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 whol