, 166, 211, 244, 255, 274, 291, 308, 346, 371, 393, 423, 465 Mind-body problem, 36, 466, 468 Mindfulness, 24, 25, 166, 167, 202–206, 404, 429–431, 433, 460, 473 Mindfulness of breathing, 24, 25 Mind-independence, 274 Mineness, 125 See also For-me-ness Minimal self, 125–127 Minsky, Marvin, 149 Mīmāṃsā, Bhaṭa, 8, 22 Moderate Claim, 450, 451 Modus tollens argument, 251 Mohanty, J.N., 3, 4, 8, 479 Momentariness, 65, 118, 170, 172, 187, 215, 220, 286, 291, 329 Momentary, 65, 77, 104, 118, 128, 173, 187, 198, 200, 212, 220, 225, 235, 354, 410, 464, 466, 471 awareness, 128, 172, 181 event, 126, 165, 171, 172, 185 object, 53, 172 quality-particulars, 171, 173, 181 Monism, 3 Mookerjee, Satkari, 104 Moral nihilism, 452 Moral phenomenology, 219, 393 Moral psychology, Greek, 394

491 Moral responsibility, 4, 6, 13, 51, 88, 112, 114, 115, 396, 460 Moran, Richard, 266 Mortality paradox, 156 Motion, 36, 144, 215, 216, 220–222, 224, 273, 463478 Motivational internalism, 448, 449 Much, Michael Torsten, 182 muditā, 393, 405 Mūlamadhyamakakārikā, 3, 236, 263, 275, 277, 278, 363, 378, 380, 478 Multiple realizability, 29, 451 Murdoch, Iris, 388, 390, 403, 408 Murti, T.R.V., 8 N Nadler, Steven, 70 Nāgārjuna, 3, 117, 130, 211, 235, 247, 261, 283, 295, 309, 363, 393, 463 Nāgasena, 43, 50, 52, 106, 464 Nagel, Thomas, 100 Nahmias, Eddie, 51 Naïve realism, 149, 161, 244 NaiyĀyika, 4, 104, 112, 283, 331, 332, 336, 341 Narratives, 41, 122, 124, 131, 136, 204, 389, 390, 396, 403–408, 410, 442, 464, 472 Nattier, Ian, 443 Naturalism, 5–12, 142, 242, 390, 395, 397 Naturalism, neo-Aristotelian, 390 Near-death experiences, 143 Neo-Confucianism, 302, 432 Net of Indra, 301, 302 Neural networks, 121 Neurath, Otto, 36 Nichols, Shaun, 13, 105–119, 152, 153, 155 Nicomachean Ethics, 399, 422 Nida-Rümelin, Martine, 37 Nietzsche, Friedrich, 68, 96, 97 nigamana, 329 Nihilism, 3, 6, 44, 48, 49, 94, 95, 105–112, 116, 216, 222, 224, 226, 264, 269, 275, 278, 347, 448, 449, 451, 452, 469 Nihilism, about morality, 449 nikāyas, 168, 171, 178, 365 nirodha, 124 nirodha-samāpatti, 123 See also Cessation nirvāṇa, 464, 466, 468, 470–472 nirvikalpa jñāna, 165 Nishida Kitarō, 3 Noë, Alva, 207

492 Nominalism, 181, 192, 209 Non-conceptual, 2, 37, 47, 123, 165–210, 239, 265, 266, 410 Nonconceptual content, 166, 175, 181, 192–194, 196, 197 Nonconceptual experience, 123, 165, 167, 196, 206 Nonconceptualism, 167, 192–198 Nonconceptualist account of emptiness, 364, 380 Nonconstructed objects, 145 Non-dualism, 261 Nonduality, 201, 465, 466 Non-egological, 53 Non-existence, imagining, 152–157 Non-reductionism, 49, 108, 268, 455 Non-reductive physicalism, 49 Non-self, 122, 127, 129, 157, 404, 468 See also No-self; anātman Non-substantiality, 27 No ownership, 35, 103 Normative significance, 449–451 No-self, 2, 4, 8, 11, 40, 42–50, 53, 57, 76, 88, 105, 111, 112, 118, 125, 126, 136, 388, 400, 411, 430–433, 446, 448, 452, 459–462, 466–474 See also Non-self Nothingness, 3, 301, 404 Nozick, Robert, 161, 464, 474 Nussbaum, Martha, 392, 393, 407, 426, 427, 432 Ñyāṇamoli, Bhikkhu, 91 Nyāya, 3, 97, 100, 103, 112, 125, 126, 138, 166, 252, 283, 328, 329, 332, 334, 336, 337, 423, 455 Nyāyasūtra (NS), 94, 328, 330 Nyāyasūtrabhāṣya (NSB), 330 O Object, 4, 19, 40, 68, 95, 107, 122, 145, 167, 211, 235, 248, 267, 284, 296, 308, 327, 346, 399, 447, 459 of experience, 122 external, 172 inanimate, 45 insentient, 53, 123 material, 78, 84, 171, 291 non-constructed, 145 physical, 40, 72, 78, 79, 95, 146, 290 Objective aspect, 51, 54 Oetke, Claus, 88, 103 Ontological dependence, 94, 100, 238, 239

Index Ontological paradox, 296–298 Ontological parity, 232, 239, 243 Other mind, 94 Overdetermination, 48, 49 Ownerless Suffering Argument, 438, 449–451, 454, 455 Ownership view, 37 P Pain, 7, 23, 24, 52, 55, 76, 77, 87, 88, 112, 128, 168, 202–204, 234, 273, 290, 395, 401, 402, 426, 427 Paleocompatibilism, 4, 5 Pāṇini, 277 Panpsychism, 55, 83 parabhāva, 235, 238, 377–379 Paradox, 45, 156, 207, 235, 250–253, 256, 262, 296–298, 317, 430 paramārtha, 92, 97, 98, 264, 267 paramitā, 127, 393 parārthānumāna, 332–334 Parfit, Derek, 10, 35–43, 45–47, 49, 51, 56, 57, 87, 90–96, 104, 122, 137, 268, 417, 438, 440, 449–452 Parmenides, 403 Partial limb replacement, 134 Paryudāsa, 366, 367 Past-life regressions, 143 Patañjali, 428 Perception, categorical, 176–179, 181, 187 Perceptual experience, 155, 195, 197 Perceptual object, 173, 181, 191, 461 Perduring agent, 136 Pereboom, Derk, 460 Persistence, 11, 14, 35, 40, 42, 58, 65, 66, 118, 270, 452 Personal identity, 10, 11, 35–43, 45, 47, 50–53, 57, 58, 68, 91, 97, 125, 234, 270, 273, 450, 452, 455, 461 Personal identity, physical criterion of, 40 Personalism, 13, 40, 42–51, 57, 265–274 Personalist view, 46, 47, 105, 270 Perspectivalism, 469–472 Perspective first-person, 28, 46, 125, 156, 157 inter-personal, 28 third person, 28, 30, 33, 153, 156 Pervasion, 12, 328, 329, 331–336, 341–342 See also vyāpti Phaedo, 402, 409 Phenomenal aspect, 110, 186 Phenomenal consciousness, 53, 57

Index Phenomenal content, 189, 190 Phenomenal feel, 448 Phenomenal form, 186 Phenomenal primitive, 52 Phenomenal self, 11 Phenomenology, 10, 32, 33, 66, 103, 107, 125, 219, 393, 400, 408, 460, 462, 466 Philebus, 388, 398, 399, 402, 409, 410 Phillips, Stephen, 166, 330 Physicalism, 7, 10, 28, 49, 137, 147, 449, 451 Physicalist stance, 9 Platonism, 14, 387, 388 Plato’s cave, 101, 123 Pleasure, 128, 234, 393, 395, 398, 402, 419, 426, 427, 429 Pluralism, conceptual-explanatory, 232, 243 Plutarch, 90 Porphyry, 426 Positionality, 101 Potter, Karl, 255 Pragmatism, epistemological, 232, 243–245 Prajñaptitaḥ, 94 Prajñākaramati, 446, 447 Prajñāpāramitā, 130, 250, 256, 262 prajñapti, 95 prajñapti-sat, 95 pramāṇa, 3, 166, 328, 342 Pramāṇasamuccaya (PS), 54, 328, 334, 339 Pramāṇasamuccaya-vṛtti (PSV), 171 prasajya, 366, 367 prasajya negation, 366, 367 prasaṅga, 350 Prāsaṅgika, 251, 252 Prasannapadā, 101, 240, 254 pratijñā, 262, 263, 329, 330 pratiṣṭhāpanā, 329 Pratitīyasamutpādahṛdaya, 219 pratītyasamutpādahṛdayavyākhyāna, 219 Pratītyasamutpāda/Paṭiccasamuppāda, 244, 338 pratyakṣa, 172, 173, 183, 197, 333 See also Perception, categorical Pre-reflective self-awareness, 31, 32 Pre-reflective self-consciousness, 11 Presence, 47, 51, 54, 125, 146, 153, 157, 169, 174, 177, 181, 191, 196, 201, 219, 223, 256, 285, 292, 304, 337–339, 453, 473, 474 Presentism, 286 Priest, Graham, 13, 239, 247–257, 295–300, 302, 303, 363, 365, 376, 382 Priestly, Leonard, 46, 53, 401 Private language, 311

493 Property-particulars, 99 Propriety, 419, 424, 442 Psychological connectedness, 39, 41, 450 Psychological continuity, 38–42, 57, 91, 159, 450 Pudgala, 92, 93, 114, 126 Pudgalavāda, 40, 46, 51, 126, 262, 267, 268 Pudgalavādins, 46, 93, 94, 114, 126, 272, 300, 459, 469–472 Pure consciousness experience, 83 Pyrrhonian Skeptics, 423 Q Qualia, 31, 124 Quality-particulars, 168–171, 173, 174, 177, 181, 182 Quine, W.V.O., 313, 323 R Raghunātha Śiṛomaṇi, 103 Rāmakaṇṭha, 104 Rand, Ayn, 460 Ratié, Isabelle, 182 Rationality, 10, 381, 463 Ratnākaraśānti, 342 Ratnakīrti, 104 Ratnāvalī, 219, 393, 401 Read, Rupert, 14, 307–323 Realism, 7, 13, 49, 161, 210, 212, 226, 237–240, 242, 244, 245, 261, 265, 274, 278, 283–293, 349, 451, 452 Reality, 4, 20, 46, 65, 91, 107, 146, 181, 218, 232, 247, 262, 286, 296, 347, 370, 388, 451, 464 Reality, criterion of, 279 Rebirth, 13, 19, 40, 114, 115, 117, 137, 141–143, 159, 160, 351, 398, 406, 469, 471 Recognition, 4, 148, 168, 169, 172–181, 183–185, 187, 189–191, 194, 196, 197, 208, 209, 410 See also saṃjñā Reconstructivist naturalism, 142 Reductio argument, 213, 218, 250, 251 Reductionism, 4, 20, 35, 91, 106, 169, 225, 231, 260, 345, 417, 450 Buddhist, 7, 13, 35, 36, 42–51, 53, 57, 109, 111, 116, 117, 231–236, 245, 345–360 explanatory, 231, 244 mereological, 106, 111, 271 ontological, 225 physicalist, 260

494 Reductionist view, 6, 35, 37, 42–44, 50, 53, 56, 92, 102, 236, 242, 459 Reduction ontological, 225 semantic, 233, 234 Reductive physicalism, 7, 49 Reflective attitudes, 54 Reflexive cognition, 22 Reflexive consciousness, 20, 30, 31, 53, 55, 56 See also svasaṃvitti Reflexive self-consciousness, 55, 56 Reflexivism, 55, 56 Reification, 14, 23, 32, 199, 202, 210, 264, 308 Relativism, 241, 278 Repetti, Rick, 459–474 Representation, 80, 107, 150–152, 154, 156, 157, 176–179, 184–191, 196–198, 200, 205, 207, 218 Representational framework, 152, 154 Representational interface, 150, 152, 154, 158 Republic, 388, 389, 398–403, 406, 407, 409 Res cogitans, 19, 69, 78 Res extensa, 69 Residual impressions, 186, 188, 192, 197 Robo-Buddha, 9, 22, 25 Robots, 21, 22, 149 Rousseau, Jean-Jacques, 4 Rozemond, Marleen, 66, 70 Ruegg, David Seyford, 363, 365, 367 Rūpa, 98, 99, 137, 232 Russell, Bertrand, 65, 83, 292, 368 Ryle, Gilbert, 312 S sâdhya-dharma (property to be proven), 12, 328, 331, 332, 336 sākāra, 126, 172, 173, 186, 187, 190, 192 See also Aspect Śākyabuddhi, 190 Śālistambakakārikā, 217 Śālistamba-sūtra, 213 sāmānyalakṣana (generic charactristic), 173 Saṃghabhadra, 100 saṃjñā, 52, 110, 169, 174, 178, 180, 181, 232 sāmkarya, 289 saṁtāna, 219 samvṛti, 236, 244, 264 samvṛti-sat, 98, 101, 267 See also Conceiling way Saṃyutta Nikāya, 88, 93, 94, 99, 100 saṅketa, 92, 96

Index Śāntarakṣita, 46, 283, 453 Śāntideva, 13, 391, 393, 399, 405, 431, 433, 437–455, 469 Sartre, Jean-Paul, 55, 56, 125 Sarvāstivāda, 126, 134, 137, 170, 232, 323, 346 sat (being, existence, reality), 264, 267, 273 satya (truth), 13, 247, 264 Sauchelli, Andrea, 40, 348 Sautrāntika, 74, 211–226 Schlick, Moritz, 68 Schmidthausen, Lambert, 401 Searle, John, 367 Second law of thermodynamics, 158 Seeds, 74, 220, 222, 339, 374, 396 Self, 6, 21, 37, 70, 87, 105, 122, 153, 168, 215, 234, 255, 260, 301, 312, 358, 392, 417, 446, 459 See also ātman Self-appropriation, 401 Self-awareness, 30–32, 37 Self-consciousness, 11, 50–56, 58, 153 Self-cultivation, 14, 405, 417–433 Self-deception, 88 Self, experiential, 11, 194 Self-givenness, 125 Self-identity, 124, 135, 137 Self-illumination, 110 Self-knowledge, 54–56 Self, minimal, 125–127 Self-transparency, of mind, 80, 81, 83 Self-transparency thesis, 80, 81 Sellars, Wilfrid, 267, 352 Selves, substantive, 46 Semantic insulation, 5, 44, 345, 350, 352, 355 Semantic interpretation, 6, 7, 10, 400 Semantic memory, 41 Semantic realism, 234, 347 Seneca, 90 Sense of self, 53, 88, 395, 448, 454, 464, 470, 473 Sensorimotor, 8, 180, 208, 209, 462 Sentience, 31, 66, 74, 80, 123 Sharf, Robert, 13, 166, 167, 206 Shear, Joseph, 195, 196 Sherlock Holmes, 284, 356–358 Shoemaker, Sydney, 2, 7, 41 Siderits, Mark, 1, 20, 35, 64, 87, 105, 125, 160, 165, 212, 231, 247, 259, 283–293, 295, 323, 327, 345, 363, 387, 417, 437, 459 Sider, Ted, 48, 347–354, 360 Sidgwick, Henry, 390, 442, 448, 450

Index Simulacra, 147 Simultaneous causation, 171, 172 Śiṣkāsamuccaya, 391 Sixth Meditation, 84 skandha-santāna, 90, 95 Smith, John Edwin, 3 Socrates, 398, 401, 402, 406, 407, 421, 425 Sorabji, Richard, 90, 93, 426 Sorites paradox, 45 Soteriological, 9, 10, 107, 119, 124, 142, 160, 199, 202, 213, 239, 240, 266, 323, 352, 469, 470, 478 Soteriology, 101, 115, 118, 160, 219, 480 Soul, 10, 37, 39, 56, 64, 67, 68, 71, 72, 75, 76, 78, 88, 90–92, 125, 126, 128, 159, 284, 285, 394, 396, 400, 402, 406, 409, 410, 418, 426, 427, 450, 452–455 immortality of the, 284 pellet, 159 Space allocentric, 195 egocentric, 195 Spackman, John, 11, 13, 363–382, 477 Spinoza, Baruch, 29, 67, 68, 78, 79, 84 Spiritual aspirant, 287 Split brain, 57 Śrīdhara, 103, 104 Stalnaker, Robert, 364, 369, 370, 372–374, 379, 419 States of consciousness, 67 Stcherbatsky, Theodore, 43, 92, 96 Stevenson, Ian, 142, 143 Stimulus magnitudes, 176, 187 Stoicism, 396 Storehouse consciousness, 74, 80, 81 See also ālayavijñāna Strawson, Galen, 63–85, 156, 460 Strawson, Peter F., 54, 103, 165, 290, 367–369, 371 Stream of consciousness, 27, 57, 67, 71, 81, 84, 156 Stripping argument, 99 Subject self-conscious, 46, 56 thin, 75–77 Subjective aspect, 29, 32, 52, 54, 55 Subjective character, 11, 54 Subjective experience, 23, 29, 124, 191 Subjectivity, 23, 26–28, 33, 54, 56, 57, 77, 101, 111, 194, 480 Subjectivity, reification of, 23 Subject-object duality, 199, 200, 248 Substance dualism, 27, 63, 64, 142

495 Substance, 27, 29, 39, 56, 63–65, 67–73, 77, 78, 84, 90, 106, 107, 112, 117, 125, 159, 198, 235, 273, 287, 288, 309, 312, 330–332, 340, 354, 364, 395, 450, 452, 453, 455 Suffering, 10, 87–89, 91, 93, 107, 118, 119, 160, 161, 214, 234, 279, 286, 299, 351, 359, 388, 392, 393, 396, 400, 401, 404, 405, 408, 410, 411, 427–430, 432, 438, 439, 444, 446–451, 454, 455, 462, 467, 468, 479 śūnyatā, 3, 166, 198, 255, 256 Superimposition, 177, 200 Supervenience, 46, 50, 159, 348, 351 Suzuki, D.T., 2 svābhāsa, 54 See also Subjective aspect svabhāva, 4, 13, 47, 211, 212, 215, 221, 231, 235–236, 238, 239, 241, 243–245, 263, 264, 269, 278, 296, 297, 346, 364, 371–374, 376–380 See also Intrinsic nature svabhāva-presupposition, 13, 364, 371–376, 380, 381 svārthānumāna, 332, 333 svasaṃvitti, 55 See also Reflexive consciousness Svātantrika, 252 Synchronic unity, 11 T Taber, John, 169, 171, 182 Tanaka, Koji, 13, 295–304, 477 tan boluomi, 127 Target property, 328, 330–332, 335–337, 339, 341, 342 tarka, 350 Tarkabhāṣā, 332–336 Teleportation, 39, 45, 57, 124, 136, 137 Teletransportation, 38 Temporal dependence, 147 Tetralemma, 249, 295, 303, 363, 365–367, 372, 373, 375, 376, 382 Thakchoe, Sonam, 13, 211–226, 477 The Cowherds, 12, 217, 278, 455 The Questions of Milinda, 43, 50 Thomasson, Amie, 44, 55 Thompson, Evan, 11, 13, 20, 26, 30, 121, 138, 165–210, 273, 390, 477, 480 Thought experiment, 57, 83, 121–138, 312, 472 Thurman, Robert, 207

496 Tiantai, 302 Tibetan Buddhism, 166, 206 Tillemans, Tom, 12, 20, 165, 182, 184, 186, 187, 209, 239, 240, 253, 254, 365, 391, 477, 479 Timeaus, 403 Time, Kantian idea of, 155 trairūpya, 330, 339 Trance states, 464, 466, 472 Transcendence, 94, 297, 298 Transcendental argument, 155, 263 condition, 263, 302 feature, 158 stance, 3 unity of apperception, 52 Transparency, 392, 407 Trope, 9, 52, 99, 107, 109, 169, 186, 212, 464 Truth, 5, 41, 68, 96, 107, 122, 151, 168, 211, 231, 247, 259, 302, 321, 345, 363, 398, 423, 449, 460 conventional, 101, 151, 211, 213, 219, 231, 233, 234, 236, 238, 240, 241, 248, 253, 259, 261, 262, 264–267, 273, 346, 349, 351–355, 357–360, 364, 371, 372, 376, 379, 381, 382, 463, 477, 480 correspondence theory of, 379 meta-linguistic, 345 pragmatic conception of, 13, 372 ultimate, 5, 6, 13, 151, 211, 231, 233, 234, 236–238, 241, 248, 250, 253, 256, 262–265, 267, 346–349, 351, 355, 357–360, 363, 364, 366, 371, 375, 376, 379–382, 400, 430, 463 See also satya (truth) Truth-maker, 234–236, 238, 239, 262–265, 272 Tsongkhapa/Tsong kha pa, 376, 462, 469, 471, 472 Turing test, 124 Two truths, 5, 12, 13, 43–45, 98, 101, 106–109, 211, 218, 231, 233, 236, 241, 248, 253, 259–262, 267, 302, 345–347, 363–382, 463, 472 Tzohar, Roy, 199, 206 U udāharaṇa, 329–331 Uddyotakara, 125, 183, 330, 332 Ultimate truth, 5, 6, 13, 151, 211, 231, 233, 234, 236–238, 241, 248, 250, 253, 256, 262–265, 267, 346–349, 351, 355,

Index 357–360, 363, 364, 366, 371, 375, 376, 379–382, 400, 430, 463 Unconscious cognitive processes, 22–25 Unconscious information processes, 23 Unger, Peter, 7, 159 Unified, 52, 55–57, 126, 146, 157, 297, 301–303, 400, 409, 410 Unity, 10, 11, 36, 50–57, 89, 125, 195, 244, 302, 395, 408, 410 of consciousness, 50–57 diachronic, 11, 125 synchronic, 11 Universalizability constraint, 445 Upaniṣad, 255, 256 upādāna, 94, 270, 271, 273, 276, 277 See also Appropriation upādāya, 94, 244, 268, 270, 271, 276, 277 upekkhā, 405 Utilitarianism, 437, 438, 443, 445, 448 Utpaladeva, 200 V Vagueness, 356, 357, 369 Vague terms, 45 Vaibhāṣika, 170–172, 174, 212, 215, 217, 286 Vaidharmya, 330, 331 Vaiśeṣika, 103 Varela, Francisco, 121, 168, 208 vāsanā, 186, 204 Vasubandhu, 43, 46, 47, 87, 91–105, 111–117, 170, 172, 174, 175, 198–200, 212–215, 219–222, 225, 233, 267–270, 283, 286–289, 299, 300, 302, 339, 346, 352, 395, 401, 454, 455 Vātsīputrīya, 46, 114, 138, 267 Vātsyāyana, 94, 125, 330, 331 vedanā, 98, 109–111, 169, 174, 180, 181, 232 Vedānta, 8, 101, 255–256 Vegetative coma, 123 Veil of representation, 150 Veridical cognition, 4, 327 Vetālapañcaviṃśati, 136 vikalpa, 166, 173–175, 177, 191, 198 Vigrahavyāvartanī, 262, 297 vimalakīrti-nirdeśa, 138 Virtual reality, 150, 151 Virtue ethics, 388–398, 444, 445, 448 Aristotelian, 389, 395 Confucian, 390 Greek, 390 viṣayābhasa, 54 See also Intentional aspect

Index viśiṣṭa-pratyakṣa, 333 Visuddhimagga (VM), 43, 91, 111, 178, 388, 404, 405, 428, 431 Vlastos, Gregory, 391, 398, 399, 408 Volition, 52, 95, 109, 393, 431, 461, 462, 467, 468 vyatireka-vyāpti, 341 vyāpaka, 334 vyāpti, 12, 328, 332, 333, 335, 341–342 See also Pervasion W Waldron, William S, 480 Walton, Kendall, 345, 352–357 Warring States Period, 425 Watson, Alex, 88, 104, 317 Wegner, Daniel, 51 Weil, Simone, 390 Westerhoff, Jan, 4, 13, 117, 141–162, 237, 248, 297, 363–367, 374, 395, 477, 480 Whole-body transformation, 134 Williams, Bernard, 38, 45, 97, 421, 432, 442 Williams, Paul, 99, 267, 347, 377, 478, 480 Winch, Peter, 311, 391 Wirklichkeit, 290 Witness, 21, 104, 131, 133, 138, 284, 390

497 Wittgenstein, Ludwig, 2–4, 28, 190, 207, 308–312, 314, 315, 317, 320, 321, 323, 478 Wuqing foxing, 123 wu wei, 206, 462 X Xenoglossy, 143 xin, 420 Xunzi, 418–426, 429, 432 Y Yoga, 428 Yogācāra, 8, 30, 121, 126, 166, 167, 180, 198–201, 204–206, 248, 267, 276, 302, 374 Yogācāra-Sautrāntika, 181, 197 Yoga Sūtra, 428 Yogic perception, 465 Z Zahavi, Dan, 11, 31, 54, 56, 125–127, 138, 194, 477, 480 Zen, 2, 3, 63, 202, 203, 259, 297, 303, 304, 309, 478 Zombies, 21, 22, 26, 124, 463, 480